CA2571180A1

CA2571180A1 - Computer systems and methods for constructing biological classifiers and uses thereof

Info

Publication number: CA2571180A1
Application number: CA002571180A
Authority: CA
Inventors: Choong-Chin Liew; Tom Yager; Adam Dempsey; Samuel Chao
Original assignee: Chondrogene Ltd
Current assignee: StageZero Life Sciences Ltd
Priority date: 2004-06-19
Filing date: 2005-06-20
Publication date: 2006-01-05
Also published as: WO2006002240A3; WO2006002240A8; WO2006002240A2; EP1766043A2; EP1766043A4

Abstract

The present invention provides systems and method for constructing classifiers that distinguish between trait subgroups using molecular marker data from blood samples. The invention further encompasses the use of the classifiers and combinations of molecular markers identified by the classifiers in a wide variety of applications including: diagnosis; prognosis; prediction of disease, stage of disease or disease risk; monitoring disease progression and/or regression; monitoring disease reoccurrence and identifying risk of disease reoccurrence; determining and/or predicting response to treatment and/or treatment outcomes; monitoring and/or predicting treatment compliance or non compliance and the like. The invention further provides a variety of selected molecular markers and a means to identify combinations of the selected molecular markers useful for diagnosing particular traits of interest.

Description

DEMANDE OU BREVET VOLUMINEUX
LA PRESENTE PARTIE DE CETTE DEMANDE OU CE BREVET COMPREND
PLUS D'UN TOME.

NOTE : Pour les tomes additionels, veuillez contacter 1e Bureau canadien des brevets JUMBO APPLICATIONS/PATENTS
THIS SECTION OF THE APPLICATION/PATENT CONTAINS MORE THAN ONE
VOLUME

NOTE: For additional volumes, please contact the Canadian Patent Office NOM DU FICHIER / FILE NAME
NOTE POUR LE TOME / VOLUME NOTE:

COIVIPi<1'T~~ S'~'~'lVr~~METHODS FOR CONSTRUCTING BIOLOGICAL
CLASSIFIERS AND USES THEREOF
This application is entitled to and claims priority benefit under 35 U.S.C.
Section 119(e) to U.S. provisional application Number 60/581,312, filed June 19, 2004, U.S.
provisional application Number 60/581,977, filed June 21, 2004, U.S.
provisional application Number 60/643,475, filed January 12, 2005, and U.S. provisional application Number 601663,722, filed March 22, 2005, each of which is incorporated herein by reference in its entirety.
I. FIELD OF THE INVENTION
The field of this invention relates to computer systems and methods to identify classifiers using data obtained from blood. The invention further encompasses the use of the classifiers and combinations of molecular markers identified by the classifiers in a wide variety of applications including: diagnosis; prognosis; prediction of disease, stage of disease or disease risk; monitoring disease progression and/or regression;
monitoring disease reoccurrence and identifying risk of disease reoccurrence; determining and/or predicting response to treatment and/or treatment outcomes; monitoring and/or predicting treatment compliance or non compliance and the like.
TableDESCRIPTION SIZE Date Text File Name Recorded 1A Sequence Related Table96KB June 17, TABLElA.TXT

regarding Comorbid 2005 Hypertension 1B Sequence Related Table102KB June 17, TABLE1B.TXT

regarding Comorbid 2005 Obesity 1 Sequence Related Table49KB June 17, TABLE 1 C.TXT
C

regarding Comoxbid 2005 Allergies 1D Sequence Related TableS9KB June 17, TABLE1DTXT

regarding Comorbid 2005 Systemic Steroids 1E Sequence Related Table204KB June 17, TABLElE.TXT

regarding Hypertension 2005 (Chondro) 1F Sequence Related Table251I~B June 17, TABLE1F.TXT

regarding Obesity (Chondro) 2005 1G Sequence Related Table57KB June 17, TABLE1G.TXT

regarding CoMorbid 2005 Hypertension Only 1H Sequence Related Table23KB June 17, TABLE1H.TXT

regarding Hypertension 2005 OA

Shared 1I Sequence Related Table60KB June 17, TABLEILTXT

regarding Comorbid 2005 Obesity Only 1 J Sequence"l~e~ated 'I'able1 BIB June TABLE 1 J.TXT
17, regarding Obesity OA 2005 Shared lI~ Sequence Related Table 28KB June TABLE1K.TXT
17, regarding Comorbid Allergy 2005 Only 1 L Sequence Related Table 19KB June TABLE 1 L.TXT
17, regarding Allergy OA 2005 Shared 1M Sequence Related Table 40KB June TABLE1M.TXT
17, regarding Comorbid Steroid 2005 Shared 1N Sequence Related Table 23KB June TABLElN.TXT
17, regarding Steroid OA 2005 Shared IO Sequence Related Table 49KB June TABLElO.TXT
17, regarding Differentiating 2005 Systemic Steroids 1P Sequence Related Table 118KB June TABLE1P.TXT
17, regarding Diabetes 2005 1 Q Sequence Related Table 165KB June TABLE1 Q.TXT
17, regarding Hyperlipidemia 2005 1R Sequence Related Table 102KB June TABLE1R.TXT
17, regarding Lung Disease 2005 1 S Sequence Related Table 830KB June TABLE1 S.TXT
17, regarding Bladder Cancer 2005 1T Sequence Related Table 483KB June TABLE1T.TXT
17, regarding Bladder Cancer 2005 Staging IU Sequence Related Table 657KB June TABLElU.TXT
17, regarding Coronary Artery 2005 Disease 1V Sequence Related Table 380T~B June TABLE1V.TXT
17, regarding Rheumatoid 2005 Arthritis 1 W Sequence Related Table 183KB June TABLE1 W.TXT
17, regarding Rheumatoid 2005 Arthritis 1X Sequence Related Table 165KB June TABLE1X.TXT
17, regarding Depression 2005 1Y Sequence Related Table 32KB June TABLElY.TXT
17, regarding OAStaging 2005 1Z Sequence Related Table 19KB June TABLEIZ.TXT
I7, regarding Liver Cancer 2005 1Zb Seqeuence Related Table430KB June TABLE1Z.TXT
17, regarding Liver Cancer 2005 lAA Sequence Related Table 592KB June TABLElAA.TXT
17, regarding Schizophrenia 2005 lAB Sequence Related Table 142I~B June TABLElAB.TXT
17, regarding Chagas Disease 2005 lAC Sequence Related Table 64KB June TABLEIAC.TXT
17, regarding Asthma (Chondro) 2005 lAD Sequence Related Table 57I~B June TABLElAD.TXT
17, regarding Asthma (Affy) 2005 lAE Sequence-Related Table 118KB June TABLE lAE.TXT
17, regarding Lung Cancer 2005 lAG Sequence Related Table 157KB June TABLElAG.TXT
17, regarding Hypertension 2005 (Affymetrix) lAH Sequence Related Table 203KB June TABLElAH.TXT
17, regarding Obesity 2005 (Affymetrix) lAI Sequence Related Table 267KB June TABLElALTXT
17, regarding Ankylosing 2005 Spondylitis (Affy) 2 Sequence Related Table 19KB June TABLE2.TXT
17, regarding OA Only 2005 Subtraction 3A Sequence Related Table 228KB June TABLE3A.TXT
17, regarding Schizophrenia 2005 v.

MDS

3B Sequence Related Table 347KB June TABLE3B.TXT
17, regarding Hepatitis 2005 v. Liver Cancer 3C Sequence Related Table 470KB June TABLE3C.TXT
17, regarding Bladder Cancer 2005 v.

Kidney Cancer 3D Sequence Related Table 556KB June TABLE3D.TXT
17, regarding Bladder Cancer 2005 v.

Testicular Cancer 3E Sequence Related Table 588KB June TABLE3E.TXT
17, regarding Testicular 2005 Cancer v.

Kidney Cancer 3F Sequence Related Table 84KB June TABLE3F.TXT
17, regarding Liver Cancer 2005 v.

Stomach Cancer 3G Sequence Related Table 149KB June TABLE3G.TXT
17, regarding Liver Cancer 2005 v.

Colon Cancer 3H Sequence Related Table 166KB June TABLE3H.TXT
17, regarding Stomach Cancer 2005 v.

Colon Cancer 3I Sequence Related Table 214KB June TABLE3LTXT
17, regarding OA v. RA 2005 3K Sequence Related Table 16KB June TABLE3K.TXT
17, regarding Chagas Disease 2005 v.Heart Failure 3L Sequence Related Table 19KB June TABLE3L.TXT
17, regarding Chagas Disease 2005 v.

CAD

3N Sequence Related Table 13KB June TABLE3N.TXT
17, regarding CAD v. Heart 2005 Failure 3P Sequence Related Table 68KB June TABLE3P.TXT
17, regarding Asymptomatic 2005 Chagas v. Symptomatic Chagas 3Q Sequence Related Table 56KB June TABLE3Q.TXT
17, regarding Alzheimers' 2005 v.

Schizophrenia 3R Sequence Related Table S1KB June TABLE3R.TXT
17, regarding Alzheimers' 2005 v.

Manic Depression 4A Sequence Related Table 538KB June TABLE4A.TXT
17, regarding OA v. Control 2005 (ChondroChip) 4B Sequence Related Table SSOKB June TABLE4B.TXT
17, regarding OA v. Control 2005 (Affy) 4C Sequence Related Table 321KB June TABLE4C.TXT
17, regarding OA mild v. 2005 Control (ChondroChip) 4D Sequence Related Table 587KB June TABLE 4D.TXT
17, regarding OA mild v. 2005 Control (Affy) 4E Sequence Related Table 198KB June TABLE4E.TXT
17, regarding OA moderate 2005 v.

Control (ChondroChip) 4F Sequence Related Table 576KB June TABLE4F.TXT
17, regarding OA moderate 2005 v.

Control (Affy) 4G Sequence Related Table 203KB June TABLE4G.TXT
17, regarding OA marked 2005 v.

Control (ChondroChip) 4H Sequence Related Table 679KB June TABLE4H.TXT
17, regarding OA marked 2005 v.

Control (Affy) 4I Sequence Related Table 291KB June TABLE4LTXT
17, regarding OA severe 2005 v.

Control (ChondroChip) 4J Sequence Related Table 607KB June TABLE4J.TXT
17, regarding OA severe 2005 v.

Control (Affy) 4K Sequence Related Table 113KB June TABLE4K.TXT
17, regarding OA mild v. 2005 moderate (ChondroChip) 4L Sequence Related Table 488KB June TABLE4L.TXT
17, regarding OA mild v. 2005 moderate (Affy) 4M Sequence Related Table 93KB June TABLE4M.TXT
17, regarding OA mild v. 2005 marked (ChondroChip) 4N Sequence Related Table 373KB June TABLE4N.TXT
17, regarding OA mild v. 2005 marked (Affy) 40 Sequence Related Table 177KB June TABLE40.TXT
17, regarding OA mild v. 2005 severe (ChoridroChip) 4P Sequence Related Table 687KB June TABLE4P.TXT
17, regarding OA mild v. 2005 severe (Affy) 4Q Sequence Related Table 103KB June TABLE4Q.TXT
17, regarding OA moderate 200 v.

marked (ChondroChip) 4R Sequence Related Table 450KB June TABLE4R.TXT
17, regarding OA moderate 2005 v.

marked (Affy) 4S Sequence Related Table 79KB June TABLE4S.TXT
17, regarding OA moderate 2005 v.

severe (ChondroChip) 4T Sequence Related Table 627KB June TABLE4T.TXT
17, regarding OA moderate 2005 v.

severe (Affy) 4U Sequence Related Table 66KB June TABLE4U.TXT
17, regarding OA marked 2005 v.

severe (ChondroChip) 4V Sequence Related Table 758KB June TABLE4V.TXT
17, regarding OA marked 2005 v.

severe (Affy) SA Sequence Related Table 80KB June TABLESA.TXT
17, regarding Psoriasis 2005 v. Control SB Sequence Related Table 373KB June TABLESB.TXT
17, regarding Thyroid Disorder 2005 v.

Control SC Sequence Related Table 87KB June TABLESC.TXT
17, regarding Irritable 2005 Bowel Syndrome v. Control SD Sequence Related Table 79KB June TABLESD.TXT
17, regarding Osteoporosis 2005 v.

Control SE Sequence Related Table 231KB June TABLESE.TXT
17, regarding Migraine 2005 Headaches v. Control SF Sequence Related Table 56KB June TABLESF.TXT
17, regarding Eczema v. 2005 Control SG Sequence Related Table 349KB June TABLESG.TXT
17, regarding NASH v. Control 2005 SH Sequence Related Table 268KB June TABLESH.TXT
17, regarding Alzheimers' 2005 v.

Control SI Sequence Related Table 298KB June TABLESLTXT
17, regarding Manic Depression 2005 v. Control SJ Sequence Related Table 45KB June TABLESJ.TXT
17, regarding Crohns' Colitis 2005 v.

Control SK Sequence Related Table 53KB June TABLESK.TXT
17, regarding Chronic Cholecystis 2005 v. Controls SL Sequence Related Table 160KB June TABLESL.TXT
17, regarding Heart Failure 2005 v.

Control SM Sequence Related Table 304KB June TABLESM.TXT
17, regarding Cervical Cancer 2005 v.

Control SN Sequence Related Table 185KB June TABLESN.TXT
17, regarding Stomach Cancer 2005 v.

Control 50 Sequence Related Table 404KB June TABLESO.TXT
17, regarding Kidney Cancer 2005 v.

Control SP Sequence Related Table 486KB June TABLESP.TXT
17, regarding Testicular 2005 Cancer v.

Control SQ Sequence Related Table 380KB June TABLESQ.TXT
17, regarding Colon Cancer 2005 v.

Control SR Sequence Related Table 140KB June TABLESR.TXT
17, regarding Hepatitis 2005 B v.

Control SS Sequence Related Table 177KB June TABLESS.TXT
17, regarding Pancreatic 2005 Cancer v. Control ST Sequence Related Table 63KB June TABLEST.TXT
17, regarding Asymptomatic 2005 Chagas v. Control SU Sequence Related Table 77KB June TABLESU.TXT
17, regarding Symptomatic 2005 Chagas v. Control SV Sequence Related Table 383I~B June TABLESV.TXT
17, regarding Bladder Cancer 2005 v.

Control 6A Sequence Related Table 163KB June TABLE6A.TXT
17, regarding Cancer (all 2005 types) v.

Control 6B Sequence Related Table 73KB June TABLE6B.TXT
17, regarding Cardiovascular 2005 Disease v. Control 6C Sequence Related Table 337I~B Juice TABLE6C.TXT
17, regarding Neurological 2005 Diseases v. Control 7A Sequence Related Table SSKB June TABLE7A.TXT
17, regarding Celebrex~ 2005 v. all Cox inhibitors except Celebrex 7B Sequence Related Table 57KB June TABLE7B.TXT
17, regarding Celebrex~ 2005 v.

Control 7C Sequence Related Table 53KB June TABLE7C.TXT
17, regarding ~'V'ioxx~ v. Control 2005 7D Sequence Related Table 49I~B June 17, TABLE7D.TXT

regarding Vioxx~ v. 2005 All Cox Inhibitors except Vioxx~

7E Sequence Related Table 72I~B June 17, TABLE7E.TXT

regarding NSAIDS v. 2005 Control 7F Sequence Related Table 208KB June 17, TABLE7F.TXT

regarding Cortisone 2005 v. Control 7G Sequence Related Table 316I~B June 17, TABLE7G.TXT

regarding Visco Supplement 2005 v. Control 7H Sequence Related Table 131KB June 17, TABLE7H.TXT

regarding Lipitor~ v. 2005 Control 7I Sequence Related Table 23KB June 17, TABLE7LTXT

regarding Smoker v. 2005 Non-Smoker 2. BACKGROUND OF THE INVENTION
The prior art is deficient in simple, non-invasive and effective methods of identifying molecular markers and the use of said molecular markers for purposes of:
diagnosis; prognosis; prediction of disease, stage of disease or disease risk;
monitor disease progression and/or regression; monitor disease reoccurrence and identifying risk of disease reoccurrence or the like. The prior art is also deficient in simple non-invasive methods of identifying molecular markers and use of said molecular markers to determine andlor predict response to treatment and/or treatment outcomes, monitor and/or predict treatment compliance or non-compliance, etc. Although progress has been made in identifying molecular markers by detecting the products of putative molecular markers using expression arrays in a variety of diagnostic areas and therapeutic areas, such progress has been primarily limited to studying non-blood tissue samples, such as primary tumors, that are difficult to obtain and thus have limited potential as a diagnostic. What is even more unsatisfactory is that retrieval of such tissue samples often requires invasive medical procedures such as surgery. Prediction of response to treatment is also a significant problem. It is well understood that, for many currently recognized treatments, only a small percentage of the population (for example approximately 20-30%) will respond positively. Amongst the remainder of the population, there are those who do not improve, and others who display a negative or toxic response to the treatment.
As a result of these detrimental effects to some, many effective treatments do not get to market. The prior art is thus deficient in simple, non-invasive methods to analyze and predict treatment and response to treatment.

Such drawbacks have made identification of molecular markers unsatisfactorily difficult. See, for example, Alon et al., 1999, Proc. Natl. Acad. Sci. USA 96, pp. 6745-6750; Schummer et al., 1999, Gene 238, pp. 375-385; and van't Veer, 2002, Nature 415, pp. 530-536.
Even where progress has been made in identifying molecular markers by monitoring molecular marker products using expression arrays - whether in blood or using tissue - the techniques utilized merely identify large number of molecular markers two or more of which may be required so as to permit categorizing an unknown sample for diagnosis. It is not clear, however, which of these molecular markers are most useful to accurately diagnose an unknown sample. In addition, techniques currently available in the art are not sufficiently robust (ie high levels of reproducibility) in accordance with scientific and regulatory standards so as to be used reliably to diagnose a test individual.
Thus what is required in the art is a means to select smaller subsets of useful molecular markers which when used in combination permit the accurate and reproducible diagnosis of an unknown sample for a particular trait of interest. Further what is required in the art is a means of translating the molecular marker data from these selected combination s so as to convert these into a diagnosis.
Discussion or citation of a reference herein will not be construed as an admission that such reference is prior art to the present invention.
3. SUMMARY OF THE INVENTION
Thus what is needed in the art is a method to identify useful combinations of molecular markers and a means of using said combinations of molecular markers (or more accurately measurement of the products of said molecular markers) so as to permit diagnosis of a test sample. Embodiments of the present invention address many of the shortcomings and drawbacks found in the prior art by the novel approach of using molecular marker measurement data from blood and methods of processing such data to screen the large numbers of candidate molecular markers in blood so as to identify useful combinations of these molecular markers. Embodiments of the present invention involve the construction of classifiers and use of these classifiers. In addition, embodiments of the invention involve the use of the molecular markers identified by these classifiers to diagnose or otherwise determine whether a test subject has a specific trait of interest.
Blood offers a surprisingly informative alternative to tissues as a source of information.
Blood includes numerous cell types including monocytes, leukocytes, lymphocytes, erythrocytes, platelets, as well as possibly many other cell types. The turnover of cells in the human circulatory system is rapid. As a consequence of continuous interactions between the blood and the body, it has been hypothesized that the changes that occur within the cells or tissues of the body will trigger specific changes in gene expression within blood. See, for example, United States patent application serial No.
10/601,518, filed June 20, 2003, United States patent application serial No.lO/802,875, filed March 12, 2004, United States patent application serial No. 10/809,675, filed March 25, 2004, United States patent application serial No. 10/268,730, filed October 9, 2002, United States patent application serial No. 09/477,148, filed January 4, 2000, and United States patent application serial No. 60/115,125, filed Jan. 6, 1999, which are hereby incorporated herein by reference in their entirety. Thus, blood has the potential to provide a powerful indicator of what is happening in the human body at any given time, but provides unique challenges to harness this vast amount of potential information available. Embodiments of the current invention help address this challenge.
4. BRIEF DESCRIPTION OF THE DRAWINGS
Fig. 1 illustrates a computer system for determining and selecting useful biological classifiers.
Fig. 2 illustrates a method for deriving biological classifiers in accordance with an embodiment of the present invention.
Fig. 3 is a flowchart of a method of applying the classifiers to a patient.
Fig. 4 illustrates a data structure for storing high throughput information for a plurality of molecular markers in accordance with one embodiment of the present invention.
Fig. 5 illustrates a data structure for storing a plurality of classifiers in accordance with one embodiment of the present invention.
Fig. 6 illustrates a patient database for storing data for molecular markers for a plurality of patients in accordance with an embodiment of the present invention.
Fig. 7 illustrates a Receiver Operating Characteristic (ROC) curve that is used to assess the discriminating ability of a molecular marker or a classifier in accordance with one embodiment of the present invention.
Fig. 8 illustrates ROC curves corresponding to two candidate classifiers for osteoarthritis computed in accordance with one embodiment of the present invention.
Descriution of Tables:
Table 1 as a group of tables identifies the molecular markers that are differentially expressed in blood samples from patients with a disease or patients who are co-morbid as compared to blood samples from healthy patients or patients without said disease, or with only one of said co-morbid diseases and also shows the sequences of selected products of the identified molecular markers. Molecular marker data from the molecular markers listed in each table or a subset of these molecular markers can be used can be used in steps 214-218 as outlined in Figure 2B so as to identify classifiers and the combinations of molecular markers which form the classifiers useful in diagnosis.
Table 1A identifies the molecular markers which are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has both osteoarthritis and hypertension as compared with a second trait subgroup wherein each member of the second trait subgroup has neither osteoarthritis nor hypertension using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1B shows the identity of those molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has both osteoarthritis and obesity as compared with a second trait subgroup wherein each member of the second trait subgroup has neither osteoarthritis nor obesity using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1C shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has both osteoarthritis and allergies as compared with a second trait subgroup wherein each member of the second trait subgroup has neither osteoarthritis nor allergies using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1D shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has both osteoarthritis and subject to systemic steroids as compared with normal patients using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1E shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has hypertension as compared to a second trait subgroup wherein each member of the second trait subgroup did not have hypertension using the ChondroChipTM

platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1F shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has obesity as compared to a second trait subgroup wherein each member of the second trait subgroup did not have obesity using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1G shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has hypertension and OA when compared with a second trait subgroup wherein each member of the second trait subgroup have OA only wherein molecular markers identified in Table 1 A have been removed so as to identify molecular markers which are unique to hypertension. The table also shows the sequences of selected products of the identif ed molecular markers.
Table 1H shows the molecular markers which were identified in Table 1A which are shared with those molecular markers differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has both hypertension and OA when compared with a second trait subgroup wherein each member of the second trait subgroup have OA only. The table also shows the sequences of selected products of the identified molecular markers.
Table 1I shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has both obesity and have OA when compared with a second trait subgroup wherein each member of the second trait subgroup have OA only and wherein molecular markers identified in Table 1B have been removed so as to identify molecular markers which are unique to obesity. The table also shows the sequences of selected products of the identified molecular markers.
Table 1J shows the molecular markers identified in Table 1B which are shaxed with those molecular markers differentially expressed in blood samples from patients who are obese and have OA when compared with patients who have OA. The table also shows the sequences of selected products of the identified molecular markers.
Table 1K shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a f rst trait subgroup where each member of the subgroup has both allergies and OA when compared with a second trait subgroup wherein each member of the second trait subgroup have OA only wherein molecular markers identified in Table 1 C have been removed so as to identify molecular markers which are unique to allergies. The table also shows the sequences of selected products of the identified molecular markers.
Table 1L shows the identify of those molecular markers identified in Table 3C
which are shared with those molecular markers differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has both allergies and OA when compared with a second trait subgroup wherein each member of the second trait subgroup having OA only. The table also shows the sequences of selected products of the identified molecular markers.
Table 1M shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup is taking systemic steroids and has OA when compared with a second trait subgroup wherein each member of the second trait subgroup have OA only wherein molecular markers identified in Table 1D have been removed so as to identify molecular markers which are unique to patients on systemic steroids. The table also shows the sequences of selected products of the identified molecular markers.
Table 1N shows the identify of those molecular markers identified in Table 1 D
which are shared with those molecular maxkers differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup who are on systemic steroids and have OA when compared with a second trait subgroup wherein each member of the second trait subgroup have OA only. The table also shows the sequences of selected products of the identified molecular markers.
Table 10 shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup are either taking birth control, on prednisone or on hormone replacement therapy and presenting with OA using the ChondroChipTM platform. The table also shows the sequences of selected products of the' identified molecular markers.
Table 1P shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has both type II diabetes as compared to a second trait subgroup wherein each member of the second trait subgroup does not have type II diabetes using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.

Table 1Q shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Hyperlipidernia as compared to a second trait subgroup wherein each member of the second trait subgroup does not have Hyperlipidemia using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1R shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has lung disease as compared to a second trait subgroup wherein each member of the second trait subgroup does not have lung disease using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1S shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has bladder cancer as compared to a second trait subgroup wherein each member of the second trait subgroup does not have bladder cancer using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1T shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has early stage bladder cancer, late stage bladder cancer with a second trait subgroup wherein each member of the second trait subgroup does not have bladder cancer using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1U shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has coronary artery disease (CAD) as compared to a second trait subgroup wherein each member of the second trait subgroup does not have not having CAD
using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1V shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has rheumatoid arthritis as compared to a second trait subgroup wherein each member of the second trait subgroup does not have rheumatoid arthritis using the ChondroChipTM platform . The table also shows the sequences of selected products of the identified molecular markers.
Table 1W shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has rheumatoid arthritis as compared to a second trait subgroup wherein each member of the second trait subgroup does not have rheumatoid arthritis using the Affymetrix~ platform . The table also shows the sequences of selected products of the identified molecular markers.
Table 1X shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has depression as compared with a second trait subgroup wherein each member of the second trait subgroup does not having depression using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
1 S Table 1Y shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has one of various stages of osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup does not have osteoarthritis using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1Z shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has liver cancer as compared with a second trait subgroup wherein each member of the second trait subgroup does not have liver cancer using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 1Z(B) shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a fixst trait subgroup where each member of the subgroup has liver cancer as compared with a second trait subgroup wherein each member of the second trait subgroup does not have liver cancer using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table lAA shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has schizophrenia as compared with a second trait subgroup wherein each member of the second trait subgroup does not have schizophrenia using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table lAB shows the molecular markers that are differentially expressed in blood samples from a training population compxised of a first trait subgroup where each member of the subgroup has Chagas disease as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Chagas disease using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table lAC shows the molecular markers that are differentially expressed in blood samples from a training population compxised of a first trait subgroup where each member of the subgroup has both asthma and osteoarthritis as compared a second trait subgroup wherein each member of the second trait subgroup has only osteoarthritis using the ChondroChipTM. The table also shows the sequences of selected products of the identified molecular markers.
Table lAD shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has asthma as compared with a second trait subgroup wherein each member of the second trait subgroup does not have asthma using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table lAE shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has lung cancer as compared with a second trait subgroup wherein each member of the second trait subgroup does not have lung cancer using the Affylnetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table lAG shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has hypertension as compared with a second trait subgroup wherein each member of the second trait subgroup does not have hypertension using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.

Table lAH shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has obesity as compared with a second trait subgroup wherein each member of the second trait subgroup does not have obesity using the AffymetrixC~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table lAI shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has ankylosing spondylitis as compared with a second trait subgroup wherein each member of the second trait subgroup does not have ankylosing spondylitis using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 2 shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has either mild or severe OA, but for which molecular markers relevant to asthma, obesity, hypertension, systemic steroids and allergies have been removed. The table also shows the sequences of selected products of the identified molecular markers.
Table 3 is a group of tables wherein each table shows those molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has a f rst disease as compared to blood samples from a second trait subgroup wherein each member of the second trait subgroup has a second disease so as to allow differential diagnosis as between said first and second disease.
Table 3A shows the molecular markers that axe differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has schizophrenia as compared with a second trait subgroup wherein each member of the second trait subgroup has manic depression syndrome (MDS) using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 3B shows the molecular maxkers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has hepatitis as compared with a second trait subgroup wherein each member of the second trait subgroup has liver cancer using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.

Table 3C shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has bladder cancer as compared with a second trait subgroup wherein each member of the second trait subgroup has liver cancer using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 3D shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has bladder cancer as compared with a second trait subgroup wherein each member of the second trait subgroup has testicular cancer using the Affymetrix~ platform.
The table also shows the sequences of selected products of the identified molecular markers.
Table 3E shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has testicular cancer as compared with a second trait subgroup wherein each member of the second trait subgroup has kidney cancer using the Affymetrix~
platform.
The table also shows the sequences of selected products of the identified molecular markers.
Table 3F shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has liver cancer as compared with a second trait subgroup wherein each member of the second trait subgroup has stomach cancer using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 3G shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has liver cancer as compared with a second trait subgroup wherein each member of the second trait subgroup has colon cancer using the Affymetrix~ platform.
The table also shows the sequences of selected products of the identified molecular markers.
Table 3H shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has stomach cancer as compared with a second trait subgroup wherein each member of the second trait subgroup has colon cancer using the Affymetrix~
platform.
The table also shows the sequences of selected products of the identified molecular markers.

Table 3I shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has Rheumatoid Arthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has Osteoarthritis using the Affymetrix~
platform.
The table also shows the sequences of selected products of the identified molecular markers.
Table 3K shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has Chagas Disease as compared with a second trait subgroup wherein each member of the second trait subgroup has Heart Failure using the Affymetrix~
platform.
The table also shows the sequences of selected products of the identified molecular markers.
Table 3L shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the 1 S subgroup has Chagas Disease as compared with a second trait subgroup wherein each member of the second trait subgroup has Coronary Artery Disease using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 3N shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has Coronary Artery Disease as compared with a second trait subgroup wherein each member ofthe second trait subgroup has Heart Failure using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
2S Table 3P shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has Asymptomatic Chagas Disease as compared with a second trait subgroup wherein each member of the second trait subgroup has Symptomatic Chagas Disease using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 3Q shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has Alzheimer's' as compared with a second trait subgroup wherein each member of the second trait subgroup has Schizophrenia using the Affymetrix~
platform.

The table also shows the sequences of selected products of the identified molecular markers.
Table 3R shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has Alzheimer's' as compared with a second trait subgroup wherein each member of the second trait subgroup has Manic Depression Syndrome using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4 tables are those which shows molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has a stage of Osteoarthritis as compared to blood samples from a second trait subgroup wherein each member of the second trait subgroup has a second stage of Osteoarthritis so as to allow monitoring of progression and/or regression of disease. Each table also shows the sequences of selected products of the 1 S identified molecular markers.
Table 4A shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has Osteoaxthritis as compared with a second trait subgroup wherein each member of the second trait subgroup is without Osteoarthritis using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4B shows the molecular markers that axe differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup is without Osteoarthritis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4C shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has mild Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup is without mild Osteoarthritis using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.

Table 4D shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has mild Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup is without Osteoarthritis using the AffymetrixC~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4E shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has moderate Osteoarthritis as compared with patients without Osteoarthritis using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4F shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has moderate Osteoarthritis as compared a second trait subgroup wherein each member of the second trait subgroup is without Osteoarthritis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4G shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has marked Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup is without Osteoarthritis using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4H shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has marked Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup is without Osteoarthritis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4I shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has severe Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup is without Osteoarthritis using the ChondroChipTM

platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4J shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has severe Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup is without Osteoarthritis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4K shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has mild Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has moderate Osteoarthritis using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4L shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has mild Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has moderate Osteoarthritis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4M shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has mild Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has marked Osteoarthritis using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4N shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has mild Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has marked Osteoarthritis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 40 shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has mild Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has severe Osteoarthritis using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4P shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has mild Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has severe Osteoarthritis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4Q shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has moderate Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has marked Osteoarthritis using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4R shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has moderate Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has marked Osteoarthritis using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4S shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has moderate Osteoarthritis as compared with patients a second trait subgroup wherein each member of the second trait subgroup has severe Osteoarthritis using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4T shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has moderate Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has severe Osteoarthritis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.

Table 4U shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has marked Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has severe Osteoarthritis using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 4V shows the molecular markers that are differentially expressed in blood from a training population comprised of a first trait subgroup where each member of the subgroup has marked Osteoarthritis as compared with a second trait subgroup wherein each member of the second trait subgroup has severe Osteoarthritis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 5 tables are those which identify molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup 1 S where each member of the subgroup has a disease or condition of interest as compared to blood samples from a second trait subgroup wherein each member of the second trait subgroup is without said disease or condition. The tables also shows the sequences of selected products of the identified molecular markers.
Table 5A shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has psoriasis as compared with a second trait subgroup wherein each member of the second trait subgroup does not have psoriasis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
2S Table SB shows the molecular markers that axe differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has thyroid disorder as compared with a second trait subgroup wherein each member of the second trait subgroup does not have thyroid disorder using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SC shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has irritable bowel syndrome as compared with a second trait subgroup wherein each member of the second trait subgroup does not have irritable bowel syndrome using the Affymetrix~J platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SD shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has osteoporosis as compared with a second trait subgroup wherein each member of the second trait subgroup does not have osteoporosis using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SE shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has migraine headaches as compared with a second trait subgroup wherein each member of the second trait subgroup does not have migraine headaches using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 5F shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has eczema as compared with a second trait subgroup wherein each member of the second trait subgroup does not have eczema using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SG shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has NASH as compared with a second trait subgroup wherein each member of the second trait subgroup does not have NASH using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 5H shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Alzheimers' disease as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Alzheimer's disease using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SI shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Manic Depression Syndrome as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Manic Depression Syndrome using the AffymetrixC~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SJ shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Crohn's Colitis as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Crohn's Colitis using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SK shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Chronic Cholecystis as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Chronic Cholecystis using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SL shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Heart Failure as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Heart Failure using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 5M shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Cervical Cancer as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Cervical Cancer using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SN shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Stomach Cancer as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Stomach Cancer using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.

Table 50 shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Kidney Cancer as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Kidney Cancer using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SP shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Testicular Cancer as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Testicular Cancer using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SQ shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Colon Cancer as compared a second trait subgroup wherein each member of the second trait subgroup does not have Colon Cancer using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SR shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Hepatitis B as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Hepatitis B using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SS shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Pancreatic Cancer as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Pancreatic Cancer using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table ST shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Asymptomatic Chagas as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Chagas using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SU shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Symptomatic Chagas as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Chagas using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table SV shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Bladder Cancer as compared with patients not having Bladder Cancer using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 6 tables axe those tables which show those molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has any one of a series of related conditions as compared to blood samples a second trait subgroup wherein each member of the second trait subgroup does not have said related conditions. The table also shows the sequences of selected products of the identified molecular markers.
Table 6A shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has Cancer as compared with a second trait subgroup wherein each member of the second trait subgroup does not have Cancer using the Affymetrix~
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 6B shows the molecular markers that axe differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup with Cardiovascular Disease as compared with a second trait subgroup wherein each member of the second trait subgroup does not have a Cardiovascular Disease using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 6C shows the molecular markers that axe differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup has a Neurological Disease as compared with a second trait subgroup wherein each member of the second trait subgroup does not have a Neurological Disease using the Affymetrix~ platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 7 tables are those tables which show molecular markers that are differentially expressed in blood samples from with a condition wherein said condition is a treatment as compared to blood samples from patients without said treatment or with a different said treatment.
Table 7A shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup is taking Celebrex~ as compared a second trait subgroup wherein each member of the second trait subgroup is taking a Cox Inhibitor which was not Celebrex~
using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 7B shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup is taking Celebrex~ as compared with a second trait subgroup wherein each member of the second trait subgroup is not taking Celebrex~ using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 7C shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup is taking Vioxx~ as compared a second trait subgroup wherein each member of the second trait subgroup is not taking Vioxx~ using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified 2S molecular markers.
Table 7D shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a f rst trait subgroup where each member of the subgroup is taking Vioxx~ as compared with a second trait subgroup wherein each member of the second trait subgroup on a Cox inhibitor but not on Vioxx~ using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 7E shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup is taking NSAIDS as compared with patients not on NSAIDS using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 7F shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup is taking Cortisone as compared with a second trait subgroup wherein each member of the second trait subgroup on not on Cortisone using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 7G shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup is taking Visco Supplement as compared with a second trait subgroup wherein each member of the second trait subgroup not on Visco Supplement using the ChondroChipTM platform. The table also shows the sequences of selected products of the identified molecular markers.
Table 7H shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup is taking Lipitor~ as compared with a second trait subgroup wherein each member of the second trait subgroup not on Lipitor~ using the ChondroChipTM
platform.
The table also shows the sequences of selected products of the identified molecular markers.
Table 7I shows the molecular markers that are differentially expressed in blood samples from a training population comprised of a first trait subgroup where each member of the subgroup is who are smokers as compared with a second trait subgroup wherein each member of the second trait subgroup who are not smokers using the ChondroChipTM
platform. The table also shows the sequences of selected products of the identified molecular markers.
To further clarify, Tables 1 AA; 1 AB; 1 AD; 1 AE; 1 AG; 1 AH; 1 AT; 1 S; 1 T;
1 U;
1W; 1Z(b); 3A; 3B; 3C; 3D; 3E; 3F; 3G; 3H; 3I; 3K; 3L; 3P; 3Q; 3R; 4B; 4D; 4F;
4H; 4J;
4L; 4N; 4P; 4R; 4T; 4V; SA; SB; SC; SD; SEE; SF; SG; SH; SI; SJ; SK; SL; SM;
SN; 50;
SP SQ; SR; SS; ST; SU; SV; 6A; 6B; 6C; 7F; and 7G each identify the molecular markers identified using the Affymetrix~ genechip to screen the products of the majority of molecular markers of the human genome in accordance with step 202.
Tables 1A; 1 AC; 1B; 1C; 1D; 1E; 1F; 1G; 1H; 1I; 1J; 1K; 1L; 1M; 1N; 10; 1P;
1Q; 1R; 1V; 1X; 1Y; 1Z; 2; 4A; 4C; 4E; 4G; 4I; 4K; 4M; 40; 4Q; 4S; 4V; 7A; 7B;
7C;

7D; 7E; 7H; and 7I each identify molecular markers identified using our own ChondroChipTM genechip to screen the products of the majority of the molecular markers of the human genome.
Like reference numerals refer to corresponding parts throughout the several views of the drawings.

5. DETAILED DESCRIPTION
The embodiments of the present invention use novel approaches to screen and select molecular markers and develop classifiers that can be used to harness the use of molecular marker data from blood. The present invention thus provides systems and methods for constructing biological classifiers using molecular marker data from blood by providing a method to screen and select from a laxge variety of potential molecular markers so as to identify a small subset of molecular markers. The classifiers and the combinations of molecular markers identified using aspects of the current invention are useful for a wide vaxiety of purposes including: diagnosis; prognosis;
prediction of disease, stage of disease or disease risk; monitoring disease progression and/or regression;
monitoring disease reoccurrence and identifying risk of disease reoccurrence;
determining and/or predicting response to treatment and/or treatment outcomes; monitoring and/or predicting treatment compliance or non-compliance and the like. As used herein, a "condition" includes a mode or state of being including a physical, emotional, psychological or pathological state. A condition can be as a result of both "genetic" (ie genetically inherited) and/or "environmental" factors (ie the result of exposure to internal or external influences). In one embodiment of the invention, a condition is a disease. In another embodiment of the invention, a condition is a stage of a disease. In yet another embodiment of the invention, a condition is a mode or state of being which is not a disease. For example in one embodiment, a condition which is not a disease is a condition resulting from the progression of time. A condition resulting from progression of time can include, but is not limited to: memory loss, Loss of skin elasticity, loss of muscle tone, and loss of sexual desire. In a further embodiment of the invention a condition which is not a disease is the response to treatment. A treatment can include, but is not limited to disease modifying treatments as well as treatments useful in mitigating the symptoms of disease.
For example treatments can include drugs specific for a disease of the invention.
As used herein, the term "data" or "molecular marker data" generally refers to data reflective of the abundance of a product of a molecular marker in blood including either or both of RNA and protein.

As used herein, "diagnosis" includes the ability to determine that an individual has or does not have a specific condition or conditions. Diagnosis also refers to the ability to determine that an individual has one condition or conditions as compared with one or more other condition or conditions. In some embodiments, diagnosis refers to the ability to S demonstrate an increased likelihood that an individual has a specific condition.
"diagnosis" refers to the ability to demonstrate an increased likelihood that an individual has one condition as compared to a second condition. More particularly "diagnosis" refers to a process whereby there is an increased likelihood that an individual is properly characterized as having a condition ("true positive") or is properly characterized as not having a condition (or is properly characterized as having the second condition where the diagnosis is as between two conditions) ("true negative") while minimizing the likelihood that the individual is improperly characterized with said condition ("false positive") or improperly characterized as not being afflicted with said condition (or improperly characterized as having the second condition)("false negative").
As used herein, the term "differential expression" refers to a difference in the level of expression of the RNA and/or protein products of a molecular marker of the invention, as measured by the amount or level of RNA or protein. In reference to RNA, it can include difference in the level of expression of mRNA, and/or one or more spliced variants of mRNA of the biomarker in one sample as compared with the level of expression of the same one or more biomarkers of the invention as measured by the amount or level of RNA, including mRNA and/or one or more spliced variants of mRNA in a second sample.
"Differentially expressed" or "differential expression" can also include a measurement of the protein, or one or more protein variants encoded by a molecular marker of the invention in a sample or population of samples as compared with the amount or level of protein expression, including one or more protein variants of a molecular marker of the invention. Differential expression can be determined as described herein and as would be understood by a person skilled in the art. The term "differentially expressed"
or "changes in the level of expression" refers to an increase or decrease in the measurable expression level of a given product of a molecular marker as measured by the amount of RNA and/or the amount of protein in a sample as compared with the measurable expression level of a given product of the molecular marker in a second sample. The first sample and second sample need not be from different patients, but can be samples from the same patient taken at different time points. The term "differentially expressed" or "changes in the level of expression" can also refer to an increase or decrease in the measurable expression level of a given molecular marker in a population of samples as compared with the measurable expression level of the molecular marker in a second population of samples. As used herein, "differentially expressed" when referring to a single sample can be measured using the ratio of the level of expression of a given molecular marker in said sample as compared with the mean expression level of the given molecular marker of a control population wherein the ratio is not equal to 1Ø Differentially expressed can also be used to include comparing a first population of samples as compared with a second population of samples or a single sample to a population of samples using either a ratio of the level of expression or using p-value. When using p-value, a nucleic acid transcript including hnRNA and mRNA is identified as being differentially expressed as between a first and second population when the p-value is less than 0.1, less than 0.05, less than 0.01, less than 0.005, less than 0.001 etc. When determining differential expression on the basis of the ratio of the level of molecular marker product - expression of an RNA or protein product of the molecular marker is differentially expressed if the ratio of the level of the RNA or protein product in a first sample as compared with that in a second sample is greater than or less than 1Ø For instance, a ratio of greater than 1, for example 1.2, 1.5, 1.7, 2, 3, 4, 10, 20, or a ratio of less than 1, for example 0.8, 0.6, 0.4, 0.2, 0.1. 0.05, of RNA or protein product of a molecular marker would be indicative of differential expression. In another embodiment of the invention, a molecular maxker is differentially expressed if the mean level of expression of a nucleic acid transcript including the hnRNA and/or mRNA
transcript in a first population as compared with its mean level of expression of the transcript in a second population is greater than or less than 1Ø For instance, a ratio of greater than 1, for example 1.2, 1.5, 1.7, 2, 3, 4, 10, 20, or a ratio less than 1, for example 0.8, 0.6, 0.4, 0.2, 0.1. 0.05 would be indicative of differential expression.
In another embodiment of the invention a molecular marker is differentially expressed if the ratio of the level of the hnRNA and/or mRNA transcript in a first sample as compaxed with the mean level of the transcript of the second population is greater than or less than 1.0 and includes for example, a ratio of greater than 1, for instance 1.2, 1.5, 1.7, 2, 3, 4, 10, 20, or a ratio less than 1, for example 0.8, 0.6, 0.4, 0.2, 0.1. 0.05. "Differentially increased expression" refers to 1.1 fold, 1.2 fold, 1.4 fold, 1.6 fold, 1.8 fold, or more, relative to a standard, such as the mean of the expression level of the second population.
"Differentially decreased expression" refers to less than 1.0 fold, 0.8 fold, 0.6 fold, 0.4 fold, 0.2 fold, 0.1 fold or less, relative to a standard, such as the mean of the expression level of the second population.

As used herein, the term "molecular marker" (or sometimes referred to as a "biomarker") refers to a gene or a genetic element. In some embodiments, the molecular marker of interest is identified by the Gene ID (formerly Locus Link ID) as is published by the National Center for Biotechnology Information (NCBI) Database as would be understood by a person skilled in the art.
As used herein, the term "oligonucleotide" is defined as a molecule comprised of two or more deoxyribonucleotides and/ or ribonucleotides, and preferably more than three.
Its exact size will depend upon many factors which, in turn, depend upon the ultimate function and use of the oligonucleotide. The oligonucleotides may be from about 8 to about 1,000 nucleotides long. Although oliognucleotides of 8 to 100 nucleotides are useful in the invention, preferred oligonucleotides range from about 8 to about 15 bases in length, from about 8 to about 20 bases in length, from about 8 to about 25 bases in length, from about 8 to about 30 bases in length, from about 8 to about 40 bases in length or from about 8 to about 50 bases in length.
The term, "primer", as used herein refers to an oligonucleotide, whether occurring naturally as in a purified restriction digest or produced synthetically, which is capable of acting as a point of initiation of synthesis when placed under conditions in which synthesis of a primer extension product, which is complementary to a nucleic acid strand, is induced, i.e., in the presence of nucleotides and an inducing agent such as a DNA
polymerase and at a suitable temperature and pH. The primer may be either single-stranded or double-stranded and must be sufficiently long to prime the synthesis of the desired extension product in the presence of the inducing agent. The exact length of the primer will depend upon many factors, including temperature, source of primer and the method used. For example, for diagnostic applications, depending on the complexity of the target sequence, the oligonucleotide primer typically contains 15-25 or more nucleotides, although it may contain fewer nucleotides. The factors involved in determining the appropriate length of primer are readily known to one of ordinary skill in the art. In general, the design and selection of primers embodied by the instant invention is according to methods that are standard and well known in the art, see Dieffenbach, C.W., Lowe, T.M.J., Dveksler, G.S.
(1995) General Concepts for PCR Primer Design. In: PCR Primer, A Laboratory Manual (Eds. Dieffenbach, C.W, and Dveksler, G.S.) Cold Spring Harbor Laboratory Press, New York, 133-155; Innis, M.A., and Gelfand, D.H. (1990) Optimization of PCRs. In:
PCR
protocols, A Guide to Methods and Applications (Eds. Innis, M.A., Gelfand, D.H., Sninsky, J.J., and White, T.J.) Academic Press, San Diego, 3-12; Sharrocks, A.D. (1994) The design of primers for PCR. In: PCR Technology, Current Innovations (Eds.
Griffin, H.G., and Griffin, A.M, Ed.) CRC Press, London, 5-11.
As used herein, the term "probe" means oligonucleotides and analogs thereof and refers to a range of chemical species that recognise polynucleotide target sequences through hydrogen bonding interactions with the nucleotide bases of the target sequences.
The probe or the target sequences may be single- or double-stranded RNA or single- or double-stranded DNA or a combination of DNA and RNA bases. A probe is at least nucleotides in length and less than the length of a complete gene. A probe may be 10, 20, 30, 50, 75, 100, 150, 200, 250, 400, 500 and up to 2000 nucleotides in length.
Probes can include oligonucleotides modified so as to have a tag which is detectable by fluorescence, chemiluminescence and the like. The probe can also be modified so as to have both a detectable tag and a quencher molecule, for example Taqman~ and Molecular Beacon probes.
As used herein, the term "product of the molecular marker" or "molecular marker product" refers to the RNA or protein found in blood which corresponds to the molecular marker (ie is transcribed from the gene or genetic element or is translated from RNA
which is transcribed from the gene or genetic element). For example, in some embodiments RNA resulting from the molecular marker can include one or more of the following species; hnRNA, mRNA, and/or one or more spliced variants of mRNA.
In some embodiments, proteins resulting from the molecular marker can include any proteins found in blood which correspond to the RNA resulting from the molecular marker.
As used herein, the term "selectively amplified" or "selective amplification", refers to a process whereby one or more copies of a particular target nucleic acid sequence is selectively generated from a template nucleic acid. Selective amplification or selectively amplified is to be compared with amplification in general which can be used as a method in combination with, for example, random primers and an oligodT primer to amplify a population of nucleic acid sequences (e.g. mRNA). Selective amplification is preferably done by the method of polymerase chain reaction (Mullis and Faloona, 1987, Methods Enzymol. 155:335).
As used herein, the term "selectively binds" in the context of proteins encompassed by the invention refers to the specific interaction of any two of a peptide, a protein, a polypeptide, and an antibody, wherein the interaction preferentially occurs as between any two of a peptide, protein, polypeptide and antibody preferentially as compared with any other peptide, protein, polypeptide and antibody. For example, when the two molecules are protein molecules, a structure on the first molecule recognises and binds to a structure on the second molecule, rather than to other proteins. "Selective binding", "Selective binding", as the term is used herein, means that a molecule binds its specific binding partner with at least 2-fold greater affinity, and preferably at least 10-fold, 20-fold, 50-fold, 100-fold or higher affinity than it binds a non-specific molecule.
As used herein "selective hybridization" in the context of this invention refers to a hybridization which occurs as between a polynucleotide encompassed by the invention and an RNA , and its complement thereof (ie a cDNA copy), of the molecular marker of the invention, wherein the hybridization is such that the polynucleotide preferentially binds to the RNA products of the molecular marker of the invention relative to the RNA
products of other molecular markers or other genes in the genome in question. In a preferred embodiment a polynucleotide which "selectively hybridizes" is one which hybridizes with a selectivity of greater than 70%, greater than 80%, greater than 90% and most preferably of 100% (i.e. cross hybridization with other RNA species preferably occurs at less than 30%, less than 20%, less than 10%). As would be understood to a person skilled in the art, a polynucleotide which "selectively hybridizes" to the RNA product of a biomarker of the invention can be determined taking into account the length and composition.
As used herein, "specifically hybridizes", "specific hybridization" refers to hybridization which occurs when two nucleic acid sequences are substantially complementary (at least about 65% complementary over a stretch of at least 14 to 25 nucleotides, preferably at least about 75% complementary, more preferably at least about 90% complementary). See I~anehisa, M., 1984, Nucleic acids Res., 12:203, incorporated herein by reference. As a result, it is expected that a certain degree of mismatch is tolerated. Such mismatch may be small, such as a mono-, di- or tri-nucleotide.
Alternatively, a region of mismatch can encompass loops, which are defined as regions in which there exists a mismatch in an uninterrupted series of four or more nucleotides.
Numerous factors influence the efficiency and selectivity of hybridization of two nucleic acids, for example, the hybridization of a nucleic acid member on an array to a target nucleic acid sequence. These factors include nucleic acid member length, nucleotide sequence and/or composition, hybridization temperature, buffer composition and potential for steric hindrance in the region to which the nucleic acid member is required to hybridize. A positive correlation exists between the nucleic acid length and both the efficiency and accuracy with which a nucleic acid will anneal to a target sequence. In particular, longer sequences have a higher melting temperature (TM) than do shorter ones, and are less likely to be repeated within a given target sequence, thereby minimizing non-specific hybridization. Hybridization temperature varies inversely with nucleic acid member annealing efficiency. Similarly the concentration of organic solvents, e.g., formamide, in a hybridization mixture varies inversely with annealing efficiency, while increases in salt concentration in the hybridization mixture facilitate annealing. Under stringent annealing conditions, longer nucleic acids, hybridize more efficiently than do shorter ones, which are sufficient under more permissive conditions.
As used herein, the term "specifically binds" refers to the interaction of two molecules, e.g., a ligand and a protein or peptide, or an antibody and a protein or peptide wherein the interaction is dependent upon the presence of particular structures on the respective molecules. For example, when the two molecules are protein molecules, a structure on the first molecule recognises and binds to a structure on the second molecule, rather than to proteins in general. "Specific binding", as the term is used herein, means that a molecule binds its specific binding partner with at least 2-fold greater affinity, and preferably at least 10-fold, 20-fold, 50-fold, 100-fold or higher affinity than it binds a non-specific molecule.
As herein used, the term ''standard stringent conditions" and ''stringent conditions"
means hybridization will occur only if there is at least 95% and preferably, at least 97%
identity between the sequences, wherein the region of identity comprises at least 10 nucleotides. In one embodiment, the sequences hybridize under stringent conditions following incubation of the sequences overnight at 42 0 C, followed by stringent washes (0.2X SSC at 65 ~ C). The degree of stringency of washing can be varied by changing the temperature, pH, ionic strength, divalent cation concentration, volume and duration of the washing. For example, the stringency of hybridization may be varied by conducting the hybridization at varying temperatures below the melting temperatures of the probes. The melting temperature of the probe may be calculated using the following formulas:
For oligonucleotide probes, between 14 and 70 nucleotides in length, the melting temperature (Tm) in degrees Celcius may be calculated using the formula:
Tm=~ 1.5+16.6(log [Na+]) + 0.41 (fraction G+C)-(600/N) where N is the length of the oligonucleotide.
For example, the hybridization temperature may be decreased in increments of 5°C
from 6~°C to 42°C in a hybridization buffer having a Na+
concentration of approximately 1M. Following hybridization, the filter may be washed with 2X SSC, 0.5% SDS at the temperature of hybridization. These conditions are considered to be "moderate stringency"
conditions above 50°C and "low stringency" conditions below 50°C. A specific example of "moderate stringency" hybridization conditions is when the above hybridization is conducted at 55°C. A specific example of "low stringency" hybridization conditions is when the above hybridization is conducted at 45°C.
If the hybridization is carried out in a solution containing formamide, the melting temperature of the annealing nucleic acid strands may be calculated using the equation Tm=81.5+16.6(log [Na +]) + 0.41 (fraction G + C)-(0.63% formamide)-(600/N), where N
is the length of the probe.
If the hybridization is carried out in a solution containing formamide, the melting temperature of the annealing nucleic acid strands may be calculated using the equation Tm=81.5+16.6(log [Na +]) + 0.41 (fraction G + C)-(0.63% formamide)-(600/N), where N is the length of the probe.
For example, the hybridization may be carried out in buffers, such as 6X SSC, containing formamide at a temperature of 42°C. In this case, the concentration of formamide in the hybridization buffer may be reduced in 5% increments from 50%
to 0%
to identify clones having decreasing levels of homology to the probe.
Following hybridization, the filter may be washed with 6X SSC, 0.5% SDS at 50 ~C.
Hybridization conditions axe considered to be "moderate stringency" conditions when hybridization fluids are comprised of above 25% formamide and "low stringency" conditions when hybridization fluids are comprised of below 25% formamide. A specific example of "moderate stringency" hybridization conditions is when the above hybridization is conducted at 30% formamide. A specific example of "low stringency"
hybridization conditions is when the above hybridization is conducted at 10% formamide.
As used herein, the term "responder" is used to mean an individual who responds to treatment. The use of the term "responds to treatment" depends upon the context of the treatment and the disease or condition, but in some embodiments indicate a sufficiently effective and safe response by an individual to the administration of treatment.
As used herein, the term "non-responder" is used to mean an individual who does not respond positively to treatment. The use of the phrase "does not respond to treatment"
also depends upon the context of the treatment and the disease or condition, but in some embodiments indicates an ineffective or unsafe response by an individual to the administration of treatment.

As used herein, the term "trait" is a mode or state of being including a physical, emotional, psychological or pathological state. A trait can include both "genetic" and/or "environmentally" influenced factors. The term "genetic factors" means genetically inherited elements which affect one or more traits as a result of the genetic makeup of the individual. The term "environmental factors" includes exposure to internal or external influences including but not limited to medical treatments, non-medical drugs, pollution, environmental toxins, lead poisoning, mercury poisoning, exposure to genetically modified organisms, radioactivity, pesticides, insecticides, cigarette smoke, alcohol, or exercise and can affect abundance of RNA or affect gene expression as a result of epigenetic mutations and/or non genetic mutations. A physiological or pathological trait can include the status with regards to a condition including having a condition including a disease, having risk factors of a disease having a certain stage of disease or having a certain response to treatment or a risk of a certain response to treatment. In some cases a displayed trait can actually be the result of one or more underlying traits. A trait also includes clinically measurable parameters including those clinically measurable parameters which are indicators of state of health or disease. For example, a clinically measurable parameter includes blood pressure, lung capacity, electrolyte level, enzyme levels (e.g.
Serum Glutamic Oxaloacetic Transaminase, alkaline phosphatase, Gamma-Glutamyltransferase or Gamma-Glutamyl Transpeptidase, Lactic dehydrogenase) hormone levels (e.g.
thyroid stimulating hormone); protein levels (e.g. Prostate specific antigen PSA) and the like.
Clinically measurable parameters can include disease specific clinical indicators, for example prostate specific antigen as an indicator of prostate cancer; insulin levels as an indicator of diabetes; thyroid stimulating hormone levels as an indicator of thyroid disorder and the like.
As used herein, the term "trait subgroup" is used to define a group of subjects where each subject has at least one trait or group of traits in common, for example, each subject has a disease, a specific stage of disease, same response to treatment, taking the same drug, etc.
As used herein, the terms "treatment", "treat", and "treating" includes administration of one or more compounds, combination of one or more compounds, application of a non-compound based therapeutic regimen, or any combination thereof where administration includes application of a single treatment, a regiment or course of treatment etc. to reduce or amelioration of the progression, severity and/or duration of a disease or condition and/or the reduction or amelioration of the symptoms of a disease or condition resulting from the use of a treatment and/or treatment regime.
5.1 INVENTIVE SYSTEMS AND ALGORITHMS
Fig. 1 shows an exemplary system according to an embodiment of the invention that supports the functionality described herein. The system is preferably a computer system 10 having:
~ one or more central processors 22;
~ a main non-volatile storage unit 14, for example a hard disk drive, for storing software and data, the storage unit 14 controlled by storage controller 12;
~ a system memory 36, preferably high speed random-access memory (RAM), for storing system control programs, data, and application programs, comprising programs and data loaded from non-volatile storage unit 14; system memory 36 may also include read-only memory (ROM);
~ an optional user interface 32, comprising one or more input devices (e.g., keyboard 28) and a display 26 or other output device;
~ an optional network interface card 20 for connecting to any wired or wireless communication network 34 (e.g., a wide area network such as the Internet);
~ an internal bus 30 for interconnecting the aforementioned elements of the system; and ~ a power source 24 to power the aforementioned elements.
Operation of computer 10 is controlled primarily by operating system 40, which is executed by central processing unit 22. Operating system 40 can be stored in system memory 36. In addition to operating system 40, in a typical implementation, system memory 36 includes various components described below. Those of skill in the art will appreciate that such components can be wholly resident in RAM 36 or non-volatile storage unit 14. Furthermore, at any given time, such components can partially reside both in RAM 36 and non-volatile storage unit 14. Further still, some of the components shown in Fig. 1 as being resident in RAM 36 can be resident in another computer (a remote computer) that is addressable by computer 10 over wide area network 34. It will be appreciated that such a remote computer may physically be resident in the same room as computer 10 or in another physical location. As illustrated in Fig. 1, in one exemplary embodiment of the invention, RAM 36 comprises pro~,rams and data to interact with components in the computer system 10 for confi~gnring:
~ file system 42 for controlling access to the various files and data structures used by embodiments of the present invention;
~ a training population 44 for use in construction of one or more classifiers ;
~ a molecular marker data processing module I- 54 for processing molecular marker data representative of a genome or a portion thereof for members of training population 44;
~ a molecular marker screening module A (56A) for identifying molecular markers whose molecular marker data individually discriminates between two or more trait subgroups of the training population using molecular marker data of module 54;
a molecular marker screening module B (56B) for identifying molecular markers whose molecular marker data do not individually discriminate, but demonstrate ability to differentiate between two or more trait subgroups of the training population when used in combination using molecular marker data of module 54;
a candidate molecular marker data structure 58 for storing information about candidate molecular markers identified by molecular marker candidate screening module 56A and optionally molecular marker candidate screening module 56B;
a second molecular marker data processing module II 61 for processing additional molecular marker data for a selection of the candidate molecular markers identified in screening module 56A and, optionally, screening module 56B for members of a training population;
an outlier selection module 57 for evaluating molecular marker data identified in either module 56A and/or 56B or module 61 so as to remove one or more individuals from the training population as outliers;
a combination module 61-5 which selects combinations of molecular markers from candidate molecular markers identified in module 56A and optionally module 56B
a molecular marker classifier construction module 62 for constructing candidate classifiers from combinations of molecular markers identified by molecular marker combination module 61-5;

a molecular marker classifier evaluation module 64 for evaluating and selecting candidate classifiers constructed by molecular marker construction module 62;
a classifier polling and reporting module 66 for receiving patient or subject molecular marker data and polling one or more classifiers selected by evaluation module 64 in order to determine whether a patient or subject has the disease or trait associated with each of the respective classifiers;
a patient database 68 for storage of molecular marker data for diagnostic, prognostic or predictive use; and ~ a classifier database 70 for storage of one or more classifiers selected by molecular marker classifier evaluation module 64.
As illustrated in Fig. l, computer 10 comprises software program modules and data structures. The data structures either stored in computer 10 or accessible to computer I O
include a training population 44, candidate molecular marker data structures 58, patient database 68, and classifier database 70. Each of these data structures can comprise any form of data storage system including, but not limited to, a flat ASCII or binary file, an Excel spreadsheet, a relational database (e.g. SQL), or an on-line analytical processing (OLAP) database (MDX and/or variants thereof). In some specific embodiments, such data structures are each in the form of one or more databases that include hierarchical structure (e.g., a star schema). In some embodiments, such data structures are each in the form of databases that do not have explicit hierarchy (e.g., dimension tables that are not hierarchically arranged).
In some embodiments, each of the data structures stored or accessible to system 10 are single data structures. In other embodiments, such data structures in fact comprise a plurality of data structures (e.g., databases, files, archives) that may or may not all be hosted by the same computer 10. For example, in some embodiments, training population 44 comprises a plurality of Excel spreadsheets that are stored either on computer 10 and/or on computers, that are addressable by computer 10 across wide area network 34.
In another example, patient database 68 comprises a database that is either stored on computer 10 or is distributed across one or more computers that are addressable by computer 10 across wide area network 34. Section 5.9 describes exemplary architectures for training population 44, candidate molecular marker data structure 58, patient database 68, and/or classifier database 70.

It will be appreciated that many of the modules and data structures illustrated in Fig. 1 can be located on one or more remote computers. For example, some embodiments of the present application are web service-type implementations. In such embodiments, classifier polling and reporting module 66 and other modules can be used by a physician to treat a patient and can reside on a client computer that is in communication with computer via network 34. In some embodiments, for example, classifier polling and reporting module 66 can be an interactive web page.
In some embodiments, training population 44, candidate molecular marker data structure 5~, patient database 68 and/or classifier database 70 and modules (e.g. modules 10 54, 56A, 56B, 57, 61, 61-5, 62, 64, and 66) illustrated in Fig. 1 are on a single computer (computer 10) and in other embodiments one or more of such data structures and module are hosted by one or more remote computers (not shown). Any arrangement of the data structures and software modules illustrated in Fig. 1 on one or more computers is within the scope of the present invention so long as these data structures and software modules are addressable with respect to each other across network 34 or by other electronic means.
Thus, the present invention fully encompasses a broad array of computer systems.
Now that an overview of a system in accordance with one embodiment of the present invention has been described, various advantageous methods in accordance with embodiments of the present invention will now be disclosed in conjunction with Figs. 2 through 5. Figure 2 is a flowchart showing a method of selecting molecular markers and developing one or more classifiers or groups of classifiers according to an embodiment of the invention.
Step 202.
Referring to Fig. 2A, in step 202, molecular marker data reflective of the abundance of each of a plurality of RNA and/or proteins found in the blood for members of training population 44 is obtained using one or more of the techniques as described in Section 5.3 and/or 5.4. In some embodiments, the data is reflective of the abundance of RNA products of the molecular marker. In some embodiments, the RNA products are those expressed in blood. In other embodiments, the RNA products are those which are found in blood, but may not necessarily be expressed in blood (e.g. in instances where sufficient mRNA is transported into the blood to be detected. In some embodiments, the data is reflective of the abundance of protein products of the molecular markers. In some embodiments, the data is reflective of the level of proteins expressed in blood. In other embodiments, the data is reflective of proteins found in blood in sufficient quantity to be detected. Measuring of molecular marker data (ie data reflective of the level of the product of the molecular marker) can be done using those techniques known to persons skilled in the art. Note that in some embodiments, data may be obtained using public sources or other sources of data rather than performing one or more of the techniques described. For example, it is anticipated that databases of microarray data collected from blood may be available in future.
In some embodiments, the molecular marker data for each molecular marker is obtained using the same technique to allow greater comparability. In some instances a priori information is known about all or a portion of such genes and in some embodiments a priori information about such genes is either not known or not considered in step 202. In some embodiments the molecular markers resulting from step are those molecular markers identified in Tables 1 A through to 7I.
Measurement of molecular marker data of a plurality of the molecular markers in the blood of each member of training population 44 can be done using any known technique and preferably is done using large scale techniques which allow for the ability to obtain data for a large number of molecular markers and/or for a large number of individuals quickly and efficiently and at a relatively low cost. For example, microarray techniques and RT-PCR and/or Quantitative RT-PCR can be useful large scale techniques.
For example, a high throughput or large scale technique is a technique which allows one to obtain data for large numbers of genes concurrently e.g.1,000 genes, 5,000 genes, 10,000 genes, 15,000 genes 30,000 genes; or all of the genes of the genome of interest. It is expected that additional techniques are being developed and are also useful in embodiments of the invention to screen large numbers of genes quickly and efficiently.
Training population 44 includes a population of individuals made up of one or more trait subgroups with each individual in such trait subgroups having one or more traits.
In some embodiments, each trait subgroup represented in training population 44 includes molecular marker data from at least 3-4 different subjects. More preferably, each trait subgroup represented in training population 44 includes molecular marker data for at least ten different subjects. Still more preferably each trait subgroup represented in the training population 44 includes molecular marker data for at least 30, 40, 50, 100, 200, 500, 1000 or more subjects.
Each training population is selected to include two or more trait subgroups, each subgroup comprising trait subgroup members. Each of these two or more trait subgroups differs with respect to a trait of interest and/or an aspect of a trait of interest. In one embodiment, the members of each of the trait subgroups have been diagnosed as having or not having the trait of interest by one or more known techniques. In another embodiment, the members of each trait subgroup are diagnosed for having or not having the trait of interest using a well accepted methodology for diagnosing of said trait.
For example, each member of a first trait subgroup of the training population has liver cancer, whereas each member of a second trait subgroup of the training population does not have liver cancer. In another embodiment, each member of a first trait subgroup of the training population has Alzheimer's, whereas each member of a second trait subgroup of the training population has manic depressive disorder, and each member of a third trait subgroup of the training population has schizophrenia, and each member of a fourth trait subgroup does not have any of the above conditions. In another example, the trait of interest is a disease such as prostate cancer, and the aspect of interest is the degree of advancement of the prostate cancer. Thus, each member of the first trait subgroup can be those subjects that have early stage prostate cancer, each member of a second trait subgroup can be those subjects that have later stage pxostate cancer and each member of a third trait subgroup can be those subjects that do not have prostate cancer.
In another example, the trait of interest is responsiveness of individuals having musculoskeletal disorders to a Cox 2 inhibitor. A first trait subgroup is comprised of individuals who are responsive to a treatment, a second trait subgroup is comprised of individuals who are responsive to treatment but demonstrate a toxic side-effect, and a third trait subgroup is comprised of individuals who are nonresponsive to treatment. In another embodiment, one trait subgroup can include those subjects that have not yet undergone treatment but who are later identified as being responders to treatment and a second trait subgroup can include those subjects that have not yet undergone treatment but who are later identified as non responders (e.g. demonstrates a toxic side-effect, demonstrates no improvement in condition, demonstrates a worsening of condition, etc.).
In some embodiments, members of each trait subgroup of the training population 44 are preferably selected such that each trait subgroup of the training population 44 has a similar distribution with respect to at least one, two, three, four, five, six, one or more, two or more, three or more, four or more, five or more, six or more, between one and 1000 other traits. For example, age, sex, body mass index (BMI), genetic variation information (e.g., gene SNP mutations, restriction fragment length polymorphisms, microsatellite markers, restriction fragment length polymorphisms, and presence, absence or characterization of short tandem repeats.), treatment regimens; co-morbidities;
concentrations of metabolites, blood chemistry levels, and/or other indicators of health and/or wellness.
A treatment can include, but is not limited to, disease modifying treatments as well as treatments useful in mitigating the symptoms of disease and includes administration of one or more compounds, combinations of one or more compounds, application of a non-compound based therapeutic regimen, or any combination thereof where administration includes application of a single treatment, a regimen or course of treatment and the like.
For example, treatments can include drugs specific for a disease such as drugs specific for Alzheimer's, cardiovascular disease, manic depression syndrome, schizophrenia, diabetes cancers including liver cancer, testicular cancer, bladder cancer, prostate cancer, kidney cancer, breast cancer, colon cancer, osteoarthritis, rheumatoid arthritis, osteoporosis, ankylosing spondylitis, or any other disease including those listed herein. .
For example, treatments can include but are not limited to administration of VIOXX~, Celebrex~, non-steroidal anti-inflammatory drugs (NSAIDS), cortisone, visco supplement, Lipitor~, Adriamycin~, Cytoxan~, Herceptin~, Nolvadex~, Avastin~, Erbitux~, Fluorouracil~, LargactilC~, Sparine~, Vesprin~, Stelazine~, Fentazine~, Prolixin~, Compazine~, Tindal~, Modecate~, Moditen~, Mellarin, Serentil, Norvane, D, Fluanxol~, Clopixol~, Taractan~, Depixol~, Clopixol~, Haldol~, Haldol~, Decanoate, Orap~, Inapsine~, ImapC~, Semap~, Loxitane~, Daxol~, lithium, anticonvulsants (e.g., carbamazepine), antidepressants, and/or Moban~. More generally, a treatment can include any treatment or drug described in the Compendium of Pharmaceuticals and Specialties, Canadian Pharmaceutical Association; 26th edition, June, 1991; I~rogh, Compendium of Pharmaceuticals and Specialties, Canadian Pharmaceutical Association; 27th edition, April, 1992. In another embodiment, a treatment can include administration of any compound described in the United States Food and Drug Administration list of approved drug products (the "Orange Book") that is found at http://www.mco.edu/research/fda.html.
In some embodiments, molecular marker data is not obtainable from each member of a training population or each member of a trait subgroup (for example, using microarray technology there may be an insufficient signal for one or more molecular markers for any particular member of the training population). Nevertheless, as would be understood by a person skilled in the art, candidate molecular markers can still be selected on the basis of the molecular marker data so long as data is obtainable for a su~cient number of molecular markers from a sufficient number of members of the training population. For example, for each molecular marker, it is sufficient if data is available for at least 75%, 80%, 85%, 90% or 95% of the each trait subgroup of the training population.
Section 5.2 provides details on the types of blood samples from subjects in the training population that can be used to obtain data for molecular markers.
Section 5.2 further provides details on how such blood samples can be obtained. Section 5.2 also provides details on the types of subjects that can be used to form training population 44 and the types of subpopulations that can be used in the training population.
Fig. 1 illustrates the data structure of a training population 44 in accordance with one embodiment of the present invention. There is a record 46 for each subject in training population 44. Each record 46 includes an optional subject identifier 48 for uniquely identifying the subject. Each record 46 includes a molecular marker data file 50 for storage of the molecular marker data measured in step 202. Fig. 4 provides more details on a molecular profile 50 resulting from Molecular Marker data processing module I in accordance with one embodiment of the present invention. The molecular profile 50 of Fig. 4 includes an identifier 302 for each molecular marker 302 tracked by profile 50.
Then, for each respective molecular marker 302 in profile 50, there exists one or more measurements of molecular marker data 304. In some embodiments, more than one data point is measured for molecular markers 302. If more than one data point is measured for a molecular marker, then a statistical measure of central tendency (e.g. mean, median, average etc.) can be computed. Accordingly, such measurements for molecular marker data 304 can be stored in data structure 50.
There exists a trait characterization field 52 for each subject in training population 44. Preferably as many as possible traits of each member of training population 44 are documented in a trait characterization record (52). Documented traits include known condition or clinically measurable parameters; genetic likelihood of disease or condition;
medications both past and current; environmental exposures, ethnicity, age, sex and the like. In some embodiments, training population 44 includes only two trait subgroups and trait characterization 52 is a binary choice between two values, where one value indicates that the corresponding subject belongs in a first trait subgroup and a second value indicates that the subject belongs in a second trait subgroup. In some embodiments, training population 44 is divided into a plurality of lists, where each list in the plurality of lists represents a different trait subgroup. In such embodiments, there is no need for a phenotypic characterization field 52.

Although not illustrated in Fig. 1, in some embodiments, there exists a scoring population in addition to the training population. The scoring population is used to evaluate each of the classifiers derived from the training population. The scoring population is made up of one or more individuals that have at least two trait subgroups in common with the training population. In some embodiments, multiple scoring populations are generated from the training population using one or more resampling or cross validation procedures including: bootstrapping; leave one out; leave n out;
percent split and the like so as to evaluate the classifiers derived from the training population. In preferred embodiments, the members of the scoring population are not the members used in the training population.
In some embodiments, some aspects of step 202 are performed by first data molecular processing module I 54. As such, first data molecular processing module 54 can be a known software program, such as commercially available and/or academically available data processing programs.
Step 204.
In step 204, using the data measured in step 202, individual candidate molecular markers are identified, where the molecular marker data allows the differentiation as between two of the trait subgroups of the training population. In some embodiments, step 204 represents a series of pairwise comparisons, where each pairwise comparison is between molecular marker data for subjects from two different trait subgroups.
In other words, data associated with molecular markers from a population of samples having one aspect of a trait of interest (a first trait subgroup) are compared with a population of samples having a second aspect of a trait of interest (a second trait subgroup) so as to identify molecular markers that are able to differentiate between the two trait subgroups (ie the molecular marker data enables the ability to differentiate between the two trait subgroups.
In instances where more than two trait subgroups are represented by a training population 44, more than two pairwise comparisons can be performed on the molecular marker data to identify lists of candidate molecular markers for each of the possible pairwise comparison.
A number of statistical techniques can be used to perform the pairwise comparisons of step 204. In some embodiments, standard statistical techniques such as a t-test are used.
Methods based on conventional t-tests provide the probability (P) that a difference in measured values for the data of a molecular marker between two different trait subgroups occurs by chance. See, for example, Baldi et al., 2001, Bioinformatics 17, pp.
509-519, 2001, which is hereby incorporated herein by reference in its entirety. The t-test compares the actual difference between two means in relation to the variation in the data (expressed as the standard deviation of the difference between the means). For instance, to determine whether a particular molecular marker discriminates between a first trait subgroup and a second trait subgroup, the mean of the data for the molecular marker in the first trait subgroup is compared to the mean of the data for the molecular marker in the second subgroup in accordance with the t-test. In some embodiments, the molecular marker data is such that it is deemed to discriminate between two trait subgroups when the t-test yields a score that matches or exceeds the p = 0.05 level (95% confidence, "significant confidence"), the p = 0.01 level (99% confidence, "highly significant confidence") or p =
0.001 (99.9% confidence, "very highly significant confidence"). In some embodiments, the t-test is applied with a Bonferroni correction or similar form of correction. In some embodiments, rather than using a t-test, nonparametric equivalents such as the Wald-Wolfowitz runs test, the Mann-Whitney U test, the I~olmogorov-Smirnov two-sample test or ROC are used.
In some embodiments, training population molecular marker data is obtained using a microarray and the Significant Analysis of Microarrays technique of Tusher et al. is used to identify molecular markers whose data discriminates between trait subgroups. See, for example, Tusher et al., 2001, Proc. Natl. Acad. Sci. USA 98, 5116-5121. In some embodiments, Manduchis' algorithms for assigning confidence to differentially expressed genes is used to identify molecular markers whose data discriminates trait subgroups. See, for example, Manduchi et al., 2000, Bioinformatics 16, 685-598.
In some embodiments, there are a number of different trait subgroups represented within the training population. Thus, in such embodiments, application of a series of pairwise t-test can become computationally intensive and prone to underinclusiveness in the identification of discriminating molecular markers resulting for each binary comparison since the number of pairwise comparisons that must be performed grows quickly as a function of the number of trait subgroups present in training population 44.
For example, if there are seven trait subgroups present in training population 44, a total of 21 pairwise (t-test class) test can be performed. In such embodiments, analysis of variance (ANOVA) can be used to simultaneously consider whether the data for a molecular marker produces statistically different means for each of the phenotypic data structures. ANOVA
considers the data for a given molecular marker from each of the trait subgroups present in training population 44 and produces a single number (the F-statistic) that can be evaluated for significance at any desired confidence value (e.g., the p = 0.05 Level, the p = 0.01 level, the p=0.001 level, etc). ANOVA is described, for example, in Draghici, Data Analysis Tools For DNA Mic~oarrays, 2003, Chapman & Hall, CRC Press, New York, pp. I55-I1;7, which is hereby incorporated herein by reference in its entirety. In some embodiments, nonparametric equivalents to ANOVA, such as the Kruskal-Wallis analysis of r~arcks test, the Median test, Friedman's two-way analysis of variance, or the Cochran Q
test, are used to identify molecular markers that have statistically different means as between two of the raft subgroups. In some embodiments, after running ANOVA, a means comparisons test such as Duncan's, Student-Newman-Keuls (SNK), Tukey-Kramer, Tukey's HSD, or Least Significant Difference (LSD) are run to determine which molecular markers have data that statistically differentiates as between one or more of the trait subgroups tested (e.g. has statistically different values in one or more of the trait subgroups tested).
In some embodiments, such tests are performed instead of ANOVA or pairwise t-tests to identify molecular markers that discriminate two or more trait subgroups.
In some embodiments, a parametric test and a nonparametric test are run in step 204. If there are only two trait subgroups being compared (e.g., non-disease versus disease) then a Welch t-test (parametric test) and a Mann-Whitney test are used without multiple test corrections at decreasing p-values (e.g., p< 0.05, p < 0.01, p <
0.005, p <
0.001, etc.) until the number of candidate molecular markers is less than 50, less than 100, less than 150, less than 200, less than 250 etc.. If three or more trait subgroups are being compared, the Kruskal-Wallis (nonpararnetric) and Welch ANOVA (parametric) tests are used. Following the same procedure for instances where only two trait subgroups axe being compared, additional molecular marker sets are produced in conjunction with multiple test corrections. An example of a test correction is the Benjamin-Hochberg false discovery rate. In some embodiments the post-hoc statistical tests for groups of three or more provided in GeneSpring version 6.1+ is used. The post-hoc test provides a means for determining which molecular markers are different between particular trait subgroups.
Such post-hoc tests include the Student Newman Keuls test and the Tukey test.
In some embodiments, data for a test molecular marker from all possible trait subgroups in training population 44 is not used. Rather, similar to the case where pairwise t-tests are used, ANOVA is used to determine whether the data for a molecular marker can discriminate between some or all of the trait subgroups. For example, consider the case in which training population 44 includes five trait subgroups. In one approach, a pairwise t-test can be used to identify molecular markers that statistically discriminate (e.g. p = 0.05) between one of the possible pairs of trait subgroups in the training population. Similarly, ANOVA can be used to identify molecular markers that statistically discriminate between one of the possible triplets in the training population. Alternatively, ANOVA
can be used to identify molecular markers that statistically discriminate between one of the possible quadruplets in the training population.
In some embodiments, either in addition to using statistical methods, or independently of statistical methods such as described herein, processing step 204 to select candidate molecular markers is done by looks for differential data abundances (e.g., differential expression). Such methods differ from those described above in the sense that variance as between the level of expression as between members of the same trait subgroup are not necessarily considered. In order to compare differential expression as between two or more trait subgroups, a statistical measure of central tendency (e.g. mean, median, average etc.) can be computed for each molecular marker within any trait subgroup. In some embodiments, the measure of differential data abundance (or differential expression) for a molecular marker product as between a first trait subgroup and a second trait subgroup is determined by measuring the fold change, for example a fold change of greater than 1.5, 2.0, 2.5, 3.0, 4,0 or higher can be selected.. In another embodiment, the measure of differential data abundance need not be quantified, but molecular markers products which on visual inspection display a clear difference in expression (ie abundance levels) as between two trait subgroups can be selected. To illustrate, consider a population P' that comprises trait subgroups "A" and "B". That is, each member of P' is classified into either subgroup "A" or "B" based on whether or not they exhibit or have a particular trait. In this situation, products of molecular markers that are present in large quantities/abundance in one group ("A" or "B") but not the other group are identified. For instance, molecular markers that strongly express in subgroup "A" but not in subgroup "B" can be identified from the measurements taken for each member of each subgroup in step 202. Likewise, molecular markers that express strongly in subgroup "B" but not subgroup "A" can be identified. These patterns of differential abundance of the products of the molecular markers can be used to identify candidate molecular markers.
For an illustration of this approach, see Glob et al. 1999, Science 286: 531.
In embodiments where molecular markers are identified in select pairwise tests, such as pairwise t-tests or ANOVA tests of subsets of the total number of trait subgroups in training population 44, there might be several different lists of molecular markers. For example, molecular marker list A might include molecular markers whose measured data (including normalized data etc.) discriminates between a first and second trait subgroup as determined by a first pairwise t-test. Molecular marker list B might include molecular markers whose data discriminates between a first and third trait subgroup as determined by a second pairwise t-test. Molecular marker list C might include molecular markers whose data discriminates between a first, second and third trait subgroup as determined by ANOVA. In some embodiments, each candidate molecular marker list 60 is preserved as an independent list in candidate molecular marker data structure 58 (Fig. 1).
In some embodiments, processing step 204 identifies a total of between 10 and 4000 candidate molecular markers for any particular training population. In other embodiments, step 204 identifies a total of between 500 and 2000 candidate molecular markers. In yet another embodiments, processing step 204 identified a total of between 100 and 1000 candidate molecular markers.
In some embodiments, some aspects of step 204 are performed by molecular marker candidate identification module 56A. Additionally, there exist known programs that can perform some of the functionality described in step 204. Such programs include those formerly sold by Silicon Genetics (e.g. GeneSpringTM); now Agilent Technologies.
In some embodiments, candidate molecular markers identified using the above-described statistical or nonstatistical tests are clustered in order to visualize relationships between the genes. For example, in some embodiments GeneSpringTM is used to perform hierarchical clustering using the Spearman correlation statistic. In some embodiments, QT
clustering is used to identify genes that have a similar pattern of expression across the specimens in training population 44. Clustering that can be employed in step 204 is described generally in Section 5.12. Further, Section 5.12 gives examples of some clustering techniques that can be used in step 204. Molecular markers for use in subsequent sections (e.g., for quantitative measurement using methods as described in step 206 below) can be ranked and selected by one or more criteria, including, but not limited to, fold change differences in molecular marker data between two or more trait subgroups, standard deviations of molecular marker data as between two or more trait subgroups using the above-described statistical tests, coefficient of variation, statistical significance (e.g., p-value from ANOVA and/or t-test, or other tests described above), level of expression as determined using molecular marker data, gene function, reproducibility of molecular marker data (includes intra- and inter-experimental), elucidated pathways /
networks of the molecular markers as would be understood by a person skilled in the art (e.g.
selecting a molecular marker on the basis of an understanding of how said gene is known to function in the body) and the like. Thus in some embodiments, molecular markers for use in subsequent sections are chosen on the basis of the p value identified as a result of step 204 as a measure of the likelihood that the molecular marker data can distinguish as between the two trait subgroups and more particularly molecular markers are chosen wherein the p value is less than 0.5; less than 0.1, less than 0.05, less than 0.01, less than 0.005, less than 0.001, less than 0.0005, less than 0.0001, less than 0.00005, less than 0.00001, less than 0.000005, less than 0.000001 etc. In some embodiments, molecular markers for subsequent steps are chosen on the basis of the level of differential expression displayed by the molecular marker products as between the two or more trait subgroups.
Note that in measuring differential fold change in blood, the fold change differences can be quite small, thus in some embodiments, selection of molecular markers is based on a differential fold change where the fold change is greater than 1.2, greater than 1.3, greater than 1.4, greater than 1.5, greater than 1.6, greater than 1.7, greater than 1.8, greater than 1.9, greater than 2.0, greater than 2.1, greater than 2.2, greater than 2.3, greater than 2.4, greater than 2.5, greater than 2.6, greater than 2.7, greater than 2.8, greater than 2.9, greater than 3.0, greater than 3.1, greater than 3.2. greater than 3.3, greater than 3.4 greater than 3.5, greater than 4.0 and the like. In some embodiments, it is helpful to select moleculax markers on a basis of the combination of both p value and fold change as would be understood by a person skilled in the art. Thus in some embodiments, molecular maxkers are first selected as outlined above on the basis of the p value resulting from the molecular marker data and then a subselection of said molecular markers is chosen on the basis of the differential fold change determined from the molecular marker data. In other embodiments, molecular markers are first selected on the basis of differential fold change, and then subselection is made on the basis of p value. In some embodiments, the use of one or more of the selection criteria and subsequent ranking permits the selection of the top 2.5%, 5%, 7.5%, 10%, 12.5%, 15%, 17.5%, 20%, 30%, 40%, 50% or more of the ranked molecular markers for use in subsequent steps. In other embodiments, the selection criteria noted above can be set on the basis of the desired number of selected molecular markers for use in steps 206 and or other steps leading to the selection of the available set of maxkers fox step 2I4.
As would be understood, a selection criteria based on the desired number of selected molecular markers will depend upon the resources available for obtaining the molecular marker data for step 206 and/or the computer resources available for calculating and evaluating classifiers of all or a portion of possible combinations of the selected moleculax markers. In some embodiments, the desired number of selected molecular markers for use in step 214 can be 4,000; 3,000; 2,000; 1,000; 900; 800; 700; 600; 500; 400;
300; 200;
190; 180; 170; 160; 150; 140; 130; 120; 110; 100; 90; 80; 70; 60; 50; 40; 30;
20; 10. The more molecular markers which can be selected for use in step 214; the greater the likelihood of identifying classifier or classifiers which are particularly useful for diagnosis.
In some embodiments, one or more subjects of the training population are identified as outliers and are removed prior to identifying individual candidate molecular markers as described herein. These outlier members can then be removed from the training population prior to proceeding to later steps. As described herein, in one embodiment a neural network is used to identify such outliers. A neural network has a layered structure that includes, at a minimum, a layer of input units (and the bias) connected by a layer of weights to a layer of output units. Such units are also referred to as neurons. For output along a single dimension, the layer of output units includes just one output unit. However, neural networks can handle multiple quantitative responses (outputs along multiple dimensions) in a seamless fashion by providing multiple units in the layer of output units.
In multilayer neural networks, there are input units (input layer), hidden units (hidden layer), and output units (output layer). There is, furthermore, a single bias unit that is connected to each unit other than the input units. Neural networks are described in Duda et al., 2001, Pattern Classification, Second Edition, John Wiley & Sons, Inc., New York; and Hastie et al., 2001, The Elements of Statistical Learning, Springer-Verlag, New York.
The basic approach to the use of neural networks to identify outliers is to start with an untrained network. A training pattern is then presented to the untrained network. This training pattern comprises a training population and, for each respective member of the training population, an association of the respective member with a specific trait subgroup.
Thus, the training pattern specifies measured molecular marker data from blood for molecular markers for each member of a training population as well as an indication as to which trait subgroup each member of the training population belongs. In preferred embodiments, training of the neural network is best achieved when the training population includes members from more than one trait subgroup.
In the training process, individual weights in the neural network are seeded with arbitrary weights and then the molecular marker data for each member of the training population is applied to the input layer. Signals are passed through the neural network and the output determined. The output is used to adjust individual weights. A
neural network trained in this fashion classifies each individual of the training population with respect to one of the input trait subgroups. In typical instances, the initial neural network does not correctly classify each member of the training population. Those individuals in the training population that are misclassified identify and determine an error or criterion function for the initial neural network. This error or criterion function is some scalar function of the trained neural network weights and is minimized when the network outputs match the desired outputs. In other words, the error or criterion function is minimized when the network correctly classifies each member of the training population into the correct trait subgroup. Thus, as part of the training process, the neural network weights are adjusted to reduce this measure of error. For regression, this error can be sum-of squared errors. For classification, this error can be either squared error or cross-entropy (deviation). See, e.g. Hastie et al., 2001, The Elements of Statistical LearhifZg, Springer-Verlag, New York. Those individuals of the training population which are still incorrectly classified by the trained neural network, once training of the network has been completed, are identified as outliers and can be removed prior to proceeding.
In some embodiments, an ensemble of neural networks can be used on the training population and individuals ranked on the basis of the number of times an individual is misclassified by each neural network. In order to create the ensemble, each neural network can differ with respect to the initial seeded weighting. In another embodiment, each neural network can differ on the basis of randomly generated noise added to the molecular marker data for one or more molecular markers of each individual of the training population added to the input layer. In such an embodiment, this randomly generated noise can be applied by changing the amount of each of the measured molecular marker data of each member of the training population by a scaled random amount. When larger amounts of noise are required, the magnitude of the scaled random amount is increased. In any of these embodiments, the number of layers and the number of units in each layer can be adjusted in order to provide optimal results for any given set of conditions. In this manner, one can identify outliers which are misclassified in as many asl0%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or more of the neural networks used.
Step 205.
Step 205 is optional. We have surprisingly found that certain molecular markers whose molecular marker data fails to discriminate individually as between two trait subgroups in step 204, are still more than incrementally useful when utilized in combinations with other candidate molecular markers selected in step 204. In particular we have been able to identify molecular markers whose data fails to individually discriminate as between two trait subgroups in the pairwise comparison of step 204 but contribute to a classifier identified in step 216 from a combination which includes one or more candidate molecular markers identified in step 204 ("combination friendly molecular markers"). Thus, in order to ensure that molecular markers which may be useful in combinations are not removed prematurely, optional step 205 is performed.
In step 205, all or a portion of the molecular markers for which data has been or can be obtained in at least two of the trait subgroups of the training population are utilized ("putative combination molecular markers"). For purposes of step 205, in one embodiment, the data is obtained using a technique which allows for fast and efficient data generation for all of the molecular markers of the genome of interest chosen.
In another embodiment the data is obtained using one or more of the techniques as described in Section 5.3 and/or 5.4. In another embodiment, the data is obtained using microarray technology. In a preferred embodiment, the data used is the data obtained for step 202 In order to identify additional candidate molecular markers, combinations of molecular markers are chosen and a mathematical model applied to the molecular marker data for each molecular marker of the combination resulting in a classifier for each combination. The mathematical model applied can be selected from those defined in Section 5.14. In some embodiments, each possible combination of 2, and/or 3, and/or 4, and/or 5, and/or 6, and/or 7, and/or 8, and/or 9 and/or 10 or more of the putative combination friendly molecular markers are tested. For example, if there are 8,000 putative combination friendly molecular markers, each possible combination of 8 or less molecular markers can be written as follows:
8 000!/((8,000 - 8)! (8~!) ach classifier resulting from each combination is scored as described more fully in step 220. In some embodiments, the classifiers are scored using the training population so as to permit time and cost savings. In other embodiments, the classifiers are scored using a scoring population. In yet other embodiments, the classifiers are scored using other resampling or cross validation procedures so as to generate multiple scoring populations.
Having scored the classifiers, a subset of classifiers are then selected based on the score.
In some embodiments, the subset of classifiers is any number less than the total number of combinations evaluated. In some embodiments, the top 10%, top 20%, top 30%, top 40%, or top 50% of the classifiers generated are chosen. In some embodiments, wherein scoring is done using ROC area under the curve, those classifiers with an ROC
area under the curve of 0.5, 0.6, 0.7, 0.8, 0.9 or 1.0 are selected.
Having selected a number of classifiers, each representing a combination of molecular markers, the number of occurrences of each putative combination friendly molecular marker in the combinations of the selected classifiers are determined. Putative combination friendly molecular markers can then be selected as combination friendly molecular markers so as to used as candidate molecular markers based on the number of reoccurrences of said molecular marker in the selection of combinations evaluated. In one embodiment, 5, 10, 15, 20, 30, 50, 100, 150, 200 or more additional candidate molecular markers not previously selected in step 204 are chosen to proceed to step 206.
In other embodiments, the top 10%, top 20%, top 30%, top 40%, or top 50% of combination friendly molecular markers as determined by reoccurrence statistics are selected to be included in the selected set of candidate molecular markers for purposes of choosing combinations to create classifiers in accordance with steps 214 to 218.
Step 206.
In step 206, second molecular marker data is obtained for the candidate molecular markers identified in step 204. In one embodiment, the second molecular marker data is measured using any technique described in Section 5.3 or 5.4 or equivalents thereof. In another embodiment, the same technique used to measure the first molecular marker data in step 202 is used to obtain the second molecular marker data. In other embodiments, an alternative technique is used to measure the second molecular marker data. In other embodiments, the first molecular marker data is obtained using one of the following techniques: microarray, and/or RT-PCR and the second molecular marker data is obtained using any technique but microarray. In other embodiments, the first molecular marker data is obtained using microarray and the second molecular marker data is obtained using any technique except for microarray. The use of second molecular marker data is preferred because the changes in differential expression or abundance of the product of the molecular marker in blood as between trait subgroups can be as low as a 1.1 fold, 1.2 fold, 1.3 fold, 1.4 fold, 1.5 fold etc. which makes less sensitive and reproducible techniques less reliable. In addition, techniques which are preferable for the data collection to allow large scale screening in step 202 such as microarray have been shown to have significant inherent reproducibility issues with high standard deviation as between experiments. As such it is necessary to obtain second molecular marker data so as to ensure the accuracy of the ultimate classifiers identified.

Techniques utilized to obtaining second molecular marker data for the plurality of molecular marker products are those techniques known to measure abundance of RNA
and/or protein including the techniques described in Section 5.3 and 5.4.
In some embodiments, it is helpful to obtain a third series of molecular marker data (ie third molecular marker data, fourth molecular marker data etc.) which can be molecular marker data of the training population used in steps 202 or of a different training or scoring population using any known technique including those techniques described in sections 5.3 and 5.4. Preferably more expensive and/or time consuming techniques are used once smaller numbers of candidate molecular markers have been identified In some embodiments, some aspects of step 206 are performed by molecular marker data processing module II - 61. The exact nature of the functionality of molecular processing module 61 will depend on the type of measurement assay used in step 206.
However, it is contemplated that module 61 will be used to record measurement values for molecular marker data in a profile similar to molecular marker data processing module I -50, perform any necessary error correction techniques, normalization techniques (e.g., techniques described in Bevington and Robinson, Data Reduction acrd E~~o~
Analysis for the Physical Sciences, Second Edition, WCB/ McGraw-Hill, 1992, etc.) and/or perform any measurement techniques that can be coded in a digital computer .
Step 208.
Step 208 is optional and allows for the selection of individual candidate molecular markers of step 206 which can be removed prior to the process of selecting and evaluating combinations of molecular markers in steps 214/216. In optional step 208, one or more candidate molecular markers in data structure 58 (Fig. 1) are eliminated. In optional step 208, the same types of tests that were performed in step 204 can be performed.
The main difference is that in step 208, the quantitative data measured in step 206 using low throughput methods is used whereas in step 204, the high throughput data measured in step 202 is used. Data measured in step 206 is used in step 208 to validate the candidate molecular markers.
In one specific embodiment, training population 44 consists of a first trait subgroup and a second trait subgroup and step 208 comprises performing a t-test or a nonparametric equivalent of the t-test on each candidate molecular marker using the molecular marker data measured in step 206 to verify for each candidate molecular marker that the molecular marker data differentiates between the first trait subgroup and the second trait subgroup with some measure of statistical confidence. Candidate molecular markers whose molecular marker data are less effective in differentiating between the two trait subgroups (e.g. have a p value that is greater than .OS) are removed from data structure 58 and are no longer considered as candidate molecular markers.
In another specific embodiment, training population 44 consists of a first trait subgroup, a second trait subgroup, and a third trait subgroup. In this specific embodiment, ANOVA or a nonparametric equivalent is performed independently on each candidate molecular marker in data structure 58 to verify that each molecular marker differentiates between the three subgroups using the molecular marker data. Candidate molecular markers whose molecular marker data are less effective in differentiating between the three trait subgroups (e.g. have a p value that is greater than .OS) are removed from data structure 58 and are no longer considered as candidate molecular markers.
In some embodiments, each molecular marker is validated by using the data for each molecular marker to generate a Receiver Operating Characteristic (ROC) curve.
ROC curves are generally discussed in Park et al., Korean J. Radiol. 5, p. 11, which is hereby incorporated herein by reference in its entirety. In one embodiment of the present invention, an ROC curve is computed for each candidate molecular marker in training population 44 using the molecular marker data measured in step 206.
As noted in step 202, training population 44 includes, for each specimen in the training population, an indication 52 as to whether or not the specimen has a particular trait under study.
Each respective ROC curve graphs the True Positive Fraction (TPF) as compared with 1/ the False Positive Fraction (FPF). For example, consider the case in which molecular marker A is being validated and the data for molecular marker A that was measured in step 206 is expression level of the molecular marker A in blood samples from subjects. Table B provides hypothetical values for the abundance of A across training population 44.
Table B: Values for molecular marker data of A across training set 44.
[A] Presence / Absence of Disease 0.54 N

In Table B, each line represents a different specimen in the training population. If the relationship between [A] (data of cellular constituent A) and the presence of disease in subjects in the training population is statistically very significant, all positive results (where specimens have the disease) would be at the top of Table B and all negative results (where biological specimens do not have the disease) at the bottom of the Table B.
To plot the ROC curve corresponding to the test illustrated in Table B, the table is divided into a number of cutoff levels. Then, the sensitivity (TPF) and specificity (TNF) of each cutoff level is computed. Sensitivity and specificity are defined with reference to the decision matrix of Table C.
Table C: Decision matrix.
True Feature Status Test result Positive Negative Total Positive TP FP T+
Negative FN T N T-Total D+ D-In Table C, TP means the number of true positives, FP means the number of false positives, FN means the number of false negatives, and TN means the number of true negatives.
Sensitivity is the proportion of subject with a trait (e.g., a disease or particular biological phenotype) who test positive for the feature. In probability notation sensitivity is P(T~~D~) = TP l (TP+FN). Specificity is the proportion of patients without the trait who test negative for the feature. In probability notation specificity is P(T-~D-) = TN ! (TN +
FP).
The ROC curve is defined as a plot of the sensitivity as the y-coordinate versus 1-specificity (false positive rate) as the x-coordinate. Thus, for Table B, where each line of the Table B represents an independent cutoff level, the following ROC data points are derived.
Table D: ROC data points for Table B.
Ratio Cutoff Level Sensitivity 1-Specificity No row 0 0 First row 0.2 0 First two rows 0.4 0 First three rows 0.6 0 First four rows O.S 0 I~Ratio~Cutoff'~~,evell '. m~~ Sensitivity 1-Specificity First five rows 0.8 0.25 First six rows 1 0.25 First seven rows 1 0.5 First eight rows 1 0.75 First nine rows 1 1 To compute the last row of Table D, the number of TP, FP, FN, and TN are counted in Table B when the condition is imposed that the classifier predicts that no specimen in Table B is positive for presence of the trait (e.g., disease or a particular biological phenotype). This, of course, is not an accurate classifier as reflected in the respective sensitivity and specificity values of 0 and 1. Plotting sensitivity by 1-specificity yields the coordinate (0,0) as illustrated in the last row of Table D. Figure 7 illustrates the ROC curve based upon the data points illustrated in Table D. As illustrated in Fig. 7, a ROC curve begins at coordinate (0,0) and ends at coordinate (1,1).
Once an ROC curve has been computed for a molecular marker, in one embodiment, the area under the ROC curve can be quantified. Generally, an area of 1.0 represents a molecular marker that is a perfect diagnostic indicator of the presence of absence of the trait. Preferably an area of greater than 0.5 is desired for a diagnostic indicator, but it will depend upon the trait of interest. For example measurement of protein PSA levels currently used to diagnose prostate cancer has an ROC of 0.47.
In some embodiments there are as many as fifty candidate molecular markers in data structure 58 at this stage of the inventive method. In some embodiments there are more than fifty candidate molecular markers. In practice, the number of candidate molecular markers that remain can be set to any desired number by raising or lowering the criteria for eliminating molecular maxkers. For example, smaller p values from ANOVA
or t-tests, or larger ROC curve areas can be required if the total number of molecular maxkers is too large.
Steps 210 ahd 212.
Step 210 is optional and allows the additional removal of individual candidate molecular markers of step 206 by evaluating how each individual candidate molecular marker performs within a model which evaluates a combination of molecular markers prior to performing the evaluation of combinations in step 214/216.
In optional step 210 all or a portion of the remaining candidate molecular markers in data structure 58 are used to generate a regression classifier. To compute the regression classifier, measured data from step 206 for the molecular markers in two different trait subgroups in training population 44 are used. In some embodiments, the two different trait subgroups respectively represent a diseased and nondiseased state. In some embodiments, the two different trait subgroups respectively represent a first diseased state (e.g. cancer) and a second unrelated diseased state (e.g., Alzheimer's disease). In some embodiments, the two different trait subgroups represent those subjects that are responsive to drug therapy and those subjects that are not responsive to drug therapy. In still other embodiments, the two different trait subgroups from which molecular marker data is obtained represent data from subjects obtained a priori to treatment, but that have been classified into different trait subgroups on the basis of the ultimate response to treatment.
In some embodiments, the two different trait subgroups respectfully represent two different stages of a disease (e.g., moderate versus advanced).
In some embodiments, data for between ten and thirty candidate molecular markers in the two select trait subgroups is used in the logistic regression. In some embodiments, between twenty and one hundred candidate molecular makers in the two select trait subgroups is used in the logistic regression. In still other embodiments, all the candidate molecular markers in the two select trait subgroups are used in the logistic regression.
In step 210 logistic regression can be used because one of the dependent variables is binary - absence or presence of a particular phenotype. For example, consider the case in which molecular marker data from a first trait subgroup and molecular marker data from a second trait subgroup is used in step 210. The first trait subgroup is characterized by a first disease and the second trait subgroup is characterized by a second disease. In such instances, what can be considered by logistic regression is absence or presence of the first disease in subjects. Alternatively, what can be considered by logistic regression is absence or presence of the second disease in subjects.
In general, the multiple regression equation of interest can be written Y =a+,131X1 +/3zXz + ... ~..~kXk +~
where Y, the dependent variable, is presence (when Y is positive) or absence (when Y is negative) of the trait (e.g., phenotype, condition) associated with the first trait subgroup considered in step 204. This classifier says that the dependent variable Y
depends on k explanatory variables (the measured data values for the k candidate molecular markers from subjects in the first and second trait subgroups in training population 44), plus an error term that encompasses various unspecified omitted factors. In the above-identified classifier, the parameter (31 gauges the effect of the first explanatory variable Xl on the dependent variable Y, holding the other explanatory variables constant.
Similarly, (32 gives the effect of the explanatory variable ~2 on Y, holding the remaining explanatory variables constant.
In general, in the multiple regression procedure, estimates for (3; are obtained by taking into account how uncontrolled changes in other variables influence Y.
Thus, in specific embodiments of the present invention, regression is used to eliminate at least some of the candidate molecular markers rather than relying entirely on the tests described in step 208 because the regression takes into account patterns in which multiple molecular markers influence the dependent variable (absence or presence of a trait) in a concerted fashion.
Because the dependent variable data is binary, logistic regression can be used. The logistic regression classifier is a non-linear transformation of the linear regression. The logistic regression classifier is termed the "logit"
classifier and can be expressed as ln[p/(1-p)]=a+31X1 +,132X2 + ... +,(3x~k+~ or ~/(1- p)] = exp a expR~-x~ expaZ ~Z ~ .. . x expa~~~k expE
where, In is the natural logarithm, loge, where e=2.71828..., p is the probability that the event Y occurs, p(Y=1), (1-p), the probability that the event Y does not occur, p (Y~)0, p/(1-p) is the "odds ratio", ln[p/(1-p)] is the log odds ratio, or "logit", and all other components of the classifier are the same as the general regression equation described above. It will be appreciated by those of skill in the art that the term for a and s can be folded into a single constant. Indeed, in preferred embodiments, a single term is used to represent a and s . The "logistic" distribution is an S-shaped distribution function. The logit distribution constrains the estimated probabilities (p) to lie between 0 and 1.
In some embodiments of the present invention, the logistic regression classifier is fit by maximum likelihood estimation (MLE). In other words, the coefficients (e.g., a, (31, [3 2, ...) are determined by maximum likelihood. A likelihood is a conditional probability (e.g., P(Y~X), the probability of Y given X). The likelihood function (L) measures the probability of observing the particular set of dependent variable values (Y1, YZ, ..., Yn) that occur in the sample data set. It is written as the probability of the product of the dependent variables:
L = Prob (Y1 * Y2 * * * Y") The higher the likelihood function, the higher the probability of observing the Ys in the sample. MLE involves finding the coefficients (a, (31, (3 Z, ...) that makes the log of the likelihood function (LL < 0) as large as possible or -2 times the log of the likelihood function (-2LL) as small as possible. In MLE, some initial estimates of the parameters a, (31, [3 2, . . . are made. Then the likelihood of the data given these parameter estimates is computed. The parameter estimates are improved the likelihood of the data is recalculated.
This process is repeated until the parameter estimates do not change much (for example, a change of less than .0l or .001 in the probability). Examples of logistic regression and fitting logistic regression classifiers are found in Hastie, The Elements of Statistical Lea~v~ing, Springer, New York, 2001, pp. 95-100, which is hereby incorporated by reference in its entirety.
Step 212.
In specific embodiments, all or a portion of the candidate molecular markers are used and the molecular marker data fit using logistic regression. Then, in a stepwise fashion, some of the molecular markers are eliminated from the classifier using backward stepwise regression. Backward stepwise regression begins with a full or saturated classifier and variables are eliminated from the classifier in an iterative process. The fit of the classifier is tested after the elimination of each variable (molecular marker) to ensure that the classifier still adequately fits the molecular marker data. When no more variables can be eliminated from the classifier or a desired number of molecular markers remain in the classifier, the analysis has been completed. In specific embodiments, the regression applied in step 210 is used to refine the candidate molecular marker list to less than 25, less than 24, less than 23, less than 22, less than 21, or less than 20 molecular markers.
In one embodiment, a logistic regression classifier is computed using all or a portion of the available candidate molecular markers in data structure 58.
Then, coefficients are tested for significance for inclusion or elimination from the classifier using a Wald test, a likelihood-ratio test (chi-squared statistic), a Hosmer-Lemshow Goodness of Fit Test, or the like. For example, the likelihood-ratio test uses the ratio of the maximized value of the likelihood function for the full classifier (L1) over the maximized value of the likelihood function for the simpler classifier (Lo) in which one or more molecular markers have been removed. The likelihood-ratio test statistic equals:
_clog L~
L, This log transformation of the likelihood functions yields a chi-squared statistic.
Step 213.
Step 2I3 is optional. We have found that performing optional step 213 provides a significant improvement in identifying classifiers which are particularly useful in diagnosis of a disease or condition of interest. In optional step 213, clinically measurable parameters are identified which are thought to be relevant to the trait of interest for which a classifier is desired. For example, where the trait of interest is prostate cancer, clinically measurable parameters chosen are those that are known or are shown to be relevant to the trait of interest. For example in one embodiment, clinically measurable parameters relevant to prostate cancer can include age of subject, level of prostate specific antigen (PSA); and volume of prostate. In yet another embodiment, where the trait of interest is osteoarthritis, some relevant clinically measurable parameters are age and body mass index (BMI). The selected clinically measurable parameters are then included as part of the "selected set of candidate molecular markers" and are treated as molecular markers for the purpose of selecting combinations and developing classifiers in step 214 through 218 as described below.
In order to treat the clinically measurable parameter as a molecular marker for purposes of step 214 through 218, the clinically measurable parameter must have associated data. In some embodiments, where the clinically measurable parameter is one which has an associated value - for example age, blood glucose level, PSA
level, blood pressure, body mass index, etc., the value can be treated as the molecular marker data for purposes of steps 214 through 2I 8. In some embodiments there is no value associated with the clinically measurable parameter, for example where the relevant clinical parameter is determinable but does not provide a value. In those cases, a value can be assigned to represent each aspect of the clinically measurable parameter. For example where the sex of a person is the clinically measurable parameter, a value of 1 can be assigned to represent that the person is male, and a value of 0 can be assigned to represent that the person is female. As yet another example, where the relevant clinically measurable parameter is ethnicity, a different value can be assigned to each ethnicity (e.g.

1 Caucasian, 2 asian, 3ashkanazi jew etc. and said value can be used as the molecular marker data associated with ethnicity for purposes of step 214 through 218.
Step 214.
Steps 214 through 218 provide an approach in which all or a portion of the possible combinations of the selected set of candidate molecular markers resulting from steps 202-213 are chosen. Molecular marker data from each candidate molecular marker in a elected combination is applied to a mathematical model as described more fully in Section 5.14. If there are N selected molecular markers at this stage then, in some embodiments, as many as 2N-1 different combinations can be selected and classifiers can be computed for each of these combinations. For example, consider the case in which three molecular markers are selected after any combination of steps 201 through 212 have been performed and logistic regression is used in step 216. In this case, the following 23-1 mathematical models can be used to form 23-1 corresponding classifiers:
ln[p/(1-p)]-a+31X1 +,132X2 +3X3 +s ln[p/(1-p)]=a+,13IX1 +,32X2 +E
ln[p/(1-p)]=e~+31X1 +3X3+~
ln[p/(1-p)]=a+~ZXZ +N3 X3 +~
ln[p/( 1- p)] = a +,f~1 X1 + ~
ln[p/( 1- p)] = cz + biz Xz + s ~ and ln[p/(1-p)]=a+33X3 +s In these mathematical models, a, (31, (32, ..., (3N represent coefficients that are regressed against molecular marker data whereas Xl, X2, ..., XN each represent a different RNA or protein (or more generally, a molecular marker) for which molecular marker data is available. In some embodiments any one of elements Xl, XZ, ..., XN can represent a clinically measurable parameters. In a preferred embodiment for each combination chosen, at least one of the molecular markers of the series of Xl, XZ, ..., XN
does not represent a clinically measurable parameter. In some embodiments, additional interaction terms are also considered, producing non linear behaviour and resulting in greater than or less than as 2N-1 different combinations. In some embodiments, additional interaction terms are also considered, producing non linear behaviour. For instance, in the example above, another mathematical model to which molecular marker data can be applied in order to form a classifier is:
In[p/(1-p)~ = OG + ~Z~? + ~3~3 + ~4~2~3 + E
where the coefficient j~4 represents the interaction between molecular marker X2 and X3.
In such embodiments, more than 2N-1 "combinations" and thus more than 2N-1 classifiers are considered. In addition to the possibility of interaction terms, the present invention encompasses nonlinear variables. Examples of nonlinear variables include variables that are squared, squared rooted, or in fact, taken to any power. For instance, additional examples of mathematical models to which molecular marker data can be applied include:
In[p/(1-p)~ = a + ~2(~2)Z '~ ~3~3 + ~4~2~3 + ~
In[p/(1-p)~ = oc + ~ia 02)1/2 + a3~3 + ~ja~zX3 + s In some embodiments, a logarithmic or expho'nential function is applied to one or more of the variables. In some embodiments, ratios of molecular marker data can be used as a mathematical model. For example, consider the case in which regression is used to apply molecular marker data to the following equation in order to develop a classifier:
In[p/(1-p)~=G~+1X1 +2X2 +N3 X3 +~
Above, it was noted that Xl, X2, .. ., XN each represent the product of a different gene (or more generally, a molecular marker or molecular marker like element) for which molecular marker data is available. However, in the case where ratios are selected for use in mathematical models which are subsequently scored, each Xl, X2, ..., XN can in fact represent a ratio of abundance and/or expression levels for two different molecular marker products e.g. RNA or proteins, or any other type of molecular marker. For example, Xl can represent the ratio between molecular marker data measured in step 206 representative of gene (molecular marker) A and data for gene B in training population 44.
Mixed forms of mathematical models are also possible. For example, some variables X can represent ratios between the molecular marker data of two molecular markers whereas other variables X can represent molecular marker data of discrete molecular markers as opposed to ratios of molecular marker data. In one specific embodiment a "ratio" of the RNA
products of two molecular markers can be used and all or some of the possible combinations of said "ratios" can be utilized. For example, where the molecular marker data is a measure of abundance of RNA is determined using quantitative RT-PCR, the measure of the expression level of gene (molecular marker) A and the measure of the expression level of gene B can be used as a single term. In some embodiments this is done by first determining the level of expression of each gene individually as compared with an internal housekeeping control such as /3-actin:
e.g. ~Ct = Ctgene A - Ct(3actin where Ctgene A is the threshold cycle of amplification of GeneA and Ct[3actin is the threshold cycle of amplification of the internal control (3-actin.
Similarly the level of expression of gene B is also determined in comparison with an internal housekeeping control e.g. ~Ct = Ctgene B - Ct(iactin In order to combine the terms into a single term for purposes of creating a classifier using "ratios" of the two terms - the terms are combined as follows to form a single variable (e.g. X).
OCt = Ctgene A - Ctgene B
This is commonly described as the use of ratios given the logarithmic nature of the measure of Ct. Thus in some embodiments, the Xl, X2, .. ., X~; of a classifier, each term represents the "ratio" of two molecular markers. In other embodiments the Ct scores are compared directly rather than compared with an internal control.
For example, consider the case in which the desire is to form a classifier that discriminates between a first trait subgroup and a second trait subgroup wherein each term of the classifier represents molecular marker data which is derived from a combination of molecular markers using ratios as outlined above (e.g. OCt = Ctgene A - Ctgene B). In such a case, each variable actually represents two molecular markers - the ratio of the molecular marker data for two molecular markers. Therefore in instances, for example, where 10 molecular markers have been identified by the funneling process of steps 202-210, 45 possible combinations of "ratios" of molecular markers can be formed e.g. n!
(n-2)!2!
where n is the number of possible molecular markers (e.g. 10).
In some embodiments it is particularly useful to select molecular markers where the molecular marker data will work well in the form of a "ratio" thus, as part of step 214, in one embodiment, prior to selecting combinations of molecular markers, molecular markers whose molecular marker data can be combined as ratios are first identified.

~1'he use of molecular marker data as variables in which the variable representing the product of the molecular marker is in the form of ratios or raised to some arbitrary power (e.g., a, 2, N, ete.) is not limited to mathematical models based on regression. Such variables can be used in any of the mathematical models described herein (e.g., neural networks). For example, consider the case in which the desire is to form a classifier that discriminates between a first trait subgroup and a second trait subgroup wherein each term of the classifier is a combination of molecular markers evaluated as a ratio.
In step 214, in one embodiment, prior to selecting ratios of molecular markers, molecular markers which can be combined as ratios are first identified. Molecular markers which can be combined as ratios are those molecular markers wherein the ratio as between said molecular markers is a value which is not equal t 1.0 (or in one embodiment wherein ~Ct = Ctgene a - Ctgene B
does not equal zero. In one embodiment, a first set of molecular markers and a second set of molecular markers are selected to create ratios such that the first set of molecular marker data demonstrates the molecular marker is upregulated in the first trait subgroup (relative to the second trait subgroup) in training population 44 and a second set of molecular marker data demonstrates that the molecular marker is downregulated in the first trait subgroup in training population 44 (relative to the second trait subgroup). Thus, for example an upregulated gene is one in which OCt = Ctgene B - Ctpactin >0 ~d a downregulated gene is one in which OCt = Ctgene A ' Ct~actin <0, Here, the term upregulated or downregulated generally means that such up or down regulation is observed in the training population with some measure of statistical confidence, for example a t-test having ap value of 0.05, less than 0.01, less than 0.005, less than 0.001, less than 0.0005, less than 0.0001, less than 0.00005, less than 0.00001, or less. Then, ratios of molecular markers can be formed using one molecular marker from the first set and a second molecular marker from the second set.
In another embodiment, a ratio for use in a selected combination is one in which the numerator represents the molecular marker data that demonstrates the molecular marker is upregulated in a first trait subgroup (as compared with a second trait subgroup) and the denominator represents the molecular marker data that demonstrates the molecular marker is upregulated in a second trait subgroup (as compared with a first trait subgroup).
With such a ratio, a value greater than "1" indicates that the organism from which the molecular marker data was measured is a member of the first trait subgroup whereas a value less than "1" indicates that this organism is a member of the second trait subgroup.
Thus, in some embodiments, step 214 comprises obtaining combinations of ratios of molecular marker data. In some embodiments, step 214 comprises obtaining some multiple of molecular markers and forming a plurality of ratios of the molecular marker data so as to generate a plurality of combinations of molecular markers.
In step 214, a combination of molecular markers is selected. In some embodiments, this combination of molecular markers consists of a single molecular marker. In some embodiments, this combination of molecular markers comprises 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, as many as 30, as many as 40, as many as 50 or more molecular markers. In some embodiments, this combination of molecular markers consists of a combination of ratios of molecular markers wherein the combination comprises 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, as many as 30, as many as 40, as many as 50 or more molecular markers. For each candidate molecular marker added, the number of possible combinations grows exponentially. The limitation to the number of combinations selected for evaluation is dependent upon the capacity of the computer, network of computers or supercomputers utilized. In one embodiment, all possible combinations of molecular markers resulting from steps 202-213 (or resulting from some subset of steps 202-213) are chosen. In another embodiment, all possible combinations of ratios of molecular markers resulting from steps 202-213 (or resulting from some subset of steps 202-213) are chosen. In another embodiment, one can subject all possible pairs of candidate molecular markers; all possible combinations of three molecular markers, all possible combinations of four molecular markers;
all possible combinations of five molecular markers, all possible combinations of six molecular markers, all possible combinations of seven molecular markers etc. In another embodiment, all possible combinations of two sets of ratios are chosen, in another embodiment, all possible combinations of three sets of ratios are chosen, in another embodiment, all possible combinations of four sets of ratios are chosen, in another embodiment, all possible combinations of five sets of ratios are chosen. Each of the combinations of molecular markers is evaluated in subsequent processing steps.
Step 216.
In step 216, a classifier is computed using each combination of molecular markers chosen in the last instance of step 214 and by applying the classifier to the molecular marker data measured for each molecular marker of this combination of molecular markers to a mathematical model, such as the mathematical models defined in Section 5.14 resulting in one or more classifiers for each combination. As described more thoroughly in Step 204, in some embodiments, one or more subjects of the training population are identified as outliers and are removed prior to computing classifiers fox each combination of molecular markers chosen.
In order to compute a classif er, in some embodiments, the mathematical model is a regression model, a neural network, a clustering model, principal component analysis, nearest neighbor classifier analysis, linear discriminant analysis, quadratic discriminant analysis, a support vector machine, a decision tree, a genetic algorithm, classifier optimization using bagging, classifier optimization using boosting, classifier optimization using the Random Subspace lVlethod, Bayesian Networks (see F. V. Jensen.
"Bayesian Networks and Decision Graphs". Springer. 2001, which is incorporated herein by reference in its entirety), a projection pursuit, weighted voting, a ratio or combination of ratios, or any combination of the above. Representative mathematical models that can be used in the present invention are described in Section 5.14. In the case where the mathematical model is a ratio or combination of ratios, steps 214 and 216 involve using the low throughput molecular marker data from training population 44 that was measured in step 206 to determine which molecular markers should be in the numerator of the ratios and which molecular markers should be in the denominator of the ratios. In some embodiments, the mathematical model used comprises a plurality of ratios of the molecular marker data. In such embodiments, the molecular marker data used in the numerators of the plurality of ratios can be the same or different than the molecular marker data used in the denominators of the plurality of ratios. In other words, a given molecular marker can be represented in the numerator of more than one ratio in the plurality of ratios or represented in the denominator of more than one ratio in the plurality of ratios.
Step 218.
In optional step 218 a determination is made as to whether all of the possible desired combinations of molecular markers to be tested have been considered.
As discussed above in step 214, all or a portion of possible combinations may be tested. If not (218-No) process control returns to step 214 where another combination of molecular markers is selected and, at step 216, this new combination of molecular markers is evaluated using a mathematical model applied to the molecular marker data of the new combination. In some embodiments, the candidate molecular marker list comprises less than 25, less than 24, less than 23, less than 22, less than 21, or less than 20 molecular markers at step 214. In some embodiments, step 218 requires that a classifier be computed for all possible combinations of molecular markers. In other embodiments, step requires that classifiers for only a portion of the possible combinations of molecular markers be considered.
In some embodiments, some aspects of steps 214-218 are performed by molecular marker classifier evaluation module 62. In fact, in some embodiments, several different software programs, such as Microsoft (Redmond, Washington) Excel, are used in steps 214-218.
Step 220.
Once all the desired classifiers have been computed by loop 214-218, the classifiers are evaluated to determine which of the classifiers are most effective. In one embodiment the resulting classifiers of loop 214-218 are scored. In some embodiments, scoring is done using the training population 44. In other embodiments, scoring is done using a "scoring population" wherein the scoring population includes at least some members not present in the training population. In one embodiment, the scoring population includes members of the training population in addition to one or more members not used in the training population. In some embodiments, five percent or less, ten percent or less, twenty percent or less, thirty percent or less, fifty percent or less, or ninety percent or less of the members of the training population are common to the scoring population.
In some embodiments, the Percent Correct Predictions statistic is used to score each classifier. The "Percent Correct Predictions" statistic assumes that if the estimated p is greater than or equal to 0.5, then the event is expected to occur and to not occur otherwise. By assigning these probabilities zeros and ones, a comparison can be made to the values of the samples in the training population to determine what percentage of the training population was sampled correctly.
In some embodiments, ROC analysis is performed and is used to score the classifiers. In one embodiment, the area under the ROC curve is used to judge the quality of the classifier. As would be understood by those of skill in the relevant arts, area under the curve converts the two dimensional information contained in the ROC curve into one dimensional information. In other embodiments, information from the two dimensional aspect of the ROC curve is utilized directly. For example, the ROC curve also provides information with respect to the sensitivity and specificity of the classifier.
In some embodiments, classifiers are selected on the basis of either sensitivity or specificity. This can be an important scoring indicator. For example, a diagnostic classifier with high sensitivity (ie high true positive rate and low false negative rate may be important in situations where it is safer to misdiagnosis an individual as having disease rather than misdiagnosing a disease bearing person as normal. Therefore in some embodiments, a cutoff can be set for either sensitivity or specificity and the classifier ranked or scored on the basis of the remaining variable. In some embodiments, ROC curves are generated for each model computed in an instance of step 216 using data obtained in step 206 for members of either the training population or scoring population or both.
In some embodiments, the classifier resulting from the application of the mathematical itself results in a score as to the accuracy of the model. In this embodiment, the score is based on the accuracy of the model within the training population only.
In some embodiments, a classifier is a weighted logistic regression model characterized by a multicategory logit model. For example, in some embodiments, a classifier discriminates between two different trait groups. In other embodiments, a classifier discriminates between more than two different trait groups. Logit models, including multicategory logit models are described in Agresti, A~ Introduction to Categorical Data Analysis, John Wiley & Sons, Inc., 1996, New York, Chapters 7 and 8, which is hereby incorporated by reference. Table E illustrates the data that is used to form an ROC curve based on expression data applied to a mathematical model that uses the logit:
ln[p/(1-p)]=a+,31X1 +,132X2 +s Table E: Values for the logit ln[p/(1-p)]=a+/.31X1 +,132X2 +s using hypothetical values for training population 44 ln[p/( 1- p)] Presence / Absence of a Trait 0.98 Y

0.97 Y

0.95 Y

0.93 Y

0.91 N
0.11 Y
0.07 N
0.03 N
Each row in Table E corresponds to a different specimen in the scoring population. The left column represents the results of the logit for the classifier being sampled. The specimens in Table E are ranked by the logit score listed in the left hand column. The right hand column details the presence or absence of the trait that is being considered by the regression equation. Table E can be used to compute a ROC curve using the same techniques disclosed in step 208 (in which each row in Table E is considered a threshold cutoff value in order to compute ROC curve datapoints). Then, the area under the ROC
curve can be computed in order to assess the predictive quality of the classifier.
In step 220, each classifier is scored using any of the techniques disclosed here or that are known in the art. The classifiers can then be ranked based on their score. For example, they can be ranked based on the percent correct predictions, area under the ROC
curve, sensitivity or specificity or some weighted or unweighted combination of the two scoring techniques. In some embodiments, step 220 is performed by molecular marker evaluation module 64.
Step 222.
Step 222 is optional. Optional step 222 provides additional filtering to eliminate some of the candidate classifiers computed in loop 214-218. In one such filter, limited to the case in which the classifiers computed in steps 214-218 are based on application of data to regression based mathematical models, classifiers that have at least one coefficient that is large are eliminated. Such classifiers have the potential to magnify small errors in the data. In some embodiments, determination as to whether or not a coefficient is large can require multiple computation steps. In instances where a coefficient uniquely represents a molecular marker, the maximum value (MAX) for the data measured for the molecular marker in a trait subgroup associated with the classifier is identified. For example, consider the case in which a given coeff cient uniquely represents the expression of gene A in blood. Further suppose that low throughput data for gene A from individuals of a particular trait subgroup was measured in step 206. The value MAX
would be the largest expression value observed for gene A in the ten individuals from the subject trait subgroup. For example, if individual #7 in the set of ten individuals exhibited the highest expression level for gene A as determined by the methods of step 206, then the expression value measured for gene A in individual #7 will represent MAX.
Next, the minimum value (MIN) for the data measured for the molecular marker in the subject trait subgroup is identified. In the example presented above, MIN is the expression Level of gene A in the subject having the lowest expression level for gene A in the set of ten subjects as determined by the low throughput measurement methods of step 206.
Next, the coefficient derived in the regression for the unique molecular marker (e.g., the coefficient for gene A) is multiplied by the difference between (MAX) and (MIN) in order to obtain the test value (TEST). In other words, for each coefficient i in a classifier, the following equation is computed:
TEST= coefficient; * [MAX - MINA.
As an example, consider the case in which a classifier is used to determine whether or not a subject has a particular cancer. In this case, one of the trait subgroups in training population 44 will represent patients that have this cancer. To evaluate a coefficient of a classifier in this case, the low value and high value for the measured data of the molecular marker i in the trait subgroup is obtained and the difference between these two values is multiplied against the coefficient value in order to obtain the value of TEST;. In some embodiments, a coefficient i is considered large and the classifier that includes the coefficient is discarded when the value is greater than 5, greater than 10, greater than 100 or greater than 1000.
In typical embodiments, a classifier will determine whether a subject falls into one of at least two different trait subgroups. In other words, the classifier will discriminate between at least two different trait subgroups. A test has been presented above for determining whether a coefficient in a regression derived classifier is too large. This test used one of the trait subgroups that the classifier discriminates. In some embodiments, the test is repeated for each of the trait subgroups that the classifier can discriminate. For example, in the case of a classifier that can discriminate between the cancerous trait subgroup and the non-cancerous trait subgroup, the test is independently run using data from each trait subgroup. That is, the test is first run using only data from the cancerous trait subgroup and then the test is run a second time using only data from the non-cancerous trait subgroup. If a coefficient is too large in any such independent test, the classifier is eliminated from further consideration. In some embodiments, the test is run against only one of the possible trait subgroups that the subject classifer can discriminate.
In another embodiment, some classifiers are eliminated on the basis of the score.
For example, where the scoring system used is receiver operating characteristic (ROC) curve score determined by an area under the ROC curve, in some embodiments, those classifiers with scores of less than 0.95, 0.9, 0.85, 0.8, 0.7, 0.65, 0.6, 0.55 0.5 or 0.45 or less can be eliminated. In other embodiments, where specificity is important to the use of the classifier, a sensitivity threshold can be set and classifiers ranked on the basis of the specificity. for example with a cutoff for specificity of less than 0.95, 0.9, 0.85, 0.8, 0.7, 0.65, 0.6, 0.55 0.5 or 0.45 or less can be eliminated. Similarly, the specificity threshold can be set and classifiers ranked on the basis of sensitivity for example with a cutoff for sensitivity of less than 0.95, 0.9, 0.85, 0.8, 0.7, 0.65, 0.6, 0.55 0.5 or 0.45 or less can be eliminated. Thus in some embodiments, only the top 10 ranking classifiers, the top 20 ranking classifiers, or the top 100 ranking classifiers are selected and the remaining classifiers eliminated.
Step 224.
After classifiers have been scored and ranked and some classifiers optionally eliminated; one or more classifiers can be combined to create a classifier group. For instance, in some embodiments, the top 10 ranking classifiers, the top 20 ranking classifiers, or the top 100 ranking classifiers are selected. In some embodiments, any of the top 1 to top 500 ranking classifiers is selected. In instances where more than one classifier is selected, in one embodiment, each classifier contributes one vote to the diagnosis of the test subject such that diagnosis of the test subject is determined as a result of a combination of classifiers. In other embodiments, multiple classifiers can be used and different weighting schema applied to each classifier. For example, weighting schema can include weighting on the basis of factors such as the original score the classifier, the logs odd ratio ("logit"), the size of the coefficients for each classifier, some combination thereof and the like.
Step 226.
Step 226 is optional. Optional step 226 is useful if training population 44 is comprised of more than two trait subgroups. In cases where there axe more than two trait subgroups, multiple binary classifiers (or groups of said classifiers) can be developed wherein each binary classifier (or group of said classifiers) is directed towards differentiating as between two traits. In one embodiment, each round of the funnel (e.g.
steps 202, 204, 206, 214, 216, 220 and 224) produces a set of binary classifiers. In another embodiment, multiple lists of binary candidate molecular markers are developed by performing step 202, and then binary classifiers (or groups of binary classifiers) are developed by proceeding with multiple rounds of the remainder of the funnel (e.g. steps 204, 206, 214, 216, 220 and step 224). Because each classifier represents only a binary test e.g. the absence or presence of a single trait, in step 226 a determination is made as to whether all classifiers have been developed for training population 44. If not (226-No), process control returns to step 210 and work is initiated to develop a classifier for a different trait represented by training population 44. Therefore, steps 210-224 can optionally be repeated until one or more classifiers or groups of classifiers have been selected for each of the trait subgroups represented by training population 44.

In some embodiments, each classifier or group of classifiers developed in accordance with embodiments of the present invention is stored in classifier database 70.
Fig. 5 illustrates an exemplary classifier database 70 in accordance with one embodiment of the present invention. Database 70 includes an entry 400 for each classifier 400. Each classifier 400 is optionally given a classifier name 402. Each classifier 400 is part of classifier. For example, a given classifier can consist of only a single classifier. In other embodiments, a given classifier can consist of one or a plurality of classifiers. Therefore, each classifier 402 includes an indicator 403 to indicate which classifier the classifier is in.
Further, each classifier 400 has an optional indicator 404 to indicate the trait that the classifier can discriminate. In some embodiments, optional indicator indicates the trait subgroups that the classifier can discriminate. In some embodiments such information can be inferred from the classifier identifier field 403 since each classifier represents the absence or presence of a particular trait (e.g., absence or presence of cancer). In addition to this header information, each classifier includes the identity 406 of one or more molecular markers and the -respective coefficients 408 for each of the molecular markers.
Once classifiers or classifier groups have been developed, the classifiers can be used to diagnose a patient that has presented as possibly having a disease that can be differentiated by the classifiers. Figure 3 is a flowchart of a method of applying the classifiers to a patient.
Step 328.
Step 328 can be performed after the previously described steps, or can be used in conjunction with a classifier or classifier groups derived using the methods disclosed herein. As such, steps 328-334 represent a completely independent method of the present invention and can be performed at any time once suitable classifiers have been developed using, for example, steps 301 through 326. Step 328 is used in conjunction with step 330 to diagnose a trait of interest of an individual not represented in either the training population or the scoring population (a "test individual"). Each classifier or classifier group identified previously can be used to determine whether a test individual has a trait of interest. In order to perform such tests, molecular marker data for each molecular marker of the classifier or classifier group of interest is required. To obtain such data, a sample of blood from the subject is obtained using any of the techniques described in Section 5.2.
The sample is used to measure molecular marker data for each molecular marker in the sample using any of the techniques described in Sections 5.3 or 5.4. Thus in step 328, once classifiers have been identified, molecular marker data for use with the classifier or classifier groups can be obtained using either high throughput or low throughput techniques.
Advantageously, the molecular marker data obtained in step 328 can be stored in patient database 68 for later use. In fact, in some embodiments, rather than obtaining molecular marker data from a patient sample in step 328, the data is obtained from a subject in patient database 68. In such embodiments, the molecular marker data was previously loaded into patient database 68.
Fig. 6 illustrates a patient database 68 in accordance with one embodiment of the present invention. There is a record 500 for each patient (subject) tracked by patient database 68. Each patient record 500 optionally includes a patient identifier 502 to uniquely identify the patient. In some embodiments such unique identifiers can be inferred from the patient record value 500. Each patient record 500 includes a molecular profile 504 comprising molecular marker data collected for a plurality of molecular marker products from a sample defined in Section 5.2 using any one of the techniques described in Sections 5.3 and 5.4. In typical embodiments, a molecular profile 504 includes a ph~rality of molecular marker identities 506 and the corresponding measured molecular marker data values 508 for such molecular markers. In addition to the molecular profile, each patient record 502 can include one or more traits 510. Such trait characterizations can be assigned by observation of the subject and/or by testing the patient's molecular profile using the classifiers constructed in accordance with the methods of the present invention. Section 5.10, below, provides more details on exemplary patient databases 68.
Step 330.
In step 330 the classifier created using some or more of the previous steps is used to diagnose a test individual. In some embodiments diagnosis can be performed using a classifier group from step 224. For example, in one embodiment, where there are numerous classifiers after step 222 that provide satisfactory scores (given the purpose for use), a test subject can be diagnosed by using the results of all or some of these classifiers in the form of a classifier group as described in step 224. In one embodiment, the term diagnosis means the results of a single classifier or group of classifiers resulting from the application of the funneling method described in steps 202-224. For example, the resulting classifier or group of classifiers will enable the ability to determine whether a test individual belongs to one of two possible trait subgroups. In another embodiment, by the term diagnosis is meant the results of multiple classifiers or multiple groups of classifiers (ie classifiers resulting from the application of more than one round of the funneling method described in steps 202 - 224). For example, the resulting classifiers or groups of classifiers used in series can allow a diagnosis as to whether an individual belongs to one of three or more possible trait subgroups (e.g. results of first classifier distinguish as to whether person has schizophrenia or does not have schizophrenia - If not schizophrenia apply a second classifier or group of classifiers to determine whether individual has bipolar disorder or does not have bipolar disorder etc.) The use of the classifiers to diagnose depends upon the trait subgroups used to develop the classifier. For example, if the classifier was developed to differentiate as between two trait subgroups, the classifier can be used to diagnose a test subject as being either of the first trait subgroup or the second trait subgroup. To diagnose a test subject, preferably a quantitative technique such as quantitative RT-PCR is utilized to obtain molecular marker data measured in step 328 is used.
To illustrate the use of a classifier to diagnose, consider the case in which the classifier comprises the classifier group:
ln[p/(1-p)]=0.34+0.24X1 +0.74X3 +0.03 In[p/(1-p)] =0.54- 0.4X2 +83X~ +0.01 That is, the exemplary classifier group consisting of two classifiers. To poll the classifier, the data (e.g. abundance level, activity level, etc.) for molecular markers Xl, X2 and X3 is measured using any of the techniques described in Section 5.3 or 5.4. Then, these data measurement values are placed into the classifier equations and the equations are processed and the output is used to predict outcome In one embodiment, classifier equation values that approach a value of one indicate that the sample has the trait associated with the classifier whereas classifier equation values that approach zero indicate that the sample does not have the trait associated with the classifier. In other embodiments, the equations are regressed so that the opposite relationship hold (e.g., equation values approach one indicate absence of an associated trait). In one embodiment, each equation is assigned a "+1" vote if the equation approaches one or a "-1"
vote if the equation approaches zero. Equation votes are summed. If the net summation is positive, then the subject is deemed to have the trait associated with the classifier.
If the net summation is negative, then the subject is deemed not to have the trait associated with the classifier. In some embodiments, step 330 is performed by model polling and reporting module 66.

Step 332.
One of the advantages of the present invention is that a single sample collected in accordance with Section 5.2 can be used to test the patient for one or more of a plurality of molecular markers which may be useful for one or more traits. Accordingly, in step 332, a determination is made as to whether the patient has been tested for each pair of traits fox which a determination is required. If additional determinations are required (332-No), process control is returned to step 330 and the measured molecular marker data from the patient is used to help determine the likelihood as to whether a subject has other traits represented by a trait subgroup in training population 44.
Step 334.
When the patient sample has been used to determine if the subject has any of a plurality of different traits as determined by one or more classifiers or classifier groups, a report is generated. In some embodiments, this report includes the results of each classification test. In other words, the report provides an indication as to whether it is likely the tested subject has any one of a plurality of different traits. In some embodiments, step 334 is performed by classifier polling and reporting module 66.
Section 5.5 provides a summary of some of the applications for classifiers constructed using the methods of the present invention.
5.2 SOURCE OF MOLECULAR MARKER DATA
The present invention provides methods fox identifying molecular markers by obtaining molecular marker data which represents the products of molecular markers found in a blood sample. Molecular markers are thus identified that correlate with, are associated with, or indicate a trait. The present invention also provides methods for detecting, diagnosing, monitoring, prognosing or predicting a trait or reoccurrence of a trait based upon data corresponding to the expression of molecular markers in a blood sample. As used herein, the terms "subject" and "patient" and "individual" are used interchangeably to refer to an animal (e.g., a mammal, a fish, an amphibian, a reptile, a bird, and an insect). In a specific embodiment, a subject is a mammal (e.g., a non-human mammal and a human). In another embodiment, a subject is a pet (e.g., a dog, a cat, a guinea pig, a monkey and a bird), a farm animal (e.g., a horse, a cow, a pig, a goat and a chicken) or a laboratory animal (e.g., a mouse and rat). In another embodiment, the subject is a primate (e.g., a chimpanzee and a human). In a preferred embodiment, the subject is a human.

5.2.1 ~SbLT~tCE~~OF~A BLOOD SAMPLE
A blood sample obtained from any subject may be used in accordance with the methods of the invention. Examples of subjects from which a blood sample can be obtained and utilized in accordance with the methods of the invention include, but are not limited to, asymptomatic subjects, subjects manifesting or exhibiting 1, 2, 3, 4 or more traits or symptoms of a trait, subjects clinically diagnosed as having a trait, subjects predisposed to a trait (e.g., subjects with a family history of a trait, subjects with a genetic predisposition to a trait, and subjects that lead a lifestyle that predisposes them to a trait or increases the likelihood of contracting a trait), subjects suspected of having a trait, subjects undergoing therapy for a trait, subjects non-responsive to a therapy, subjects responsive to a therapy, subjects with more than one trait (e.g., subjects with 2, 3, 4, 5 or more traits), subjects not undergoing therapy for a trait, subjects determined by a medical practitioner (e.g., a physician) to be healthy or disease-free, subjects that are in remission, subjects cured of trait, and subjects that have not been diagnosed with a condition. In specific embodiment, the condition is a disease. In another specific embodiment, a condition is any state that is codified in the Inte~r~atiohal Classification of Diseases, 9th Revision, Department of Health and Human Services (ICD-9 codes) and/or SNOMED Clinical Ternis (SNCMED CT~) which is hereby incorporated by reference, or equivalent treatise.
Non-limiting examples of disease include, but are not limited to, blood disorder, blood lipid disease, autoimmune disease, arthritis (e.g., osteoarthritis, rheumatoid arthritis, juvenile rheumatoid arthritis and the like), bone or joint disorder, lupus, an allergy, a cardiovascular disorder (e.g., heart failure, congenital heart disease, rheumatic fever, valvular heart disease, corpulmonale, cardiomyopathy, myocarditis, pericardial disease, vascular diseases such as atherosclerosis, acute myocardial infarction, ischemic heart disease and the like), obesity, respiratory disease (e.g., asthma, pneumonitis, pneumonia, pulmonary infections, lung disease, bronchiectasis, tuberculosis, cystic fibrosis, interstitial lung disease, chronic bronchitis emphysema, pulmonary hypertension, pulmonary thromboembolism, acute respiratory distress syndrome and the like), hyperlipidemias, endocrine disorder, immune disorder, infectious disease, muscle wasting and whole body wasting disorder, neurological disorder (e.g., migraines, seizures, epilepsy, cerebrovascular disease, Parkinson's, ataxic disorders, motor neuron diseases, cranial nerve disorders, spinal cord disorders, meningitis and the like), neurodegenerative disease (e.g., alzheimers, dementia and the like), neuropsychiatric disease (e.g., schizophrenia, anxiety and the like), mood disorders (e.g., bipolar disorder; manic depression and the like), skin disorder, kidney disease, scleroderma, stroke, hereditary hemorrhage telangiectasia, diabetes, disorders associated with diabetes (e.g., PVD), hypertension, Gaucher's disease, cystic fibrosis, sickle cell anemia, liver disease, stomach disease, pancreatic disease, eye disease, ear disease, nose disease, throat disease, diseases affecting the reproductive organs, S gastrointestinal diseases (including diseases of the colon, diseases of the spleen, appendix, gall bladder, and others) and the like. For further discussion of human diseases, see Mendelian Inheritance in Man: A Catalog of Human Genes and Genetic Disorders by Victor A. McKusick (12th Edition, 3 volume set, June 1998, Johns Hopkins University Press, ISBN: 0801857422) and Harrison's Principles of Internal Medicine by Braunwald, Fauci, Kasper, Hauser, Longo, & Jameson (lSth Edition 2001), the entirety of each of which is incorporated herein. Additional examples of disease are disclosed in Section 5.8, below.
In one embodiment of the invention, the disease is an immune disorder, such as those associated with overexpression of a gene or expression of a mutant gene 1S (e.g., autoimmLme diseases, such as diabetes mellitus, arthritis (including rheumatoid arthritis, juvenile rheumatoid arthritis, osteoarthritis, and psoriatic arthritis), multiple sclerosis, encephalomyelitis, myasthenia gravis, systemic lupus erythematosis, automimmune thyroiditis, dermatitis (including atopic dermatitis and eczematous dermatitis), psoriasis, Sjogren's Syndrome, Crohn's disease, aphthous ulcer, iritis, conjunctivitis, keratoconjunctivitis, ulcerative colitis, asthma, allergic asthma, cutaneous lupus ezythematosus, scleroderma, vaginitis, proctitis, drug eruptions, leprosy reversal reactions, erythema nodosum leprosum, autoimmune uveitis, allergic encephalomyelitis, acute necrotizing hemorrhagic encephalopathy, idiopathic bilateral progressive sensorineural hearing, loss, aplastic anemia, pure red cell anemia, idiopathic 2S thrombocytopenia, polychondritis, Wegener's granulomatosis, chronic active hepatitis, Stevens-Johnson syndrome, idiopathic sprue, lichen planus, Graves' disease, sarcoidosis, primary biliary cirrhosis, uveitis posterior, and interstitial lung fibrosis), graft-versus-host disease, cases of transplantation, and allergy.
In another embodiment, a disease of the invention is a cellular proliferative and/or differentiative disorder that includes, but is not limited to, cancer, e.g., carcinoma, sarcoma or other metastatic disorders and the like. As used herein, the term "cancer"
refers to cells having the capacity for autonomous growth, i.e., an abnormal state of condition characterized by rapidly proliferating cell growth. "Cancer" is meant to include all types of cancerous growths, pre-cancerous growths or lesions, oncogenic processes, metastatic tissues or malignantly transformed cells, tissues, or organs, irrespective of histopathologic type or stage of invasiveness. Examples of cancers include, but are not limited to, solid tumors, tissue specific tumors, benign cancer, metastatic cancers, early stage cancer, late stage cancer and leukemias, including: apudoma, choristoma, branchioma, malignant carcinoid syndrome, carcinoid heart disease, carcinoma (e.g., Walker, basal cell, basosquamous, Brown-Pearce, ductal, Ehrlich tumour, in situ, Krebs 2, Merkel cell, mucinous, non-small cell lung, oat cell, papillary, scirrhous, bronchiolar, bronchogenic, squamous cell, and transitional cell), histiocytic disorders, leukaemia (e.g., B cell, mixed cell, null cell, T cell, T-cell chronic, HTLV-II-associated, lymphocytic acute, lymphocytic chronic, mast cell, and myeloid), histiocytosis malignant, Hodgkin disease, immunoproliferative small, non-Hodgkin lymphoma, plasmacytoma, reticuloendotheliosis, melanoma, chondroblastoma, chondroma, chondrosarcoma, fibroma, fibrosaxcoma, giant cell tumors, histiocytoma, lipoma, liposarcoma, mesothelioma, myxoma, myxosarcoma, osteoma, osteosarcoma, Ewing sarcoma, synovioma, adenofibroma, adenolymphoma, carcinosaxcoma, chordoma, craniopharyngioma, dysgerminoma, hamartoma, mesenchymoma, mesonephroma, myosarcoma, ameloblastoma, cementoma, odontoma, teratoma, thymoma, trophoblastic tumour, adeno-carcinoma, adenoma, cholangioma, cholesteatoma, cylindroma, cystadenocarcinoma, cystadenoma, granulosa cell tumour, gynandroblastoma, hepatoma, hidradenoma, islet cell tumour, Leydig cell tumour, papilloma, Sertoli cell tumour, theca cell tumour, leiomyoma, leiomyosarcoma, myoblastoma, mymoma, myosaxcoma, rhabdomyoma, rhabdomyosarcoma, ependymoma, ganglioneuroma, glioma, medulloblastoma, meningioma, neurilemmoma, neuroblastoma, neuroepithelioma, neurofibroma, neuroma, paraganglioma, paraganglioma nonchromaffin, angiokeratoma, angiolymphoid hyperplasia with eosinophilia, angioma sclerosing, angiomatosis, glomangioma, hemangioendothelioma, hemangioma, hemangiopericytoma, hemangiosarcoma, lymphangioma, lymphangiomyoma, lymphangiosarcoma, pinealoma, carcinosarcoma, chondrosaxcoma, cystosarcoma, phyllodes, fibrosarcoma, hemangiosarcoma, leimyosarcoma, leukosarcoma, liposarcoma, lymphangiosarcoma, myosarcoma, myxosarcoma, ovarian carcinoma, rhabdomyosarcoma, sarcoma (e.g., Ewing, experimental, I~aposi, and mast cell), neoplasms (e.g., bone, breast, digestive system, colorectal, liver, pancreatic, pituitary, testicular, orbital, head and neck, central nervous system, acoustic, pelvic respiratory tract, and urogenital), neurofibromatosis, and cervical dysplasia, and other conditions in which cells have become immortalized or transformed.

5.~.2~ ~ 1VIET~IODS FOR COLLECTING BLOOD
In one aspect, a sample of blood is obtained from a subject according to methods well known in the art. The present invention can use whole blood, but can also use blood in which the serum or plasma has been removed and the RNA or mRNA isolated from the remaining sample in accordance with methods known in the art (for example, using preferably gentle centrifugation at 300-~OOxg for five to ten minutes).
In some embodiments a drop of blood is collected from a simple pin prick made in the skin of a subject. In such embodiments, this drop of blood collected from a pin prick is all that is needed. A drop of blood can include volumes of anywhere from 10u1 through to 100u1. Blood may be drawn from a subject from any part of the body (e.g., a finger, a hand, a wrist, an arm, a leg, a foot, an ankle, a stomach, and a neck) using techniques known to one of skill in the art, in particular methods of phlebotomy known in the art. In a specific embodiment, venous blood is obtained from a subject and utilized in accordance with the methods of the invention. In another embodiment, arterial blood is obtained and utilized in accordance with the methods of the invention. The composition of venous blood varies according to the metabolic needs of the area of the body it is servicing. In contrast, the composition of arterial blood is consistent throughout the body. For routine blood tests, venous blood is generally used.
Venous blood can be obtained from any source including the basilic vein, cephalic vein, or median vein. Arterial blood can be obtained from the radial artery, brachial artery or femoral artery. A vacuum tube, a syringe or a butterfly may be used to draw the blood.
Typically, the puncture site is cleaned, a tourniquet is applied approximately 3-4 inches above the puncture site, a needle is inserted at about a 15-45 degree angle, and if using a vacuum tube, the tube is pushed into the needle holder as soon as the needle penetrates the wall of the vein. When finished collecting the blood, the needle is removed and pressure is maintained on the puncture site. Usually, heparin or another type of anticoagulant is in the tube or vial that the blood is collected in so that the blood does not clot.
When collecting arterial blood, anesthetics can be administered prior to collection.
The amount of blood collected will vary depending upon the site of collection, the amount required for a method of the invention, and the comfort of the subject.
However, an advantage of one embodiment of the present invention is that the amount of blood required to implement the methods of the present invention can be so small that more invasive procedures are not required to obtain the sample. For example, in some embodiments, all that is required is a drop of blood. This drop of blood can be obtained, for example, from a simple pinprick. In some embodiments, any amount of blood is collected that is sufficient to measure molecular marker data. As such, in some embodiments, the amount of blood that is collected is 1 ~,1 or less, 0.5 ~1 or less, 0.1 ~1 or less, or 0.01 ~1 or less. However, the present invention is not limited to such embodiments. In some embodiments more blood is available and in some embodiments, more blood can be used to effect the methods of the present invention. As such, a broad range of blood volumes is contemplated and can be used to obtain the molecular marker data measurement data used in the present invention. In various specific embodiments, 0.001 ml, 0.005 ml, 0.01 ml, 0.05 ml, 0.1 ml, 0.15 ml, 0.2 ml, 0.25 ml, 0.5 ml, 0.75 ml, 1 ml, 1.5 ml, 2 ml, 3 ml, 4 ml, 5 ml, 10 ml, 15 ml or more of blood is collected from a subject. In other specific embodiments, 0.001 ml to 15m1, 0.01 ml to 10 ml, 0.1 ml to 10 ml, 0.1 ml to 5 ml, 1 to 5 ml of blood is collected from a subject.
In some embodiments of the present invention, blood is stored within a K3/EDTA
tube. In another embodiment, one can utilize tubes for storing blood which contain stabilizing agents such as disclosed in United States patent No. 6,617,170. In another embodiment, the PAXgeneTM blood RNA system provided by PreAnalytiX, a Qiagen/BD
company, can be used to collect blood. In yet another embodiment, the TempusTM
blood RNA collection tubes, offered by Applied Biosystems, can be used. TempusTM
collection tubes provide a closed evacuated plastic tube containing RNA stabilizing reagent for whole blood collection.
The collected blood is optionally but preferably stored at refrigerated temperatures, such as 4 °C, prior to molecular marker data measurement. In some embodiments, a portion of the blood sample is used for molecular measurement at a first instance of time whereas one or more remaining portions of the blood sample is stored for a period of time for later use. This period of time can be an hour or more, a day or more, a week or more, a month or more, a year or more, or indefinitely. For long term storage, storage methods well known in the art, such as storage at cryo temperatures (e.g. below -60 °C) can be used. In some embodiments, in addition to storage of the blood (or instead of storage of t he blood), isolated molecular markers (e.g., nucleic acid, protein, carbohydrates, lipids, metabolites, etc.) are stored for a period of time for later use. Storage of such molecular markers can be for an hour or more, a day or more, a week or more, a month or more, a year or more, or indefinitely.

5.2.3 1VIETHOD~-OF ISOLATING BLOOD CELLS
In some embodiments of the present invention, whole blood is used directly to isolate and analyze the products of one or more molecular markers so as to obtain molecular marker data. In other embodiments of the invention fractionated blood can be used. By fractionated blood is meant blood in which the blood cells are separated prior to isolation of the molecular markers using techniques known in the art. For example, fractionated blood includes blood wherein the blood cells are fractionated using Ficoll-Hypaque (Pharmacia) gradient centrifugation. Such centrifugation separates erythrocytes (red blood cells) from various types of nucleated cells and from plasma. As such, in some embodiments of the present invention, a blood sample of the invention is fractionated blood. In one embodiment, peripheral blood leukocytes are utilized ("PBLs"). PBLs are separated from the remainder of the blood using a Ficoll~
gradient.
By way of example but not limitation, macrophages can be obtained as follows.
Mononuclear cells are isolated from peripheral blood of a subject, by syringe removal of blood followed by Ficoll-Hypaque gradient centrifugation. Tissue culture dishes are pre-coated with the subject's own serum or with AB+ human serum and incubated at 37°C
for one hour. Non-adherent cells are removed by pipetting. Cold (4°C) 1mM EDTA in phosphate-buffered saline is added to the adherent cells left in the dish and the dishes are left at room temperature for fifteen minutes. The cells are harvested, washed with RPMI
buffer and suspended in RPMI buffer. Increased numbers of macrophages can be obtained by incubating at 37°C with macrophage-colony stimulating factor (M-CSF). Antibodies against macrophage specific surface markers, such as Mac-l, can be labeled by conjugation of an affinity compound to such molecules to facilitate detection and separation of macrophages. Affinity compounds that can be used include but are not limited to biotin, photobiotin, fluorescein isothiocyante (FITC), or phycoerythrin (PE), or other compounds known in the art. Cells retaining labeled antibodies are then separated from cells that do not bind such antibodies by techniques known in the art such as, but not limited to, various cell sorting methods, affinity chromatography, and panning.
Blood cells can be fractionated using a fluorescence activated cell sorter (FACS).
Fluorescence activated cell sorting (FACE) is a known method for separating particles, ncluding cells, based on the fluorescent properties of the particles. See, for example, I~amarch, 1987, Methods Enzymol 151:150 165. Laser excitation of fluorescent moieties in the individual particles results in a small electrical charge allowing electromagnetic separation of positive and negative particles from a mixture.
An antibody or ligand used to detect a blood cell antigenic determinant present on the cell surface of particular blood cells is labeled with a fluorochrome, such as FITC or phycoerythrin. The cells are incubated with the fluorescently labeled antibody or ligand for a time period sufficient to allow the labeled antibody or ligand to bind to cells. The cells are processed through the cell sorter, allowing separation of the cells of interest from other cells. FACS
sorted particles can be directly deposited into individual wells of microtiter plates to facilitate separation.
Magnetic beads can be also used to separate blood cells in some embodiments of the present invention. For example, blood cells can be sorted using a magnetic activated cell sorting (MACS) technique, a method for separating particles based on their ability to bind magnetic beads (0.5 100 ~m diameter). A variety of useful modifications can be performed on the magnetic microspheres, including covalent addition of an antibody which specifically recognizes a cell solid phase surface molecule or hapten. A
magnetic fzeld is then applied, to physically manipulate the selected beads. In a specific embodiment, antibodies to a blood cell surface marker are coupled to magnetic beads. The beads are then mixed with the blood cell culture to allow binding. Cells are then passed through a magnetic field to separate out cells having the blood cell surface markers of interest. These cells can then be isolated.
In some embodiments, the surface of a culture dish may be coated with antibodies, and used to separate blood cells by a method called panning. Separate dishes can be coated with antibody specific to particular blood cells. Cells can be added first to a dish coated with blood cell specific antibodies of interest. After thorough rinsing, the cells left bound to the dish will be cells that express the blood cell markers of interest. Examples of cell surface antigenic determinants or markers include, but are not limited to, CD2 for T
lymphocytes and natural killer cells, CD3 fox T lymphocytes, CD1 la for leukocytes, CD28 for T lymphocytes, CD19 for B lymphocytes, CD20 for B lymphocytes, CD21 for B
lymphocytes, CD22 for B lymphocytes, CD23 for B lymphocytes, CD29 for leukocytes, CD14 for monocytes, CD41 for platelets, CD61 for platelets, CD66 for granulocytes, CD67 for granulocytes and CD68 for monocytes and macrophages.
Whole blood can also be fractioned into cells types such as leukocytes, platelets, erythrocytes, etc. Leukocytes can be further separated into granulocytes and agranulocytes using standaxd techniques. Granulocytes can be separated into cell types such as neutrophils, eosinophils, and basophils using standard techniques.
Agranulocytes can be separated into lymphocytes (e.g., T lymphocytes and B lymphocytes) and monocytes using standard techniques. T lymphocytes can be separated from B lymphocytes and helper T
cells separated from cytotoxic T cells using standard techniques. Separated blood cells (e.g., leukocytes) can be frozen by standard techniques prior to use in the present methods.
5.2.4 BLOOD SAMPLES USED IN METHODS OF THE INVENTION
In accordance with the methods of the invention, the term "blood sample"can include any of the samples discussed in section 5.2.3. In some embodiments, this includes whole blood, fractionated blood, a sample of subsets of fractionated blood, and a sample of specific types of blood cells. In a specific embodiment, the whole blood sample can have the plasma or serum removed by centrifugation, using preferably gentle centrifugation at 300-800xg for five to ten minutes.
In another embodiment, a blood sample of the invention is a sample of peripheral blood leukocytes (PBLs). In another embodiment, a blood sample of the invention is a sample of granulocytes. In another embodiment, a blood sample of the invention is a sample of neutrophils, eosinophils, basophils or any combination thereof. In another embodiment, a blood sample of the invention is a sample of agranulocytes. In another embodiment, a blood sample of the invention is a sample of lymphocytes, monocytes or a combination thereof. In yet another embodiment, a blood sample of the invention is a sample of T lymphocytes, B lymphocytes or a combination thereof. See, e.g., Section 5.4.3 supra for methods of isolating blood cells.
A blood sample that is useful according to the invention is in an amount that is sufficient for the detection of one or more molecular markers according to the invention.
In a specific embodiment, a blood sample useful according to the invention is in an amount ranging from 1 p.1 to 100 ml, preferably 10 ~1 to 50 ml, more preferably 10 p,1 to 25 ml and most preferably 10 ~,l to 1 ml.
In one embodiment whole blood, or serum free whole blood is taken and the redblood cells are lysed with lysing buffer. In a specific embodiment, the Lysis Buffer (1L) consists of 0.6g EDTA 1.0g KHC02, and 8.2g NH4Cl adjusted to pH 7.4 (using NaOH). Once mixed with lysing buffer, the sample is centrifuged and the cell pellet retained and RNA or mRNA extracted in accordance with methods known in the art ("lysed whole blood") (see, for example, Sambrook, Fritsch & Maniatis, "Molecular Cloning: A Laboratory Manual (1982); "DNA Cloning: A Practical Approach,"
Volumes I
and II (D.N. Glover ed. 1985). The use of whole blood or lysed whole blood is preferred since it avoids the costly and time-consuming need to separate out the cell types within the blood (Liew et al. U.S. Patent Application No.US 2004/0014059). In another embodiment, whole blood is stored and stabilized using PAXgene~ tubes and RNA
can be isolated using the PAXgene~ RNA Isolation system. In yet another embodiment, RNA is isolated from whole blood which has been isolated using PAXgene~ and additional globin reduction protocols followed.
5.3 METHODS FOR MEASURING MOLECULAR MARKER DATA
The techniques described in this section are particularly useful for obtaining molecular marker data for step 202 wherein the data is reflective of the products of the molecular marker -e.g. measurement of the abundance of RNA or protein products in blood corresponding to all of the molecular markers of the genome or "a portion thereof'.
In particular, the techniques useful for step 202 are those techniques which allow the ability to comprehensively screen for candidate molecular markers quickly and effectively.
These techniques preferably provide molecular marker data for a large number of molecular markers concurrently, thereby allowing greater ability to screen molecular markers corresponding to the entire genome, or a portion thereof in a short period of time.
In addition to the techniques described in this Section 5.3, any technique known to one of skill in the art to measure the abundance of RNA or protein corresponding to the entire genome or "a portion thereof' can be used to measure such data. In one embodiment,"a portion thereof" is data corresponding to the amount of RNA or protein expressed from more than 1,000, more than 2,000, more than 5,000, more than 10,000, more than 20,000, more than 30,000 molecular markers. In another embodiment, the "a portion thereof refers to at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, or at leat 95% of the genome. See, e.g., Sambrook, Fritsch & Maniatis, 1982, Molecular Cloning: A Laboratory Manual;
DNA
Cloning: A Practical Approach, volumes I and II (I~.N. Glover ed. 1985);
Oligonucleotide Synthesis Gait ed., 1984; Nucleic Acid Hybt~idization, Hames 8i Higgins eds., 1985;
Transcription and Translation, Hames & Higgins eds., 1984, Animal Cell Culture, Freshney ed., 1986, Immobilized Cells And Enzymes, IRL Press, 1986, Perbal, 1984, A
Practical Guide To Moleculay~ Cloning, each of which is hereby incorporated by reference in its entirety. In one embodiment more than technique can be used to measure data for each molecular marker to perform step 202.
5.3.1 RNA MEASUREMENT TECHNIQUES
Any technique known to one of skill in the art may be used to measure the level of expression of a molecular marker by measuring the amount of the product of the molecular marker. In one embodiment the RNA in blood corresponding to the molecular marker is measured. By "corresponding to a molecular marker" is meant RNA transcribed from a molecular marker (or proteins translated from RNA which is transcribed from a molecular marker when referring to protein products of a molecular marker). RNA or protein which corresponds to a molecular marker are also considered the product of the molecular marker. In a specific embodiment, the level of an RNA product is measured using a echnique which permits generation of data for a large number of molecular markers.
However measured, the result is either the absolute or relative amounts of abundance of nucleic acids corresponding to the molecular markers, including but not limited to values representing abundances or abundance ratios.
5.3.1.1 MICROARRAYS
n one embodiment, nucleic acid arrays are employed for analyzing the level of RNA product of each molecular marker of the genome in a blood sample. In a specific embodiment, molecular marker data is obtained by hybridizing detectably labeled polynucleotides representing the nucleic acid sequences in mRNA transcripts present in a cell (e.g., fluorescently labeled cDNA synthesized from total cell mRNA) to a.
microarray.
In some embodiments expressed transcripts that may or may not represent genes expressed in the blood sample are analyzed.
In some embodiments, a microarray is an array of positionally-addressable binding (e.g., hybridization) sites on a support for representing many of the nucleic acid sequences in the genome of a cell or organism. Is some embodiments, a microarray represents most or almost all of the genes in a species. In some embodiments, each microarray binding site consists of polynucleotide probes bound to a predetermined region on the support.
Microarrays are described in Draghici, Data Analysis Tools Fog DNA
Micr~oa~rays, 2003, Chapman & Hall, CRC Press, New York, pp. 15-16, which is hereby incorporated by reference in its entirety.
Microarrays can be made in a number of ways. See, for example, Draghici, Data Analysis Tools For DNA Mic~oa~rays, 2003, Chapman & Hall, CRC Press, New York, pp.
16-22, which is hereby incorporated by reference in its entirety. Preferably microarrays are reproducible, allowing multiple copies of a given array to be produced and results from the microarrays compared with each other. Preferably, the microarrays are made from materials that are stable under binding (e.g., nucleic acid hybridization) conditions.
Microarrays are preferably small, e.g., between 1 cm2 and 25 cm2, preferably 1 to 3 cm2.
However, both larger and smaller arrays are also contemplated and may be preferable, e.g., for simultaneously evaluating a very large number or very small number of different probes.
In some embodiments, a given binding site or unique set of binding sites in the microarray will specifically bind (e.g., hybridize) to a nucleotide sequence in a single gene from a cell or organism (e.g., to exon of a specific mRNA or a specific cDNA
derived therefrom).
Microarrays used in the present invention can include one or more test probes.
In some embodiments each such test probe comprises a polynucleotide sequence that is complementary to a subsequence of RNA or DNA to be detected. Each probe typically has a different nucleic acid sequence, and the position of each probe on the solid surface of the array is usually known or can be determined. Microarrays useful in accordance with the invention can include oligonucleotide microarrays, cDNA based arrays, SNP
arrays, spliced variant arrays and any other array able to provide a quantitative or semi quantitative data of the invention. Some types of microarrays are addressable arrays.
More specifically, some microarrays are positionally addressable arrays. In some embodiments, each probe of the array is located at a known, predetermined position on the solid support so that the identity (e.g., the sequence) of each probe can be determined from its position on the array (e.g., on the support or surface). In some embodiments, the arrays are ordered arrays.
In some embodiments, the density of probes on a microarray or a set of microarrays is 100 different (e.g., non-identical) probes per 1 cm2 or higher. More preferably, a microarray used in the methods of the invention will have at least 550 probes per 1 cmz, at least 1,000 probes per 1 cm2, at least 1,500 probes per 1 cm2 or at least 2,000 probes per 1 cm2. In a particularly preferred embodiment, the microarray is a high density array, preferably having a density of at least 2,500 different probes per 1 cm2. The microarrays used in the invention therefore preferably contain at least 2,500, at least 5,000, at least 10,000, at least 15,000, at least 20,000, at least 25,000, at least 50,000 or at least 55,000 different (i.e., non-identical) probes.
In one embodiment, the microarray is an array (e.g., a matrix) in which each position represents a discrete binding site for a nucleotide sequence of a transcript encoded by a gene (e.g., for an exon of an mRNA or a cDNA derived therefrom). The collection of binding sites on a microarray contains sets of binding sites for a plurality of genes and thus can be used, in some embodiments to measure molecular marker data. For example, in various embodiments, the microarrays of the invention can comprise binding sites for products encoded by fewer than 50% of the genes in the genome of an organism.
Alternatively, the microarrays of the invention can have binding sites for the products encoded by at least 50%, at least 75%, at least 85%, at least 90%, at least 95%, at least 99% or 100% of the molecular markersof an organism. In other embodiments, the microarrays of the invention can having binding sites for products encoded by fewer than 50%, by at least 50%, by at least 75%, by at least 85%, by at least 90%, by at least 95%, by at least 99% or by 100% of the genes expressed by a cell of an organism. The binding site can be a DNA or DNA analog to which a particular RNA can specifically hybridize. The DNA or DNA analog can be, e.g., a synthetic oligomer or a gene fragment, e.g.
corresponding to an exon.
In some embodiments of the present invention, an expressed transcript is represented in the nucleic acid arrays. In such embodiments, a set of binding sites can include probes with different polynucleotides that are complementary to different sequence segments of the expressed transcript. Exemplary polynucleotides that fall within this class can be of length of 15 to 200 bases, 20 to 100 bases, 40-60 bases or some other range of bases. Each probe sequence can also comprise linker sequences in addition to the sequence that is complementary to its target sequence. As used herein, a linker sequence is a sequence between the sequence that is complementary to its target sequence and the surface of support. For example, in some embodiments, the nucleic acid arrays of the invention comprise one probe specific to each target gene or exon. However, if desired, the nucleic acid arrays can contain at least 2, 5, 10, 100, or 1000 or more probes specific to some expressed transcript. For example, the array may contain probes tiled across the sequence of the longest mRNA isoform of a gene at single base steps.
In specific embodiments of the invention, when an exon has alternative spliced variants, a set of polynucleotide probes of successive overlapping sequences, i.e., tiled sequences, across the genomic region containing the longest variant of an exon can be included in the exon nucleic acid arrays. The set of polynucleotide probes can comprise successive overlapping sequences at steps of a predetermined base intervals, e.g. at steps of l, 5, or 10 base intervals, span, or are tiled across the mRNA containing the longest variant. Such sets of probes therefore can be used to scan the genomic region containing all variants of an exon to determine the expressed variant or variants of the exon to determine the expressed variant or variants of the exon. Alternatively or additionally, a set of polynucleotide probes comprising exon specific probes and/or variant junction probes can be included in the exon profiling array. As used herein, a variant junction probe refers to a probe specific to the junction region of the particular axon variant and the neighboring axon. In some cases, the probe set contains variant junction probes specifically hybridizable to each of all different splice junction sequences of the axon.
In other cases, the probe set contains axon specific probes specifically hybridizable to the common sequences in all different variants of the axon, andlor variant junction probes specifically hybridizable to the different splice junction sequences of the axon.
In some cases, an axon is represented in the axon nucleic acid arrays by a probe comprising a polynucleotide that is complementary to the full length axon. In such instances, an axon is represented by a single binding site on the profiling arrays. In some preferred cases, an axon is represented by one or more binding sites on the profiling arrays, each of the binding sites comprising a probe with a polynucleotide sequence that is complementary to an RNA fragment that is a substantial portion of the target axon. The lengths of such probes are normally between 15-600 bases, preferably between bases, more preferably between 30-100 bases, and most preferably between 40-80 bases.
The average length of an axon is 200 bases (see, e.g., Lewin, Genes V, Oxford University Press, Oxford, 1994). A probe of length of 40-80 allows more specific binding of the axon than a probe of shorter length, thereby increasing the specificity of the probe to the target axon. For certain genes, one or more targeted axons may have sequence lengths less than 40-80 bases. In such cases, if probes with sequences longer than the target axons are to be used, it may be desirable to design probes comprising sequences that include the entire target axon flanked by sequences from the adjacent constitutively splice axon or axons such that the probe sequences are complementary to the corresponding sequence segments in the mRNAs. Using flanking sequence from adjacent constitutively spliced axon or axons rather than the genomic flanking sequences, i.e., intron sequences, permits comparable hybridization stringency with other probes of the same length.
Preferably the flanking sequence used are from the adjacent constitutively spliced axon or axons that are not involved in any alternative pathways. More preferably the flanking sequences used do not comprise a significant portion of the sequence of the adjacent axon or axons so that cross-hybridization can be minimized. In some embodiments, when a target axon that is shorter than the desired probe length is involved in alternative splicing, probes comprising flanking sequences in different alternatively spliced mRNAs are designed so that expression level of the axon expressed in different alternatively spliced mRNAs can be measured.

In some instances, when alternative splicing pathways and/or exon duplication in separate genes are to be distinguished, the DNA array or set of arrays can also comprise probes that are complementary to sequences spanning the junction regions of two adjacent exons. Preferably, such probes comprise sequences from the two exons which are not substantially overlapped with probes for each individual exons so that cross hybridization can be minimized. Probes that comprise sequences from more than one exons are useful in distinguishing alternative splicing pathways and/or expression of duplicated exons in separate genes if the exons occurs in one or more alternative spliced mRNAs and/or one or more separated genes that contain the duplicated exons but not in other alternatively spliced mRNAs and/or other genes that contain the duplicated exons.
Alternatively, for duplicate exons in separate genes, if the exons from different genes show substantial difference in sequence homology, it is preferable to include probes that are different so that the exons from different genes can be distinguished.
It will be apparent to one skilled in the art that any of the probe schemes, supra, can be combined on the same nucleic acid array and/or on different arrays within the same set of nucleic acid arrays so that a more accurate determination of the expression profile for a plurality of molecular marker products can be accomplished. It will also be apparent to one skilled in the art that the different probe schemes can also be used for different levels of accuracies in profiling. For example, a nucleic acid array or array set comprising a small set of probes for each expressed transcript or each region thereof may be used to identify molecular markers under certain specific conditions. An array or array set comprising larger sets of probes for the exons that are of interest is then used to more accurately determine the specific molecular marker products under such specific conditions. Other DNA array strategies that allow more advantageous use of different probe schemes are also encompassed.
In some embodiments, the microarrays used in the invention can include binding sites (e.g., probes) for sets of exons for one or more genes relevant to the condition of interest. The number of genes in a genome can be estimated from the number of mRNAs expressed by the cell or organism, or by extrapolation of a well characterized portion of the genome. When the genome of the organism of interest has been sequenced, the number of ORFs can be determined and mRNA coding regions identif ed by analysis of the DNA sequence. For example, the human genome is now known. Genome sequences for other organisms are also completed or nearly completed. Thus, in some embodiments of the invention, an array set comprising the total probes fox all known or predicted exons in the genome of an organism is provided. As a non-limiting example, the present invention provides an array set comprising one or two probes for each known or predicted exon of a molecular marker of the human genome.
It will be appreciated that when cDNA complementary to the RNA of a cell is made and hybridized to a microarray under suitable hybridization conditions, the level of hybridization to the site in the array corresponding to an exon of any particular molecular marker will reflect the prevalence in the cell of mRNA or mRNAs containing the exon transcribed from that molecular marker. For example, when detectably labeled (e.g., with a fluorophore) cDNA complementary to the total cellular mRNA is hybridized to a microarray, the site on the array corresponding to an exon of a gene (i.e., capable of specifically binding the product or products of the gene expressing) that is not transcribed or is removed during RNA splicing in the cell will have little or no signal (e.g., fluorescent signal), and an exon of a gene for which the encoded mRNA expressing the exon is prevalent will have a relatively strong signal. The relative abundance of different mRNAs produced from the same gene by alternative splicing is then determined by the signal strength pattern across the whole set of exons monitored for the gene.
The use of a two-color fluorescence labeling and detection scheme to define alterations in gene expression has been described in connection with detection of mRNAs, e.g., in Shena et al., 1995, Science 270:467-470, which is incorporated by reference in its entirety for all purposes. In some embodiments, such schemes axe used to measure molecular marker data. Such schemes are equally applicable to labeling and detection of expressed transcripts. An advantage of using cDNA labeled with two different fluorophores is that a direct and internally controlled comparison of the mRNA
or exon expression levels corresponding to each arrayed gene in two blood samples can be made, and variations due to minor differences in experimental conditions (e.g., hybridization conditions) will not affect subsequent analyses. However, it will be recognized that it is also possible to use cDNA from a single cell, and compare, for example, the absolute amount of a particular exon in, e.g., a drug-treated or pathway-perturbed cell and an untreated cell. Furthermore, labeling with more than two colors is also contemplated in the present invention. In some embodiments of the invention, at least 5, 10, 20, or 100 dyes of different colors can be used for labeling. Such labeling permits simultaneous hybridizing of the distinguishably labeled cDNA populations to the same array, and thus measuring, and optionally comparing the expression levels of, mRNA molecules derived from more than two samples. Dyes that can be used include, but are not limited to, fluorescein and its derivatives, rhodamine and its derivatives, texas red, 5'carboxy-fluorescein ("FMA"), 2',7'-dimethoxy-4',5'-dichloro-6-carboxy-fluorescein ("JOE"), N,N,N',N'-tetramethyl-6-carboxy-rhodamine ("TAMRA"), 6'carboxy-X-rhodamine ("ROX"), HEX, TET, IRD40, and IRD41, cyamine dyes, including but are not limited to Cy3, Cy3.5 and CyS; BODIPY dyes including but are not limited to BODIPY-FL, BODIPY-TR, BODIPY-TMR, BODIPY-630/650, and BODIPY-650/670; and ALEXA
dyes, including but are not limited to ALEXA-488, ALEXA-532, ALEXA-546, ALEXA-568, and ALEXA-594; as well as other fluorescent dyes which will be known to those who are skilled in the art.
In one embodiment, hybridization levels at different hybridization times are measured separately on different, identical microarrays. For each such measurement, at hybridization time when hybridization level is measured, the microarray is washed briefly, preferably in room temperature in an aqueous solution of high to moderate salt concentration (e.g., 0.5 to 3 M salt concentration) under conditions which retain all bound or hybridized polynucleotides while removing all unbound polynucleotides. The detectable label on the remaining, hybridized polynucleotide molecules on each probe is then measured by a method which is appropriate to the particular labeling method used.
The resulted hybridization levels are then combined to form a hybridization curve. In another embodiment, hybridization levels are measured in real time using a single microarray. In this embodiment, the microarray is allowed to hybridize to the sample without interruption and the microarray is interrogated at each hybridization time in a non-invasive manner. In still another embodiment, one can use one array, hybridize for a short time, wash and measure the hybridization level, put back to the same sample, hybridize for another period of time, wash and measure again to get the hybridization time curve.
In a specific embodiment, the AffymetrixC~ Human Genome U133 (HG-U133) Set, consisting of two GeneChip~ arrays, is used in accordance with known methods.
The Human Genome U133 (HG-U133) Set contains almost 45,000 probe sets representing more than 39,000 transcripts derived from approximately 33,000 well-substantiated human genes. This set design uses sequences selected from GenBank~, dbEST, and RefSeq. The sequence clusters were created from the UniGene database (Build 133, April 20, 2001).
They were then refined by analysis and comparison with a number of other publicly available databases including the Washington University EST trace repository and the University of California, Santa Cruz Golden Path human genome database (April release).

WO 2006/002240 . PCT/US2005/022071 In another embodiment, the HG-U133A array is used in accordance with the methods of the invention. The HG-U133A array includes representation of the RefSeq database sequences and probe sets related to sequences previously represented on the Human Genome U95Av2 array. The HG-U133B array contains primarily probe sets representing EST clusters. In another embodiment, the U133 Plus 2.0 GeneChip~
is used in the invention. The U133 Plus 2.0 GeneChip~ represents over 47,000 transcripts.
In another embodiment, a cDNA based microarray is used. In one embodiment the ChondroChipTM is used in accordance with the methods of the invention. The ChondroChipTM is a cDNA based microarray. One version of the ChondroChipTM
includes 14,976 distinct elements: 10,382 known genes (69%), 4,112 EST/genomic DNA
matches (28%), 328 clones with no significant match (2.2%), and 154 control spots (1.0%). Most if not all of the elements on the ChondroChipTM are complementary to ESTs identified as expressed in human chondrocytes. An article that describes the creation of a version of the ChondroChipTM is Zhang et al., 2002, Osteoarthritis and Cartilage 10, 950-960, which is hereby incorporated by reference in its entirety.
In another embodiment, the BloodChipTM is used in accordance with the methods of the invention. The BloodChip is a cDNA microarray slide with 10,368 PCR
products derived from peripheral blood cell cDNA libraries. The creation of the BloodChipTM
microarray is described in Ma and Liew, 2003, Journal of Molecular and Cellular Cardiology 8, 993-998, which is hereby incorporated by reference in its entirety.
5.3.1.2 TARGET POLYNUCLEOTIDE MOLECULES
Target polynucleotides that can be analyzed by the methods and compositions of the invention include RNA molecules such as, but by no means limited to, expressed RNA
molecules which includes messenger RNA (mRNA) molecules, mRNA spliced variants as well as other regulatory RNA, cRNA molecules (e.g., RNA molecules prepared from cDNA molecules that are transcribed in vivo) and fragments thereof. Target polynucleotides which may also be analyzed by the methods and compositions of the present invention include, but are not limited to DNA molecules such as genomic DNA
molecules, cDNA molecules, and fragments thereof including oligonucleotides, ESTs, STSs, etc The target polynucleotide molecules may be naturally occurring nucleic acid molecules such as genomic or extragenomic DNA molecules isolated from a blood sample, or RNA molecules, such as mRNA molecules, isolated from a blood sample. The sample of target polynucleotides can comprise, e.g., molecules of DNA, RNA, or copolymers of DNA and RNA. In specific embodiments, the target polynucleotides of the invention will correspond to particular genes or to particular gene transcripts (e.g., to particular mRNA
sequences expressed in specific cell types or to particular cDNA sequences derived from such mRNA sequences). The target polynucleotides may correspond to different exons of S the same gene, e.g., so that different splice variants of that gene may be detected andlor analyzed.
In specific embodiments, the target polynucleotides to be analyzed are prepared in vitro from nucleic acids extracted from a blood sample. For example, in one embodiment, RNA is extracted from a blood sample (e.g., total cellular RNA, poly(A)+
messenger RNA, fraction thereof) and messenger RNA is purified from the total extracted RNA.
Methods for preparing total and poly(A)+ RNA are well known in the art, and are described generally, e.g., in Sambrook, Fritsch 8~ Maniatis, "Molecular Cloning: A
Laboratory Manual (1982); "DNA Cloning: A Practical Approach," Volumes I and II
(D.N. Glover ed. 198S). In one embodiment, RNA is extracted from a blood sample using 1 S guanidinium thiocyanate lysis followed by CsCI centrifugation and an oligo dT
purification (Chirgwin et al., 1979, Biochemistry 18:5294-5299). In another embodiment, RNA is extracted from a blood sample using guanidinium thiocyanate lysis followed by pwification on RNeasy columns (Qiagen). cDNA is then synthesized from the purified mRNA using, e.g., oligo-dT or random primers. In specific embodiments, the target polynucleotides are cRNA prepared from purified messenger RNA extracted from a blood sample. As used herein, cRNA is defined here as RNA complementary to the source RNA. The extracted RNAs can be amplified using a process in which doubled-stranded cDNAs are synthesized from the RNAs using a primer linked to an RNA polymerase promoter in a direction capable of directing transcription of anti-sense RNA.
Anti-sense 2S RNAs or cRNAs axe then transcribed from the second strand of the double-stranded cDNAs using an RNA polymerase (see, e.g., U.S. Patent Nos. 5,891,636, 5,716,785;
S,S4S,S22 and 6,132,997; see also, U.S. Patent No. 6,271,002, and U.S.
Provisional Patent Application Serial No. 60/253,641, filed on November 28, 2000, by Ziman et al.). Both oligo-dT primers (U.S. Patent Nos. S,S4S,S22 and 6,132,997) or random primers (U.S.
Provisional Patent Application Serial No. 60/253,641, filed on November 28, 2000, by Ziman et al.) that contain an RNA polymerase promoter or complement thereof can be used. In some embodiments the target polynucleotides are short and/or fragmented polynucleotide molecules which are representative of the original nucleic acid population of the blood sample.

The target polynucleotides to be analyzed by the methods and compositions of the invention can be detectably labeled. For example, cDNA can be labeled directly, e.g., with nucleotide analogs, or indirectly, e.g., by making a second, labeled cDNA
strand using the first strand as a template. Alternatively, the double-stranded cDNA can be transcribed into cRNA and labeled.
In some embodiments the detectable label is a fluorescent label, e.g., by incorporation of nucleotide analogs. Other labels suitable for use in the present invention include, but are not limited to, biotin, imminobiotin, antigens, cofactors, dinitrophenol, lipoic acid, olefmic compounds, detectable polypeptides, electron rich molecules, enzymes capable of generating a detectable signal by action upon a substrate, and radioactive isotopes. Suitable radioactive isotopes include 32P, 355, 14C, isN and l2sl.
Fluorescent molecules suitable for the present invention include, but are not limited to, fluorescein and its derivatives, rhodamine and its derivatives, texas red, 5'carboxy-fluorescein ("FMA"), 2',7'-dimethoxy-4',5'-dichloro-6-carboxy-fluorescein ("JOE"), N,N,N',N'-tetramethyl-6-carboxy-rhodamine ("TAMRA"), 6'carboxy-X-rhodamine ("ROX"), HEX, TET, IRD40, and IRD41. Fluroescent molecules that are suitable for the invention further include:
cyamine dyes, including by not limited to Cy3, Cy3.5 and CyS; BODIPY dyes including but not limited to BODIPY-FL, BODIPY-TR, BODIPY-TMR, BODIPY-630/650, and BODIPY-650/670; and ALEXA dyes, including but not limited to ALEXA-488, ALEXA-532, ALEXA-546, ALEXA-568, and ALEXA-594; as well as other fluorescent dyes which will be known to those who are skilled in the art. Electron rich indicator molecules suitable for the present invention include, but are not limited to, ferritin, hemocyanin, and colloidal gold. Alternatively, in less some embodiments the target polynucleotides may be labeled by specifically complexing a first group to the polynucleotide. A
second group, covalently linked to an indicator molecules and which has an affinity for the first group, can be used to indirectly detect the target polynucleotide. In such an embodiment, compounds suitable for use as a first group include, but are not limited to, biotin and iminobiotin. Compounds suitable for use as a second group include, but are not limited to, avidin and streptavidin.
In a specific embodiment, the target polynucleotides are prepaxed as follows:
2~,g Oligo-dT primers are annealed to 2~,g of mRNA isolated from a blood sample of a patient in a total volume of 15p.1, by heating to 70°C for 10 min, and cooled on ice. The mRNA is reverse transcribed by incubating the sample at 42°C for 1.5-2 hours in a 100 ~.1 volume containing a final concentration of SOmM Tris-HCl (pH 8.3), 75mM KCI, 3mM
MgCl2, 25mM DTT, 25mM unlabeled dNTPs, 400 units of Superscript II (200U/~L, Gibco BRL), and lSmM of Cy3 or C.yS (Amersham). RNA is then degraded by addition of 15.1 of O.1N NaOH, and incubation at 70 0 C for 10 min. The reaction mixture is neutralized by addition of 151 of O.1N HCI, and the volume is brought to 500,1 with TE (lOmM
Tris, 1 mM EDTA), and 20 ~ g of Cotl human DNA (Gibco-BRL) is added.
The labeled target polynucleotide molecules are purified by centrifugation in a Centricon-30 micro-concentrator (Amicon). If two different target polynucleotide samples (e.g., two samples derived from a healthy patient vs. patient with a disease) are being analyzed and compared by hybridization to the same array, each target nucleic acid sample is labeled with a different fluorescent label (e.g., Cy3 and Cy5) and separately concentrated. The separately concentrated target nucleic acid samples (Cy3 and Cy5 labeled) are combined into a fresh centricon, washed with 500,1 TE, and concentrated again to a volume of less than 7~1. 1 ~.L of 10~.g/~.l polyA RNA (Sigma, #P9403) and 1 ~,1 of 10~,g/~,1 tRNA (Gibco-BRL, #15401-O11) is added and the volume is adjusted to 9.5,1 with distilled water. For final target polynucleotide preparation 2.1~,120XSSC
(1.5M
NaCI, 150mM NaCitrate (pH8.0)) and 0.35,1 10%SDS is added.
5.3.1.3 HYBRIDIZATION TO MICROARRAYS
In some embodiments, nucleic acid hybridization and wash conditions are chosen so that the polynucleotide molecules to be analyzed by the invention (e.g., "target polynucleotide molecules) specifically bind or specifically hybridize to the complementary polynucleotide sequences of the array, typically to a specific array site, where its complernentaxy DNA is located.
Arrays containing double-stranded probe DNA situated thereon can be subjected to denaturing conditions to render the DNA single-stranded prior to contacting with the target polynucleotide molecules. Arrays containing single-stranded probe DNA (e.g., synthetic oligodeoxyribonucleic acids) may need to be denatured prior to contacting with the target polynucleotide molecules, e.g., to remove hairpins or dimers which form due to self complementary sequences.
Optimal hybridization conditions will depend on the length (e.g., oligomer versus polynucleotide greater than 200 bases) and type (e.g., RNA, or DNA) of probe and target nucleic acids. General parameters for specific (i.e., stringent) hybridization conditions for nucleic acids are described in Sambrook et al., (supra), and in Ausubel et al., 1987, Current Protocols in Molecular Biology, Greene Publishing and Wiley-Interscience, New York. When the cDNA microarrays of Schena et al. are used, typical hybridization conditions are hybridization in 5 X SSC plus 0.2% SDS at 65 °C for four hours, followed by washes at 25°C in low stringency wash buffer (1 X SSC plus 0.2%
SDS), followed by minutes at 25°C in higher stringency wash buffer (0.1 X SSC plus 0.2%
SDS) (Shena et al., 1996, Proc. Natl. Acad. Sci. U.S.A. 93:10614). Useful hybridization conditions are 5 also provided in, e.g., Tijessen, 1993, Hybridization With Nucleic Acid Probes, Elsevier Science Publishers B.V. and Kricka, 1992, Nonisotopic DNA Probe Techniques, Academic Press, San Diego, CA.
Representative hybridization conditions for use with the screening and/or signaling chips in accordance with some embodiments of the present invention include hybridization 10 at a temperature at or near the mean melting temperature of the probes (e.g., within 5 °C, more typically within 2 °C) in 1 M NaCI, 50 mM MES buffer (pH 6.5), 0.5% sodium Sarcosine and 30% formamide.
In a specific embodiment, a labeled target polynucleotide molecules are denatured by heating for two minutes at 100°C, and incubated at 37°C for 20-30 min before being placed on a nucleic acid array under a 22mm x 22mm glass cover slip.
Hybridization is carried out at 65°C for fourteen to eighteen hours in a custom slide chamber with humidity maintained by a small reservoir of 3XSSC. The array is washed by submersion and agitation for between two and five minutes in 2X SSC with 0.1%SDS, followed by SSC, and O.1X SSC. Finally, the array is dried by centrifugation for 2 min in a slide rack in a Beckman GS-6 tabletop centrifuge in Microplus carriers at 650 RPM for two minutes.
5.3.1.4 SIGNAL DETECTION AND DATA ANALYSIS
It will be appreciated that when target sequences, e.g., cDNA or cRNA, complementary to the RNA of a blood sample is made and hybridized to a microarray under suitable hybridization conditions, the level of hybridization to the site in the array corresponding to an exon of any particular gene will reflect the prevalence in the cell of mRNA or mRNAs containing the exon transcribed from that gene. For example, when detectably labeled (e.g., with a fluorophore) cDNA complementary to the total cellular mRNA is hybridized to a microarray, the site on the array corresponding to an exon of a gene (i.e., capable of specifically binding the product or products of the gene expressing) that is not transcribed or is removed during RNA splicing in the cell will have little or no signal (e.g., fluorescent signal), and an exon of a gene for which the encoded mRNA
expressing the exon is prevalent will have a relatively strong signal. The relative abundance of different mRNAs produced from the same gene by alternative splicing is then determined by the signal strength pattern across the whole set of exons monitored for the gene.
Generally, any form of image processing may be used to digitize the microarrays and thereby obtain high throughput data for molecular markers in the present invention.
For example, any of the image processing techniques described or referenced in Draghici, Data Ahalysis Tools For' DNA Mic~oa~rays, 2003, Chapman & Hall, CRC Press, New York, pp. 33-58, which is hereby incorporated by reference in its entirety, can be used. In some embodiments, two-color fluorescence is used. The use of a two-color fluorescence labeling and detection scheme to define alterations in gene expression has been described in connection with detection of mRNAs, e.g., in Shena et al., 1995, Science 270:467-470, which is hereby incorporated by reference in its entirety for all purposes.
The scheme is equally applicable to labeling and detection of exons. An advantage of using target sequences, e.g., cDNAs or cRNAs, labeled with two different fluorophores is that a direct and internally controlled comparison of the mRNA or exon expression levels corresponding to each arrayed gene in two states can be made, and variations due to minor differences in experimental conditions (e.g., hybridization conditions) will not affect subsequent analyses.
In a specific embodiment, the labeled probes are scanned using a GMS Scanner 418 and Scananlzyer software (Michael Eisen, Stamford University), followed by GeneSpring software (Silcon Genetics, CA) analysis. In another embodiment, a GMS
Scanner 428 and Jaguar software are used followed by GeneSpring software analysis. In some embodiments a normalization routine, such as any of the normalization routines described in Section 5.7, is used.
5.3.2 RT-PCR AND QUANTITATIVE RT-PCR
In one aspect of the invention, the abundance or level of expression of an RNA
product of a molecular marker can be measured performing reverse transcription on the RNA from blood and subsequently amplifying the resulting product ("RT-PCR").
In another embodiment, the abundance or level of expression of RNA can be measured from a blood sample by using quantitative RT-PCR or real time PCR ("QRT-PCR") on cDNA
copy of RNA. Total RNA, or mRNA from a blood sample can be used as a template and a primer specific to the transcribed portion of a gene of the invention is used to initiate reverse transcription. Methods of reverse transcribing RNA into cDNA are well known and described in, for example, Sambrook et al., 1989, supra. Primer design can be accomplished utilizing commercially available. software (e.g., Primer Designer 1.0, academic software, etc.). The product of the reverse transcription is subsequently used as a template for PCR. In one embodiment, a one step process can be used for either the RT-PCR and/or the QRT-PCR(combining the reverse transcription and PCR in a single reaction). In another embodiment, a two step process can be used for either the RT-PCR
and/or the QRT-PCR (first doing the reverse transcription step and subsequently performing the PCR) . In some embodiments, oligo(dT)-primed first strand cDNA
synthesis is performed so as to specifically target the mRNA population (e.g.
using the Applied Biosystems High Capacity cDNA Archive I~it (cat # 4322171), on a Perkin-Elmer DNA Thermal Cycler.
For quantitative RT-PCR, in some embodiments the reportable value is the Ct-value, which is the threshold cycle at which PCR is in the logarithmic phase.
For each gene of interest in each RNA sample, a ACt value can be calculated by the formula: OCt =
(Ct, target gene) - (Ct,(3-actin). The ~Ct values from different groups of RNA
samples can then be compared by the Mann-Whitney Rank Sum test.
In some embodiments, Quantitative RT-PCR can be done using probes including Taqman~ probes (Perkin Elmer, Foster City, California), The probe is specific for the PCR product and has both a quencher and fluorescent reporter attached to the probe.
Different fluorescent markers can be utilized. In some embodiments, multiple probes can be used in the quantitative RT-PCR process to allow for multiplexing reactions (e.g. allow for measurement of two molecular markers in one reaction well or container).
When using TaqMan~ probes, Taq DNA polymerase is used which has 5'-to-3' exonuclease activity and thus will cleave of the fluorescent reporter of the probe, freeing the fluorescent molecular from the quencher molecule. Thus the emission of fluorescence is is used to measure the amount of PCR product being made. Other probes are also useful for quantitative RT-PCR including Molecular Beacons~.
Other known techniques for quantitative Rt-PCR is to use an intercolating dye such as the commercially available QuantiTectTM SYBR~ Green PCR (Qiagen, Valencia California).
Additionally, other systems to quantitatively measure mRNA expression products are known including Scorpions~ (Zeneca Limited) or Fluorescent Polarization Probes (see e.g. Ze~eca Limited, 6,007,984) etc.
5.3.3 NUCLEASE PROTECTION ASSAYS
Nuclease protection assays (including both ribonuclease protection assays and S 1 nuclease assays) can be used to detect and quantify specific products of molecular markers. In nuclease protection assays, an antisense probe (labeled with, e.g., radiolabeled or nonisotopic) hybridizes in solution to an RNA sample. Following hybridization, single-stranded, unhybridized probe and RNA are degraded by nucleases. An acrylamide gel is used to separate the remaining protected fragments. Typically, solution hybridization is more efficient than membrane-based hybridization, and it can accommodate up to 100 ~.g of sample RNA, compared with the 20-30 ~g maximum of blot hybridizations.
The ribonuclease protection assay, which is the most common type of nuclease protection assay, requires the use of RNA probes. Oligonucleotides and other single-stranded DNA probes can only be used in assays containing S 1 nuclease. The single-stranded, antisense probe must typically be completely homologous to target RNA to prevent cleavage of the probeaarget hybrid by nuclease.
5.3.4 MASS SPECTROMETRY
Mass spectrometry (e.g., electrospray ionization "ESI", matrix-assisted laser desorption-ionization "MALDI ", and Fourier-transform ion cyclotron resonance "FT-ICR") can be used to measure data (e.g., mass, charge) of molecular markers in blood samples. Such molecular markers that can be characterized by mass spectrometry include but are not limited to, proteins, nucleic acids, carbohydrates, and other biological macromolecules. This section provides brief and non-limiting examples of mass spectrometry techniques that can be used to quantitatively characterize molecular markers.
MALDI uses a pulsed laser for desorption of the ions and a time-of flight analyzer and has been used for the detection of noncovalent tRNA:amino-acyl-tRNA
synthetase complexes. See, for example, Gruic-Sovulj et al., 1997, J. Biol. Chem.
272:32084. ESI
mass spectrometry ("ESI-MS") has been used for studying non-covalent molecular interactions. ESI-MS generates molecular ions with little to no fragmentation.
See, for example, Xavier et al., 2000, Trends Biotechnol. 18:349. Fourier-transform ion cyclotron resonance ("FT-ICR") mass spectrometry provides high-resolution spectra, isotope-resolved precursor ion selection, and accurate mass assignments. See, for example, Xavier et al., 2000, Trends Biotechnol. 18:349.
Tandem mass spectrometry is described in Link et al., 1999, Nat. Biotechnol.
17, 676-682; Washburn et al. 2001, Nat. Biotechnol. 19, 242; Gaven et al., 2002, Nature 415, 141; and Ho et al., 2002, Nature 418, 180). In the case of proteins from a blood sample, the proteins can first be digested into peptides using an enzyme such as trypsin and then subjected to liquid chromatography tandem mass spectrometry (MS/MS). Liquid chromatography provides an initial separation of the peptides, which are then ionized directly into a mass spectrometer. Following an initial scan in which the mass/charge ratio of all intact (parent) ions from the peptides are measured, the mass spectrometer selects a parent ion, fragments it and obtains the mass spectrum of the generated fragments. These fragmentation patterns are called tandem mass spectra or MS/MS spectra. This process of ion selection and fragmentation is repeated throughout the liquid chromatography separation, thus generating a set of time resolved MS/MS spectra, with each spectrum representing a species eluting at a particular time from the LC separation.
The resolving power of the liquid chromatography step, combined with the high mass resolution of modern mass spectrometers typically assures that each MS/MS spectrum represents the fragmentation pattern of a unique peptide in the digest.
5.3.5 COMPARATIVE GENE EXPRESSION PROFILING
In some embodiments of the present invention quantitative measurement of molecular marker data is performed using comparative gene-expression profiling. An example of such technology is the multiplex microsphere bead assay used by Fuja et al., 2004, Journal of Biotechnology 108, 193,.
5.3.6 TRANSCRIPTION BASED AMPLIFICATION SYSTEMS
In another aspect of the invention, the level of expression of a molecular marker in blood can be measured by amplifying RNA from a blood sample using transcription based amplification systems (TAS), including nucleic acid sequence amplification (NASBA) and 3SR. See, e.g., Kwoh et al., 1989, PNAS USA 86:1173;
International Publication No. WO 88/10315; and U.S. Patent No. 6,329,179. In NASBA, the nucleic acids can be prepared for amplification using conventional phenol/chloroform extraction, heat denaturation, treatment with lysis buffer and minispin columns for isolation of DNA and RNA or guanidinium chloride extraction of RNA. These amplification techniques involve annealing a primer that has target specific sequences.
Following polymerization, DNA/RNA hybrids are digested with RNase H while double stranded DNA molecules are heat denatured again. In either case, the single stranded DNA is made fully double stranded by addition of second target specific primer, followed by polymerization. The double-stranded DNA molecules are then multiply transcribed by a polymerase such as T7 or SP6. In an isothermal cyclic reaction, the RNA's are reverse transcribed into double stranded DNA, and transcribed once with a polymerase such as T7 or SP6. The resulting products, whether truncated or complete, indicate taxget specific sequences.

5.3.7 ADDITIONAL TECHNIQUES FOR DETECTING AND
QUANTIFYING RNA
Many other techniques are known to one of skill for detecting and measuring RNA
and can be used in accordance with the methods of the invention. Non-limiting examples of such techniques include Northern blotting, nuclease protection assays, RNA
fingerprinting, polymerase chain reaction, ligase chain reaction, Qbeta replicase, isothermal amplification method, strand displacement amplification, transcription based amplification systems, nuclease protection (SI nuclease or RNAse protection assays) as well as methods disclosed in International Publication Nos. WO 88/10315 and WO
89/06700, and International Applications Nos. PCT/LJS87/00880 and PCT/US89/01025.
A standard Northern blot assay can be used to ascertain an RNA transcript size, identify alternatively spliced RNA transcripts, and the relative amounts of mRNA in a blood sample, in accordance with conventional Northern hybridization techniques known to those persons of ordinary skill in the art. In Northern blots, RNA samples are first separated by size via electrophoresis in an agarose gel under denaturing conditions. The RNA is then transferred to a membrane, crosslinked and hybridized with a labeled probe.
Nonisotopic or high specific activity radiolabeled probes can be used including random-primed, nick-translated, or PCR-generated DNA probes, in vitro transcribed RNA
probes, and oligonucleotides. Additionally, sequences with only partial homology (e.g., cDNA
from a different species or genomic DNA fragments that might contain an exon) may be used as probes. The labeled probe, e.g., a radiolabelled cDNA, either containing the full-length, single stranded DNA or a fragment of that DNA sequence may be at least 20, at least 30, at least 50, or at least 100 consecutive nucleotides in length. The probe can be labeled by any of the many different methods known to those skilled in this art. The labels most commonly employed for these studies are radioactive elements, enzymes, chemicals that fluoresce when exposed,to ultraviolet light, and others. A number of fluorescent materials are known and can be utilized as labels. These include, but are not limited to, fluorescein, rhodamine, auramine, Texas Red, AMCA blue and Lucifer Yellow. A
articulax detecting material is anti-rabbit antibody prepared in goats and conjugated with fluorescein through an isothiocyanate. Proteins can also be labeled with a radioactive element or with an enzyme. The radioactive label can be detected by any of the currently available counting procedures. Non-limiting examples of isotopes include 3H, 14C, 32P, 355, 36C1, SlCr, 57Co, 58Co, 59Fe, 90Y, 125I, 131I, and 186Re. Enzyme labels are likewise useful, and can be detected by any of the presently utilized colorimetric, pectrophotometric, fluorospectrophotometric, amperometric or gasometric techniques.
The enzyme is conjugated to the selected particle by reaction with bridging molecules such as carbodiimides, diisocyanates, glutaraldehyde and the like. Any enzymes known to one of skill in the art can be utilized. Examples of such enzymes include, but are not limited to, peroxidase, beta-D-galactosidase, unease, glucose oxidase plus peroxidase and alkaline phosphatase. U.S. Patent Nos. 3,654,090, 3,850,752, and 4,016,043 are referred to by way of example for their disclosure of alternate labeling material and methods.
Nuclease protection assays (including both ribonuclease protection assays and nuclease assays) can be used to detect and quantitate specific mRNAs. In nuclease protection assays, an antisense probe (labeled with, e.g., radiolabeled or nonisotopic) hybridizes in solution to an RNA sample. Following hybridization, single-stranded, unhybridized probe and RNA are degraded by nucleases. An acrylamide gel is used to separate the remaining protected fragments. Typically, solution hybridization is more efficient than membrane-based hybridization, and it can accommodate up to 100 pg of sample RNA, compaxed with the 20-30 ~g maximum of blot hybridizations.
The ribonuclease protection assay, which is the most common type of nuclease protection assay, requires the use of RNA probes. Oligonucleotides and other single-stranded DNA probes can only be used in assays containing S1 nuclease. The single-stranded, antisense probe must typically be completely homologous to target RNA to prevent cleavage of the probeaarget hybrid by nuclease.
Additional techniques to quantitatively measure RNA expression include, but are not limited to, ligase chain reaction, Qbeta replicase (see, e.g., International Application No. PCT/IJS87/00880), isothermal amplification method (see, e.g., Walker et a1.,1992, PNAS 89:382-396), strand displacement amplification (SDA), repair chain reaction, Asymmetric Quantitative PCR (see, e.g., U.S. Publication No. US200330134307A1) and the multiplex microsphere bead assay described in Fuja et al., 2004, Journal of Biotechnology 108:193-205.
5.3.7.1 SEPARATION OF AMPLIFICATION PRODUCTS
Some of the quantitative measurement techniques described above may require separation of amplification products. Several techniques can be used to separate such amplification products. For example, amplification products can be separated by agaxose, agarose-acrylamide or polyacrylamide gel electrophoresis using conventional methods.
Several techniques for detecting PCR products quantitatively without electrophoresis can also be used according to the invention. See, for example, CR P~otoeols, A
Guide to Methods and Applications, Innis et al., Academic Press, Inc. N.Y., 1990). For example, chromatographic techniques can be employed to effect separation. There are many kinds of chromatography that can be used in the present invention: adsorption, partition, ion-exchange and molecular sieve, HPLC, and many specialized techniques for using them including column, paper, thin-layer and gas chromatography (Freifelder, Physical Biochemistry Applications to Biochemistry and Molecular Biology, 2nd ed., Wm.
Freeman and Co., New York, N.Y., 1982).
Another example of a separation methodology is done by covalently labeling the oligonucleotide primers used in a PCR reaction with various types of small molecule ligands. In one such separation, a different ligand is present on each oligonucleotide. A
molecule, perhaps an antibody or avidin if the ligand is biotin, that specifically binds to one of the ligands is used to coat the surface of a plate such as a 96 well ELISA plate.
Upon application of the PCR reactions to the surface of such a prepared plate, the PCR
products are bound with specificity to the surface. After washing the plate to remove unbound reagents, a solution containing a second molecule that binds to the first ligand is added. This second molecule is linked to some kind of reporter system. The second molecule only binds to the plate if a PCR product has been produced whereby both oligonucleotide primers are incorporated into the final PCR products. The amount of the PCR product is then detected and quantified in a commercial plate reader much as ELISA
reactions are detected and quantified. An ELISA-like system such as the one described here has been developed by the Raggio Italgene company under the C-Track trade name.
5.3.7.2 VISUALIZATION OF AMPLIFICATION PRODUCTS
Some of the quantitative measurement techniques described above may require visualization of amplification products. Amplification products are visualized, for example, in order to confirm amplification of the marker sequences. One typical visualization method involves staining of a gel with ethidium bromide and visualization under UV light. Alternatively, if the amplification products are integrally labeled with radio- or fluorometrically-labeled nucleotides, the amplification products may then be exposed to x-ray film or visualized under the appropriate stimulating spectra, following separation.
In one embodiment, visualization is achieved indirectly. Following separation of amplification products, a labeled, nucleic acid probe is brought into contact with the amplified marker sequence. The probe preferably is conjugated to a chromophore but may be radiolabeled. In another embodiment, the probe is conjugated to a binding partner, such as an antibody or biotin, where the other member of the binding pair carries a detectable moiety.
In another embodiment, detection is by Southern blotting and hybridization with a labeled probe. The techniques involved in Southern blotting are well known to those of skill in the art and may be found in many standard books on molecular protocols. See Sambrook et al., 1989. Briefly, amplification products are separated by gel electrophoresis. The gel is then contacted with a membrane, such as nitrocellulose, permitting transfer of the nucleic acid and non-covalent binding.
Subsequently, the membrane is incubated with a chromophore-conjugated probe that is capable of hybridizing with a target amplification product. Detection is by exposure of the membrane to x-ray film or ion-emitting detection devices.
~ne example of the foregoing is described in U.S. Pat. No. 5,279,721, incorporated by reference herein, which discloses an apparatus and method for the automated electrophoresis and transfer of nucleic acids. The apparatus permits electrophoresis and blotting without external manipulation of the gel and is ideally suited to carrying out methods according to the present invention.
5.4 METHODS FOR MEASURING MOLECULAR MARKER DATA
REFLECTIVE OF ABUNDANCE OF PROTEIN PRODUCTS OF
MOLECULAR MARKERS
Measurement of the abundance of protein products of molecular markers in blood may be performed using a number of separation techniques combined with a monitoring system. For example, whole genome monitoring of protein (e.g., the "proteome,") can be carried out using commercial systems such as a SELDI~ Chip by Ciphergen. In addition, protein microarrays comprised of immobilized, preferably monoclonal, antibodies specific to a plurality of protein species encoded by the cell genome can be used.
(e.g., The ProteinChip~ Biomarker System, Ciphergen, Fremont, California). See also, for example, Lin, 2004, Modern Pathology, 1-9; Li, 2004, Journal of Urology 171, 1782-1787;
Wadsworth, 2004, Clinical Cancer Research, 10, 1625-1632; Prieto, 2003, Journal of Liquid Chromatography & Related Technologies 26, 2315-2328; Coombes, 2003, Clinical Chemistry 49, 1615-1623; Mian, 2003, Proteomics 3, 1725-1737; Lehre et al., 2003, BJU
International 92, 223-225; and Diamond, 2003, Journal of the American Society for Mass Spectrometry 14, 760-765, which are hereby incorporated by reference in their entireties.

In one embodiment, antibodies can be used to measure protein products of the candidate molecular markers. Methods for making monoclonal antibodies are well known (see, e.g., Harlow and Lane, 1988, Autib~dies: A Laboratory Manual, Cold Spring Harbor, New York, which is incorporated in its entirety for all purposes). In one embodiment, monoclonal antibodies are raised against synthetic peptide fragments designed based on genomic sequence of the cell. With such an antibody array, proteins from the cell are contacted to the array and their binding is assayed with assays known in the art.
Immunoassays known to one of skill in the art can be used to detect and quantify protein levels. For example, ELISAs can be used to detect and quantify protein levels.
ELISAs comprise preparing antigen, coating the well of a 96 well microtiter plate with the antigen, adding the antibody of interest conjugated to a detectable compound such as an enzymatic substrate (e.g., horseradish peroxidase or alkaline phosphatase) to the well and incubating for a period of time, and detecting the presence of the antigen. In ELISAs the antibody of interest does not have to be conjugated to a detectable compound;
instead, a second antibody (which recognizes the antibody of interest) conjugated to a detectable compound may be added to the well. Further, instead of coating the well with the antigen, the antibody may be coated to the well. In this case, a second antibody conjugated to a detectable compound may be added following the addition of the antigen of interest to the coated well. One of skill in the art would be knowledgeable as to the parameters that can be modified to increase the signal detected as well as other variations of ELISAs known in the art. In a preferred embodiment, an ELISA may be performed by coating a high binding 96-well microtiter plate (Costar) with 2~,g/ml of rhu-IL-9 in PBS
overnight.
Following three washes with PBS, the plate is incubated with three-fold serial dilutions of Fab at 25oC for 1 hour. Following another three washes of PBS, 1 p,g/ml anti-human kappa-alkaline phosphatase-conjugate is added and the plate is incubated for 1 hour at 25°C. Following three washes with PBST, the alkaline phosphatase activity is determined in 50~1/AMP/PPMP substrate. The reactions are stopped and the absorbance at 560 nm is determined with a VMAX microplate reader. For further discussion regarding ELISAs see, e.g., Ausubel et al, eds, 1994, Current Protocols in Molecular Biology, Vol. 1, John Wiley & Sons, Inc., New York at 11.2.1.
Protein levels may be determined by Western blot analysis. Further, protein levels as well as the phosphorylation of proteins can be determined by immunoprecitation followed by Western blot analysis. Tmmunoprecipitation protocols generally comprise lysing a population of cells in a lysis buffer such as RIPA buffer (1% NP-40 or Triton X-I00, 1 % sodium deoxycholate, 0.1 % SDS, O. I S M NaCI, 0.01 M sodium phosphate at pH
7.2, 1% Trasylol) supplemented with protein phosphatase and/or protease inhibitors (e.g., EDTA, PMSF, aprotinin, sodium vanadate), adding the antibody of interest to the cell lysate, incubating for a period of time (e.g., 1 to 4 hours) at 40° C, adding protein A and/or protein G sepharose beads to the cell lysate, incubating for about an hour or more at 40° C, washing the beads in lysis buffer and resuspending the beads in SDS/sample buffer. The ability of the antibody of interest to immunoprecipitate a particular antigen can be assessed by, e.g., western blot analysis. One of skill in the art would be knowledgeable as to the parameters that can be modified to increase the binding of the antibody to an antigen and decrease the background (e.g., pre-clearing the cell lysate with sepharose beads). For further discussion regarding immunoprecipitation protocols see, e.g., Ausubel et al, eds, 1994, Current Protocols in Molecular Biology, Vol. l, John Wiley & Sons, Inc., New York at 10.16.1.
Western blot analysis generally comprises preparing protein samples, electrophoresis of the protein samples in a polyacrylamide gel (e.g., 8%- 20%
SDS-PAGE
depending on the molecular weight of the antigen), transferring the protein sample from the polyacrylamide gel to a membrane such as nitrocellulose, PVDF or nylon, incubating the membrane in blocking solution (e.g., PBS with 3% BSA or non-fat milk), washing the membrane in washing buffer (e.g., PBS-Tween 20), incubating the membrane with primary antibody (the antibody of interest) diluted in blocking buffer, washing the membrane in washing buffer, incubating the membrane with a secondary antibody (which recognizes the primary antibody, e.g., an anti-human antibody) conjugated to an enzymatic substrate (e.g., horseradish peroxidase or alkaline phosphatase) or radioactive molecule (e,g_~ 32p or 12SI) diluted in blocking buffer, washing the membrane in wash buffer, and detecting the presence of the antigen. One of skill in the art would be knowledgeable as to the parameters that can be modified to increase the signal detected and to reduce the background noise. For further discussion regarding western blot protocols see, e.g., Ausubel et al, eds, 1994, Current Protocols in Molecular Biology, Vol. 1, John Wiley &
Sons, Inc., New York at 10.8.1.
Protein expression levels can also be separated by two-dimensional gel electrophoresis systems. Two-dimensional gel electrophoresis is well-known in the art and typically involves isoelectric focusing along a first dimension followed by SDS-PAGE
electrophoresis along a second dimension. See, e.g., Hames et al., 1990, Gel Elect~~ophor~esis of P~otei~s: A Practical Approach, IRL Press, New York;
Shevchenko et al., 1996, Pr~c. Natl. Acad. Sci. USA 93:1440-1445; Sagliocco et al., 1996, Yeast 12:1519-1533; Lander, 1996, Science 274:536-539. The resulting electropherograms can be analyzed by numerous techniques, including mass spectrometric techniques, Western blotting and immunoblot analysis using polyclonal and monoclonal antibodies, and internal and N-terminal micro-sequencing.
5.5 USES OF CLASSIFIERS IDENTIFIED
In exemplary embodiments, the classifiers constructed in accordance with the present invention can be used to detect, diagnose, prognose and/or monitor a trait in a test individual. In a specific embodiment, a classifier or classifier group constructed in accordance with Section 5.1 is used to detect, diagnose, prognose and/or monitor a disease in said test individual. In another embodiment, a classifier or classifier group constructed in accordance with Section 5.1 is used to detect, diagnose, prognose and/or monitor a reoccurrence of disease in said test individual. In another embodiment, a classifyer or classifier group constructed in accordance with the invention is used to evaluate or predict the efficacy of treatment in a subject. In another embodiment, a classifier or classifier group constructed in accordance with the invention is used to predict whether a subject will be responsive to treatment and/or treatment outcomes. In another embodiment, a classifier or classifier group constructed in accordance with the invention is used to monitor and/or predict treatment compliance or non-compliance. In another embodiment, a classifier or classifier group constructed in accordance with the methods of the invention is indicative of the responsiveness of a subject to a stimulus (whether external or internal, e.g., smoke, pollution, sunlight, heat, and mutations) and is used to evaluate or predict the response of a subject to such stimulus.
In yet another embodiment, the molecular markers identified by the classifiers of the invention can be used independently of the classifier to detect, diagnose, prognose, predict and/or monitor a trait. In such embodiments, said detection, diagnosis, prognosis, prediction and/or monitoring of a test individual can be accomplished by monitoring the gene expression pattern or profile of the molecular markers identified by the classifier or classifier group of the invention of a test individual and comparing said pattern or profile to a gene expression pattern or profile of a control individual or group of individuals who have said trait. In another embodiment, the gene expression pattern or profile of the test individual can be compared with a control individual or group of individuals who do not have said trait. In another embodiment, the gene expression pattern or profile of the test individual is compared as between individuals or group of individuals who have said trait and who do not have said trait. In yet another embodiment, the gene expression pattern or profile of the test individual is compared with the individuals used for the training population. In yet another embodiment, the gene expression pattern or profile of the test individual is compared with the individuals of the scoring population. As used herein, a "gene expression pattern" or "gene expression profile" indicates the combined pattern of the results of the analysis of the level of expression of two or more biomarkers of the invention including 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or all of the biomarkers of the classifier or classifier groups. A gene expression pattern or gene expression profile can result from the measurement of expression of the products of the biomarkers of the invention and can be done by measuring either the RNA or the proteins corresponding to said molecular marker using any any of the techniques described herein. For example techniques to measure expression of the RNA products of the biomarkers of the invention includes, PCR based methods (including RT-PCR and quantitative RT-PCR) and non PCR based method' as well as microarray analysis. To measure protein products of the biomarkers of the invention, techniques include western blotting and ELISA analysis.
5.6 KITS
One embodiment of the present invention comprises kits for measuring molecular marker data by providing the materials necessary to measure the abundance of one or more of the products of one or more molecular markers of the classifier or classifier groups identified. Such kits may comprise materials and reagents required for measuring molecular marker data where the product of the molecular marker is RNA or protein. In some embodiments, such kits include microarrays wherein the microarray is comprised of oligonucleotides and/or DNA and/or RNA fragments which hybridize to one or more of the products of one or more of the molecular markers of a classifier or classifier group. In some embodiments, such kits may include primers fox PCR of either the RNA
product or the cDNA copy of the RNA product of the molecular marker or both. In some embodiments, such kits may include primers for PCR as well as probes for Quantitative PCR. In some embodiments, such kits may include multiple primers and multiple probes wherein some of said probes have different flourophores so as to permit multiplexing of multiple products of a single molecular marker or multiple products wherein each product results from a single molecular marker. In some embodiments, such kits may further include materials and reagents for creating cDNA from RNA. In some embodiments, such kits may include antibodies specific for the protein products of a molecular marker. Such kits may additionally comprise materials and reagents fox isolating RNA and/or proteins from a blood sample. Such kits may additionally comprise materials and reagents for isolating RNA and/or proteins from a non-blood tissue sample. In addition such kits may include materials and reagents for synthesizing cDNA from RNA isolated from a blood sample. In some embodiments of the present invention such kits may include, a computer program product embedded on computer readable media for determining whether a subject has a trait of interest. In some embodiments of the present invention, the kits of the invention may include a computer program product embedded on a computer readable media along with instructions.
In some embodiments, the invention provides kits for measuring the expression of one or more nucleic acid sequences of one or more molecular markers. In a specific embodiment, such kits measure the expression of one or more nucleic acid sequences associated with a molecular marker which has been determined according to the method of the invention as being indicative of a trait of interest. In accordance with this embodiment, the kits may comprise materials and reagents that are necessary for measuring the expression of particular nucleic acid sequence products of molecular markers identified by a classifier or classifier group of the invention. For example, a microarray or RT-PCR kit may be produced for a specific condition and contain only those reagents and materials necessary for measuring the levels of specific RNA
transcript products of the molecular markers associated with the classifier or classifier groups selected in accordance with one embodiment of the invention. Alternatively, in some embodiments, the kits can comprise materials and reagents that are not limited to those required to measure the expression of particular nucleic acid sequences of any particular molecular marker. For example, in certain embodiments, the kits comprise materials and reagents necessary for measuring the levels of expression of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50 or more of the molecular markers of the invention, in addition to reagents and materials necessary for measuring the levels of the expression of at least l, at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 45, at least 50 or more genes other than the molecular markers of the invention. In other embodiments, the kits contain reagents and materials necessary for measuring the levels of expression of at least 1, at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 45, at least 50 or more of the molecular markers of the invention, and 1, 2, 3, 4, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 125, 150, 175, 200, 225, 250, 300, 350, 400, 450, or more genes that are not molecular markers of the invention, or 1-10, 1-100, 1-150, 1-200, 1-300, 1-400, 1-500, 1-1000, 2S-100, 2S-200, 2S-300, 2S-400, 25-500, 2S-1000, 100-150, 100-200, 100-300, 100-400, 100-500, 100-1000 or 500-1000 genes that are not molecular markers of the invention.
S For nucleic acid micoarray kits, the kits generally comprise probes attached to a solid support surface. In one such embodiment, probes can be either oligonucleotides or longer length probes including probes ranging from 1 SO nucleotides in length to 800 nucleotides in length. The probes may be labeled with a detectable label. In a specific embodiment, the probes are specific for one or more of the products of a specific molecular marker identified following the methods of section 5.1. The microarray kits may comprise instructions for performing the assay and methods for interpreting and analyzing the data resulting from the performance of the assay. In a specific embodiment, the kits comprise instructions for diagnosing a trait of interest. The kits may also comprise hybridization reagents and/or reagents necessary for detecting a signal produced when a 1 S probe hybridizes to a target nucleic acid sequence. Generally, the materials and reagents for the microarray kits are in one or more containers. Each component of the kit is generally in its own a suitable container.
In certain embodiments, a nucleic acid microarray kit comprises materials and reagents necessary for measuring the levels of expression of 1, 2, 3, 4, S, 6, 7, 8, 9, 10, 1S, 20, 2S, 30, 3S, 40, 4S, SO or more of the molecular markers of the invention, in addition to reagents and materials necessary for measuring the levels of the expression of at least 1, at least 2, at least 3, at least 4, at least S, at least 6, at least 7, at least 8, at least 9, at least 10, at least 1S, at least 20, at least 2S, at least 30, at least 3S, at least 40, at least 4S, at least SO
or more genes other than the molecular markers of the invention. In other embodiments, a 2S nucleic acid microarray kit contains reagents and materials necessary for measuring the levels of expression of at least 1, at least 2, at least 3, at least 4, at least S, at least 6, at least 7, at least 8, at least 9, at least 10, at least 1S, at least 20, at least 2S, at least 30, at least 3S, at least 40, at least 4S, at least SO or more of the molecular markers of the invention, and 1, 2, 3, 4, S, 10, 1S, 20, 2S, 30, 35, 40, 45, 50, SS, 60, 6S, 70, 75, 80, 8S, 90, 9S, 100, 125, 150, 175, 200, 225, 250, 300, 350, 400, 450, or more genes that are not molecular markers of the invention, or 1-10, 1-100, 1-150, 1-200, 1-300, 1-400, 1-500, I-1000, 2S-100, 25-200, 2S-300, 2S-400, 2S-500, 2S-1000, 100-150, 100-200, 100-300, 100-400, 100-500, 100-1000 or 500-1000 genes that are not molecular markers of the invention.

For Quantitative PCR, the kits generally comprise pre-selected primers specific for particular nucleic acid sequences. The Quantitative PCR kits may also comprise enzymes suitable for amplifying nucleic acids (e.g., polymerases such as Taq), and deoxynucleotides and buffers needed for the reaction mixture for amplification. The Quantitative PCR kits may also comprise probes specific for the nucleic acid sequences associated with or indicative of a condition. The probes may or may not be labeled with a flourophore. The probes may or may not be labeled with a quencher molecule. In some embodiments the Quantitative PCR kits also comprise components suitable for reverse-transcribing RNA including enzymes (e.g. reverse transcriptases such as AMV, MMLV
and the like) and primers for reverse transcription along with deoxynucleotides and buffers needed for the reverse transcription reaction. Each component of the quantitative PCR kit is generally in its own suitable container. Thus, these kits generally comprise distinct containers suitable for each individual reagent, enzyme, primer and probe.
Further, the quantitative PCR kits may comprise instructions for performing the assay and methods for 1 S interpreting and analyzing the data resulting from the performance of the assay. In a specific embodiment, the kits contain instructions for diagnosing a trait of interest.
For antibody based kits, the kit can comprise, fox example: (1) a first antibody (which may or may not be attached to a solid support) which binds to a peptide, polypeptide or protein of interest; and, optionally, (2) a second, different antibody which binds to either the peptide, polypeptide or protein, or the first antibody and is conjugated to a detectable label (e.g., a fluorescent label, radioactive isotope or enzyme).
In a specific embodiment, the peptide, polypeptide or protein of interest is associated with or indicative of a condition (e.g., a disease). The antibody-based kits may also comprise beads for conducting an immunoprecipitation. Each component of the antibody-based kits is generally in its own suitable container. Thus, these kits generally comprise distinct containers suitable for each antibody. Further, the antibody-based kits may comprise instructions for performing the assay and methods for interpreting and analyzing the data resulting from the performance of the assay. In a specific embodiment, the kits contain instructions for diagnosing a trait of interest.
5.7 EXEMPLARY NORMALIZATION ROUTINES
A number of different normalization protocols can be used to normalize molecular marker data obtained using microarrays. Some such normalization protocols are described in this section. Typically, the normalization comprises normalizing the expression level measurement of each gene in a plurality of genes that is expressed by a subject. Many of the normalization protocols described in this section are used to normalize microarray data.
It will be appreciated that there are many other suitable normalization protocols that may be used in accordance with the present invention. All such protocols are within the scope of the present invention. Many of the normalization protocols found in this section are found in publicly available software, such as Microarray Explorer (Image Processing Section, Laboratory of Experimental and Computational Biology, National Cancer Institute, Frederick, MD 21702, USA).
One normalization protocol is Z-score of intensity. In this protocol, raw expression intensities are normalized by the (mean intensity)l(standard deviation) of raw intensities for all spots in a sample. For microarray data, the Z-score of intensity method normalizes each hybridized sample by the mean and standard deviation of the raw intensities for all of the spots in that sample. The mean intensity mnI; and the standard deviation sdI; are computed for the raw intensity of control genes. It is useful for standardizing the mean (to 0.0) and the range of data between hybridized samples to about -3.0 to +3Ø
When using the Z-score, the Z differences (Z d;~) are computed rather than ratios. The Z-score intensity (Z-score;) for intensity I;~ for probe i (hybridization probe, protein, or other binding entity) and spot j is computed as:
Z-score; _ (I;~ - mnI;) / sdI;, and Zdiff (x,y) = Z-scoreX~ - Z-scorey~
where x represents the x channel and y represents the y channel.
Another normalization protocol is the median intensity normalization protocol in which the raw intensities for all spots in each sample are normalized by the median of the raw intensities. For microarray data, the median intensity normalization method normalizes each hybridized sample by the median of the raw intensities of control genes (medianI;) for all of the spots in that sample. Thus, upon normalization by the median intensity normalization method, the raw intensity I;~ for probe i and spot j, has the value Im;~ where, Im;~ _ (I;~/ medianI;).
Another normalization protocol is the log median intensity protocol. In this protocol, raw expression intensities are normalized by the log of the median scaled raw intensities of representative spots for all spots in the sample. For microarray data, the log median intensity method normalizes each hybridized sample by the log of median scaled raw intensities of control genes (medianI;) for all of the spots in that sample. As used herein, control genes are a set of genes that have reproducible accurately measured expression values. The value 1.0 is added to the intensity value to avoid taking the log(0.0) when intensity has zero value. Upon normalization by the median intensity normalization method, the raw intensity I;~ for probe i and spot j, has the value Im;~ where, Im;~ = log(1.0 + (I;a/ medianI;)).
Yet another normalization protocol is the Z-score standard deviation log of intensity protocol. In this protocol, raw expression intensities are normalized by the mean log intensity (mnLI;) and standard deviation log intensity (sdLI;). For microarray data, the mean log intensity and the standard deviation log intensity is computed for the log of raw intensity of control genes. Then, the Z-score intensity ZlogS;~ for probe i and spot j is:
ZlogS;~ _ (log(I;~) - mnLI;)/sdLI;.
Still another normalization protocol is the Z-score mean absolute deviation of log intensity protocol. In this protocol, raw expression intensities are normalized by the Z-score of the log intensity using the equation (log(intensity)-mean logarithm) / standard deviation logarithm. For microarray data, the Z-score mean absolute deviation of log intensity protocol normalizes each bound sample by the mean and mean absolute deviation of the logs of the raw intensities for all of the spots in the sample. The mean log intensity mnLl; and the mean absolute deviation log intensity madLI; are computed for the log of raw intensity of control genes. Then, the Z-score intensity ZlogA;~ for probe i and spot j is:
ZlogA;~ _ (log(I;~) - mnLI;)/madLI;.
Another normalization protocol is the user normalization gene set protocol. In this protocol, raw expression intensities are normalized by the sum of the genes in a user defined gene set in each sample. This method is useful if a subset of genes has been determined to have relatively constant expression across a set of samples. Yet another normalization protocol is the calibration DNA gene set protocol in which each sample is normalized by the sum of calibration DNA genes. As used herein, calibration DNA genes are genes that produce reproducible expression values that are accurately measured. Such genes tend to have the same expression values on each of several different microarrays.
The algorithm is the same as user normalization gene set protocol described above, but the set is predefined as the genes flagged as calibration DNA.

Yet another normalization protocol is the ratio median intensity correction protocol. This protocol is useful in embodiments in which a two-color fluorescence labeling and detection scheme is used. In the case where the two fluors in a two-color fluorescence labeling and detection scheme are Cy3 and CyS, measurements are normalized by multiplying the ratio (Cy3/Cy5) by medianCyS/medianCy3 intensities. If background correction is enabled, measurements are normalized by multiplying the ratio (C.y3/Gy5) by (medianCyS-medianBkgdCyS) / (medianCy3-medianBkgdCy3) where medianBkgd means median background levels.
In some embodiments, intensity background correction is used to normalize measurements. The background intensity data from a spot quantification programs may be used to correct spot intensity. Background may be specified as either a global value or on a per-spot basis. If the array images have low background, then intensity background correction may not be necessary.
5.8 EXEMPLARY DISEASES
As discussed supra, the present invention provides methods for developing classifiers that can be used to determine whether a patient has a certain trait including a disease. Exemplary diseases that can be identified include asthma, cancers, common late-onset Alzheimer's disease, diabetes, heart disease, hereditary early-onset Alzheimer's disease (George-Hyslop et al., 1990, Nature 347: 194), hereditary nonpolyposis colon cancer, hypertension, infection, maturity-onset diabetes of the young (Barbosa et al., 1976, Diabete Metab. 2: 160), mellihis, nonalcoholic fatty liver (NAFL) (Younossi, et al., 2002, Hepatology 35, 746-752), nonalcoholic steatohepatitis (HASH) (James & Day, 1998, J.
Hepatol. 29: 495-501 ), non-insulin-dependent diabetes mellitus, andpolycystic kidney disease (Reeders et al., 1987, Humav~ Ge~zetics 76: 348).
Disease also includes, blood disorder, blood lipid disease, autoimmune disease, arthritis (including osteoarthritis, rheumatoid arthritis, lupus, allergies, juvenile rheumatoid arthritis and the like), bone or joint disorder, a cardiovascular disorder (including heart failure, congenital heart disease; rheumatic fever, valvular heart disease;
cor pulmonale, cardiomyopathy, myocaxditis, pericardial disease; vascular diseases such as atherosclerosis, acute myocardial infarction, ischemic heart disease and the like), obesity, respiratory disease (including asthma, pneumonitis, pneumonia, pulmonary infections, lung disease, bronchiectasis, tuburculosis, cystic fibrosis, interstitial lung disease, chronic bronchitis emphysema, pulmonary hypertension, pulmonary thromboembolism, acute respiratory distress syndrome and the like), hyperlipidemias, endocrine disorder, immune disorder, infectious disease, muscle wasting and whole body wasting disorder, neurological disorders (including migraines, seizures, epilepsy, cerebrovascular diseases, alzheimers, dementia, parkinsons, ataxic disorders, motor neuron diseases, cranial nerve disorders, spinal cord disorders, meningitis and the like) including neurodegenerative and/or neuropsychiatric diseases and mood disorders (including schizophrenia, anxiety, bipolar disorder; manic depression and the like, skin disorder, kidney disease, scleroderma, stroke, hereditary hemorrhage telangiectasia, diabetes, disorders associated with diabetes (e.g., PVD), hypertension, Gaucher's disease, cystic fibrosis, sickle cell anemia, liver disease, pancreatic disease, eye, ear, nose and/or throat disease, diseases affecting the reproductive organs, gastrointestinal diseases (including diseases of the colon, diseases of the spleen, appendix, gall bladder, and others) and the like. For further discussion of human diseases, see Mendelian Inheritance in Man: A Catalog of Human Genes and Genetic Disorders by Victor A. McI~usick (12th Edition (3 volume set) June 1998, Johns Hopkins University Press, ISBN: 0801857422) and Harrison's Principles of Internal Medicine by Braunwald, Fauci, I~asper, Hauser, Longo, & Jameson (15th Edition 200I), the entirety of which is incorporated herein.
Cancers that can be identified using the inventive techniques of the present invention include, Gut are not limited to, human sarcomas and carcinomas, e.g., fibrosarcoma, myxosarcoma, liposarcoma, chondrosaxcoma, osteogenic sarcoma, chordoma, angiosaxcoma, endotheliosarcoma, Iymphangiosarcoma, lymphangioendotheliosarcoma, synovioma, mesothelioma, Ewing's tumor, Ieiomyosarcoma, rhabdomyosarcoma, colon carcinoma, pancreatic cancer, breast cancer, ovarian cancer, prostate cancer, squamous cell carcinoma, basal cell carcinoma, adenocarcinoma, sweat gland carcinoma, sebaceous gland carcinoma, papillary carcinoma, papillary adenocarcinomas, cystadenocarcinoma, medullary caxcinorna, bronchogenic carcinoma, renal cell carcinoma, hepatoma, bile duct carcinoma, choriocarcinoma, sem.inoma, embryonal carcinoma, Wilms' tumor, cervical cancer, testicular tumor, lung carcinoma, small cell lung carcinoma, bladder carcinoma, epithelial carcinoma, glioma, astrocytoma, rnedulloblastoma, craniopharyngioma, ependymoma, pinealoma, hemangioblastoma, acoustic neuroma, oligodendroglioma, meningioma, melanoma, neuroblastoma, retinoblastoma; leukemias, e.g., acute lymphocytic leukemia and acute myelocytic leukemia (myeloblastic, promyelocytic, myelomonocytic, monocytic and erythroleukemia); chronic leukemia (chronic myelocytic (granulocytic) leukemia and chronic lymphocytic leukemia); and polycythemia vera, lymphoma (Hodgkin's disease and non-Hodgkin's disease), multiple myeloma, Waldenstrom's macroglobulinemia, and heavy chain disease.
5.9 EXEMPLARY DATABASE ARCHITECTURES
In some embodiments, training population 44, candidate molecular marker data structure 5~, patient database 6~, and/or classifier database 70 comprise or are stored in one or more data warehouses. Data warehouses are typically structured as either relational databases or multidimensional data cubes. This section describes relational databases and multidimensional data cube architectures that can be used to store training data, candidate molecular marker lists, patient molecular marker data and/or classifiers of the present invention. More information on relational databases and multidimensional data cubes is found in Berson and Smith, 1997, Data Warehousing, Data Minihg and OLAP, McGraw-Hill, New York; Freeze, 2000, Unlocking OLAP with Microsoft SQL Serves ahd Excel 2000, IDG Books Worldwide, Inc., Foster City, California; and Thomson, 1997, OLAP
Solutions: Building Multidimensional Information Systems, Wiley Computer Publishing, New York.
5.9.1 DATA ORGANIZATION
Databases have typically been used for operational purposes, such as order entry, accounting and inventory control. More recently, corporations and scientific projects have been building databases, called data warehouses or large on-Iine analytical processing (OLAP) databases, explicitly for the purposes of exploration and analysis. The "data warehouse" can be described as a subject-oriented, integrated, time-variant, nonvolatile collection of data in support of management decisions. Data warehouses are built using both relational databases and specialized multidimensional structures called data cubes. In some embodiments a database stored in computer 10 or stored in a computer addressable by computer 10 across wide area network 34 is a relational database or a datacube.
5.9.2 RELATIONAL DATABASES
Relational databases organize data into tables where each row corresponds to a basic entity or fact and each column represents a property of that entity. For example, a table can represent transactions in a bank, where each row corresponds to a single transaction, and each transaction has multiple attributes, such as the transaction amount, the account balance, the bank branch, and the customer. The relational table is referred to as a relation, a row as a tuple, and a column as an attribute or field. The attributes within a relation can be partitioned into two types: dimensions and measures.
Dimensions and measures are similar to independent and dependent variables in traditional analysis. For example, the bank branch and the customer would be dimensions, while the account balance would be a measure. A single relational database will often describe many heterogeneous but interrelated entities. For example, a database designed for a restaurant chain might maintain information about employees, products, and sales. The database schema defines the relations in a database, the relationships between those relations, and how the relations classify the entities of interest.
5.9.3 DATA CUBES
A data warehouse can be constructed as a relational database using either a star or snowflake schema and will provide a conceptual classifier of a multidimensional data set.
Each axis in the corresponding data cube represents a dimension in a relational schema and consists of every possible value for that dimension. For example, an axis corresponding to states would have fifty values, one for each state. Each cell in the data cube corresponds to a unique combination of values for the dimensions. For instance, if there are two imensions, "state" and "product", then there would be a cell for every unique combination of the two, e.g., one cell each for (California, Tea), (California, Coffee), (Florida, Tea), (Florida, Coffee), etc. Each cell contains one value per measure of the data cube. So if product production and consumption information is needed, then each cell would contain two values, one for the number of products of each type consumed in that state, and one for the number of products of each type produced in that state. Dimensions within a data warehouse are often augmented with a hierarchical structure. If each dimension has a hierarchical structure, then the data warehouse is not a single data cube but rather a lattice of data cubes.
5.10 EXEMPLARY PATIENT DATABASE
This section provides a more detailed description of a patient database 68 in accordance with one aspect of the invention. As described in Section 5. l, an exemplary patient database 68 includes a plurality of patient records 500 (Fig. 6).
There is no limit on the number of patient records 500 that can be held in patient database 68.
Database 68 can hold as few as one patient record 500. More typically, database 68 holds between 1 and 100 patient records, more than I00 patient records, more than a thousand patient records, more than ten thousand patient records, more than 100 thousand patient records, between 1 patient record and one million patient records, or more. Each patient record 500 preferably includes a patient identifier 502. As those skilled in database arts will appreciate, a patient identifier 502 need not be explicitly enumerated in certain database systems.
For instance, in some systems, a patient identifier 502 can simply be a patient record 500 identifier.
However, in some embodiments, a patient identifier 502 can be a number that uniquely identifies a patient within a health care program or clinical trial.
An advantage of database 68 is that it has the capability of tracking molecular marker data profile 504 and trait characterization information 510 for each patient registered in database 68. In some embodiments, a molecular profile 504 is the abundance levels of a plurality of molecular marker products in blood specimens obtained from a patient in accordance with Section 5.2. In some embodiments, such abundance levels are normalized using any of the techniques disclosed in Section 5.7.
In some embodiments, a molecular profile 504 comprises the processed microarray image data from the biological specimen obtained from the patient. In one example, molecular profile data 504 comprises molecular marker abundance information for all or a portion of the cellular constituents represented in a microarray, optional background signal information, and optional associated annotation information describing the probes used for the respective molecular marker. Molecular markers include, but are not limited to RNA
(e.g., mRNA) and protein.
In some embodiments, a molecular profile 504 represents the transcriptional state of cellular constituents in a biological specimen. However, in other embodiments, a molecular profile 504 can track aspects of the biological state other than or in addition to transcriptional state. Such other aspects of the biological state include, but are not limited to, the translational state, the activity state of cellular constituents in a biological sample.
In some embodiments, for example, molecular profile 504 data is, in fact, protein levels for various proteins in the blood taken from the patient. Thus, in some embodiments, molecular profiles 504 comprise amounts or concentrations of the molecular markers in biological specimens obtained in accordance with Section 5.2.
In one embodiment, the amount of at least one molecular marker that is tracked in a molecular profile 504 comprises abundances of at least one RNA species present in one or more cells in the blood obtained from the patient. Such abundances can be measured by a method comprising contacting a gene transcript array with RNA derived from one or more cells of the biological specimen, or with cDNA derived therefrom. A gene transcript array comprises a surface with attached nucleic acids or nucleic acid mimics. The nucleic acids or nucleic acid mimics are capable of hybridizing with the RNA species or with cDNA
derived from the RNA species. In one particular embodiment, the abundance of the RNA

is measured by contacting a gene transcript array with the RNA from one or more cells of the biological specimen, or with nucleic acid derived from the RNA, such that the gene transcript array comprises a positionally addressable surface with attached nucleic acids or nucleic acid mimics, where the nucleic acids or nucleic acid mimics are capable of hybridizing with the RNA species, or with nucleic acid derived from the RNA
species.
In some embodiments, a molecular profile 504 can include abundance information or activity information about ten or more molecular markers (e.g., genes or proteins), between ten and one thousand molecular markers, between one thousand and twenty thousand molecular markers, or more than twenty thousand molecular markers.
In some embodiments, in addition to or rather than providing abundance information or activity information for molecular markers, a molecular profile 504 tracks polymorphism information. Such polymorphism information includes, but is not limited to, single nucleotide polymorphisms (SNPs), SNP haplotypes, microsatellite markers, restriction fragment length polymorphisms (RFLPs), short tandem repeats, sequence length polymorphisms, DNA methylation, random amplified polymorphic DNA (RAPD), amplified fragment length polymorphisms (AFLP), and "simple sequence repeats."
For more information on such polymorphisms, see generally, The DNA Revolz~tioa by Andrew H. Paterson 1996 (Chapter 2) in: Ge~ome Nlappifzg in Plants (ed. Andrew H.
Paterson) by Academic Press/R. G. Landis Company, Austin, Tex., 7-21, which is hereby incorporated herein by reference in its entirety SNPs occur approximately once every 600 base pairs in the genome. See, fox example, Kruglyak and Nickerson, 2001, Nature Genetics 27, 235. Alleles making up blocks of such SNPs in close physical proximity are often correlated, resulting in reduced genetic variability and defining a limited number of "SNP haplotypes" each of which reflects descent from a single ancient ancestral chromosome. See Fullerton et al., 2000, Arn. J. Hum. Genet. 67, 881. Such haplotype structure is used in some embodiments of the present invention. Patil et al. found that a very dense set of SNPs is required to capture all the common haplotype information. See Patil et al., 2001, Science 294, 1719-1723.
DNA methylation is described in Grunau et al., 2003, Nucleic Acids Res. 31, pp. 75-77.
RFLPs are the product of allelic differences between DNA restriction fragments caused by nucleotide sequence variability. As is well known to those of skill in the art, RFLPs are typically detected by extraction of genomic DNA and digestion with a restriction endonuclease. Generally, the resulting fragments are separated according to size and hybridized with a probe; single copy probes are preferred. As a result, restriction fragments from homologous chromosomes are revealed. Differences in fragment size among alleles represent an RFLP (see, for example, Helentjaris et al., 1985, Plant Mol.
Bio. 5:109-118, and U.S. Pat. No. 5,324,631).
The phrase "random amplified polymorphic DNA" or "RAPD" refers to the amplification product of the distance between DNA sequences homologous to a single oligonucleotide primer appearing on different sites on opposite strands of DNA.
Mutations or rearrangements at or between binding sites will result in polymorphisms as detected by the presence or absence of amplification product (see, for example, Welsh and McClelland, 1990, Nucleic Acids Res. 18:7213-7218; Hu and Quiros, 1991, Plant Cell Rep. 10:505-511). AFLP technology refers to a process that is designed to generate large numbers of randomly distributed molecular markers (see, for example, European Patent Application No. 053485$ Al).
"Simple sequence repeats" or "SSRs" axe di-, tri- or tetra-nucleotide tandem repeats within a genome. The repeat region can vary in length between genotypes while the DNA flanking the repeat is conserved such that the same primers will work in a plurality of genotypes. A polymorphism between two genotypes represents repeats of different lengths between the two flanking conserved DNA sequences (see, for example, Akagi et al., 1996, Theor. Appl. Genet. 93, 1071-1077; Bligh et al., 1995, Euphytica 86:83-85; Struss et al., 1998, Theor. Appl. Genet. 97, 308-315; Wu et al., 1993, Mol. Gen.
Genet. 241, 225-235; and U.S. Pat. No. 5,075,217). SSR are also known as satellites or microsatellites.
In addition to molecular profiles 50, patient records 500 include trait characterizations 510. In some embodiments, a trait characterization 510 comprises observations made by a patient's physician. In some instances, the observations made by a physician include a code from the International Classification of Diseases, 9th Revision, prepared by the Department of Health and Human Services (ICD-9 codes), or an equivalent, and dates such observations were made.
5.11 EXEMPLARY GENES AS CANDIDATE MOLECULAR MARKERS
Non-limiting examples of genes useful as molecular markers for use in the invention can include, but are not limited to, genes specific for or involved in a particular biological process, such as apoptosis, differentiation, stress response, aging, proliferation, etc.; cellular mechanism genes, e.g., cell-cycle, signal transduction, metabolism of toxic compounds, and the like; disease associated genes, e.g., genes involved in cancer, schizophrenia, diabetes, high blood pressure, atherosclerosis, viral-host interaction and infection and the like. Exemplary genes can also include immune responsive genes.
Further examples of genes can include, but are not limited to, oncogenes whose expression within a cell induces that cell to become converted from a normal cell into a tumor cell.
See for example Hanahan & Weinberg, 2000, Cell 100:57; Yokota., 2000, Carcinogenesis S 21:497. Further examples of genes can include, but are not limited to cytokine genes. See, for example, Rubinstein et al., 1998, Cytokine Growth Factor Rev. 9:17S-81.
Other examples of genes can include idiotype protein genes (e.g., Benezra., et al., Oncogene 20:8334-41; Norton, 2000, J. Cell Sci. 113:3897), prion genes (e.g., Prusiner et al., 1998, Cell 93:337-48; Safar & Prusiner, 1998, Prog. Brain Res. 117:421);
genes that express molecules that induce angiogenesis (e.g., Gould & Wagner, 2002, Hum.
Pathol.
33:1061); genes encoding adhesion molecules (e.g., Chothia, 8~ Jones, 1997, Annu. Rev.
Biochem. 66:823; Parise et al., 2000, Semin. Cancer Biol. 10:407-14); genes encoding cell surface receptors (e.g., Deller and Jones, 2000, Curr. Opin. Struct. Biol.
10:213); genes of proteins that are involved in metastasizing and/or invasive processes (e.g., Boyd, 1996, 1 S C~cer Metastasis Rev. 15:77; Yokota, 2000, Carcinogenesis 21:497); genes of proteases as well as of molecules that regulate apoptosis and the cell cycle (e.g., Matrisian, 1999, Curr. Biol. 9:8776; Krepela, 2001, Neoplasma 48:332; Basbaum and Werb, 1996, Curr.
Opin. Cell Biol. 8:731; Birkedal-Hansen et al., 1993, Crit. Rev. Oral Biol.
Med. 4:197-250; Mignatti and Rifkin, 1993, Physiol. Rev. 73:161; Stetler-Stevenson et al., 1993, ~u. Rev. Cell Biol. 9:541; Brinkerhoff and Matrisan, 2002, Nature Reviews 3:207;
Strasser. et al., 2000, Annu. Rev. Biochem. 69:217; Chao and Korsmeyer, 1998, Annu.
Rev. Immunol. 16:395; Mullauer et al., 2001, Mutat. Res. 488:211; Fotedar et al., 1996, Prog. Cell Cycle Res. 2:147; Reed., 2000, Am. J. Pathol. 157:1415; D'.Ari, 2001, Bioassays 23:563); or multi-drug resistance genes, such as the MDR1 gene. In one embodiment, a 2S gene is an immune response gene or a non-immune response gene such as cytokines (e.g., interleukins and interferons such as TNF-alpha, IL-10, IL-12, IL-2, IL-4, IL-10, IL-12, IL-13, TGF-Beta, IFN-gamma; immunoglobulins, complement and the like). See, for example, Bellardelli, 1995, Role of ihte~fet~ohs and other cytokines in the regulation of the immune response APMIS 103: 161.
5.12 CLUSTERING TECHNIQUES
In some embodiments, clustering is used. For instance, clustering can be used in step 204 to visualize the relationship between the data measured for a plurality of molecular markers in step 202. In some embodiments, any of the clustering techniques described in Draghici, Data Analysis Tools For DNA Mic~oar~ays, 2003, Chapman &

Hall, CRC Press, New York, pp. 263-297, which is hereby incorporated by reference in its entirety, are used in the present invnetion. Clustering is also described on pages 211-256 of Duda and Hart, Pattern Classification a~zd Scene Analysis, 1973, John Wiley & Sons, Inc., New York, which is hereby incorporated by reference. As described in Section 6.7 of Duda, the clustering problem is described as one of finding natural groupings in a dataset.
To identify natural groupings, two issues are addressed. First, a way to measure similarity (or dissimilarity) between two samples is determined. This metric (similarity measure) is used to ensure that the samples in one cluster are more like one another than they are to samples in other clusters. Second, a mechanism for partitioning the data into clusters using the similarity measure is determined.
Similarity measures are discussed in Section 6.7 of Duda, where it is stated that one way to begin a clustering investigation is to define a distance function and to compute the matrix of distances between all pairs of samples in a dataset. If distance is a good measure of similarity, then the distance between samples in the same cluster will be significantly less than the distance between samples in different clusters. However, as stated on page 215 of Duda, clustering does not require the use of a distance metric. For example, a nonmetric similarity function s(x, x') can be used to compare two vectors x and x'.
Conventionally, s(x, x') is a symmetric function whose value is large when x and x' are somehow "similar". An example of a nonmetric similarity function s(x, x') is provided on page 216 of Duda.
Once a method for measuring "similarity" or "disimilarity" between points in a dataset has been selected, clustering requires a criterion function that measures the clustering quality of any partition of the data. Partitions of the data set that extremize the criterion function are used to cluster the data. See page 217 of Duda.
Criterion functions are discussed in Section 6.8 of Duda.
More recently, Duda et aL, Pattern Classification, 2°d edition, John Wiley & Sons, Inc. New York, has been published. Pages 537-563 describe clustering in detail. More information on clustering techniques can be found in Kaufman and Rousseeuw, 1990, Finding Groups in Data: An hztroductio~a to Cluster Analysis, Wiley, New York, NY;
Everitt, 1993, Cluster analysis (3d ed.), Wiley, New York, NY; and Backer, 1995, ComputeY-Assisted Reasoning in Cluster A~talysis, Prentice Hall, Upper Saddle River, New Jersey. Now that an overview of clustering techniques has been given, more specific examples of clustering that can be performed in the methods described in Section 5.1 is presented.

5.12.1 HIERARCHICAL CLUSTERING TECHNIQUES
Hierarchical cluster analysis is a statistical method for finding relatively homogenous clusters of elements based on measured data. Consider a sequence of partitions of n samples into c clusters. The first of these is a partition into n clusters, each S cluster containing exactly one sample. The next is a partition into n 1 clusters, the next is partition into n-2, and so on until the n~', in which all the samples form one cluster. Level k in the sequence of partitions occurs when c = n - k + 1. Thus, level one corresponds to n clusters and level n corresponds to one cluster. Given any two samples x and x*, at some level they will be grouped together in the same cluster. If the sequence has the property that whenever two samples are in the same cluster at level k they remain together at all higher levels, then the sequence is said to be a hierarchical clustering. Duda et al., 2001, Pattern Classification, 2°d edition, John Wiley & Sons, New York, 2001:
SS1. Examples of hierarchical clustering includes agglomerative clustering using nearest-neighbor algorithm, farthest-neighbor algorithm, the average linkage algorithm, the centroid 1S algorithm, or the sum-of squares algorithm. See, for example W003100SS7.
5.12.1.1 CLUSTERING WITH PEARSON CORRELATION
COEFFICIENTS
In some embodiments of the present invention, molecular marker data data is clustered using agglomerative hierarchical clustering with Pearson correlation coefficients.
In this form of clustering, similarity is determined using Pearson correlation coefficients between sets of molecular marker data measurements. Other metrics that can be used, in addition to the Pearson correlation coefficient, include but are not limited to, a Euclidean distance, a squared Euclidean distance, a Euclidean sum of squares, a Manhattan metric, and a squared Pearson correlation coefficient. Such metrics can be computed using SAS
2S (Statistics Analysis Systems Institute, Cary, North Carolina) or S-Plus (Statistical Sciences, Inc., Seattle, Washington).
5.12.1.2 DIVISIVE CLUSTERING
In some embodiments, the hierarchical clustering technique used to cluster molecular marker data measurements is a divisive clustering procedure.
Divisive (top-down clustering) procedures start with all of the samples in one cluster and form the sequence by successfully splitting clusters. Divisive clustering techniques are classified as either a polythetic or a monothetic method. A polythetic approach divides clusters into arbitrary subsets.

5.12.2 K-MEANS CLUSTERING
In k-means clustering, sets of molecular marker data measurements are randomly assigned to K user specified clusters. The centroid of each cluster is computed by averaging the value of the vectors in each cluster. Then, for each i= 1, ..., N, the distance between vector x; and each of the cluster centroids is computed. Each vector x; is then reassigned to the cluster with the closest centroid. Next, the centroid of each affected cluster is recalculated. The process iterates until no more reassignments are made. See, for example, Duda et al., 2001, Pattern Classification, John Wiley & Sons, New York, NY, pp. 526-528. A related approach is the fuzzy k-means clustering algorithm, which is also known as the fuzzy c-means algorithm. In the fuzzy k-means clustering algorithm, the assumption that every set of molecular marker data measurements is in exactly one cluster at any given time is relaxed so that every set has some graded or "fuzzy"
membership in a cluster. See Duda et al., 2001, Pattern Classification, John Wiley &
Sons, New York, NY, pp. 528-530.
5.12.3 JARVIS-PATRICK CLUSTERING
Jarvis-Patrick clustering is a nearest-neighbor non-hierarchical clustering method in which a set of objects is partitioned into clusters on the basis of the number of shared nearest-neighbors. In the standard implementation advocated by Jarvis and Patrick,1973, IEEE T~ahs. Comput., C-22:1025-1034, a preprocessing stage identifies the K
nearest-neighbors of each object in the dataset. In the subsequent clustering stage, two objects i and j join the same cluster if (i) i is one of the K nearest-neighbors of j, (ii) j is one of the K nearest-neighbors of i, and (iii) i and j have at least km;" of their K
nearest-neighbors in common, where K and km;n are user-defined parameters. The method has been widely applied to clustering chemical structures on the basis of fragment descriptors and has the advantage of being much less computationally demanding than hierarchical methods, and thus more suitable for large databases. Jarvis-Patrick clustering can be performed using the Jarvis-Patrick Clustering Package 3.0 (Barnard Chemical Information, Ltd., Sheffield, United Kingdom).
5.13 MOLECULAR MARKERS
Molecular marker is used herein to mean a gene or genetic element. All genes and genetic elements are considered molecular markers, but the invention teaches how to identify molecular markers useful for diagnosing a trait of interest.

5.14 REPRESENTATIVE MATHEMATICAL MODELS THAT CAN BE
USED TO BUILD CLASSIFIERS
This section describes various mathematical models that can be used to build classifier in accordance with the methods of the present invention.
5.14.1 REGRESSION CLASSIFIERS
In some embodiments, the classifier constructed in step 216 is a regression classifier, preferably a logistic regression classifier. Such a regression classifier includes a coefficient for each of the molecular markers selected in the last instance of step 214. In such embodiments, the coefficients for the regression classif er are computed using, for example, a maximum likelihood approach. In such a computation, the data measured for the molecular markers in step 206 (e.g., RT-PCR data) is used. In particular embodiments, molecular marker data from only two trait subgroups is used and the dependent variable is absence or presence of a particular trait in the subjects for which molecular marker data is available. As in the case of step 210, the two different trait subgroups can, for example, respectively represent a diseased and nondiseased state, a first diseased state (e.g. liver cancer) and a second phenotypically similar (e.g. hepatitis B) or unrelated diseased state (~.g., Alzheimer's disease), those subjects that are responsive to drug therapy and those subjects that are not responsive to drug therapy, or subjects that have been subjected to a perturbation (e.g., drug treatment) versus those subjects that have not been subjected to a perturbation.
In another specific embodiment, training population 44 consists of a plurality of trait subgroups (e.g., three or more trait subgroups, four or more specific trait subgroups, etc.). In this specific embodiment, a generalization of the logistic regression model that handles multicategory responses can be used in step 216 to develop a classifier that discriminates between the various trait subgroups found in the training population. For example, measured data for selected molecular markers can be applied to any of the mulitcategory logit models described in Agresti, A~ Int~~oduction to Categorical Data Analysis, 1996, John Wiley & Sons, Inc., New York, Chapter ~, which is hereby incorporated herein by reference in its entirety, in order to develop a classifier capable of discriminating between any of a plurality of trait subgroups represented in a training population.

5.14.2 NEURAL NETWORKS
The present invention is not limited to the use of logistic regression. In some embodiments, the data measured for the molecular markers in step 206 (e.g., RT-PCR
data) can be used to train a neural network.
In some embodiments, a neural network is derived in each successive instance of step 216 of Fig. 2A using the combination of molecular markers selected in the corresponding instance of step 214 of Fig. 2A. A neural network is a two-stage regression or classification classifier. A neural network has a layered structure that includes a layer of input units (and the bias) connected by a layer of weights to a layer of output units. For regression, the layer of output units typically includes just one output unit.
However, neural networks can handle multiple quantitative responses in a seamless fashion.
In multilayer neural networks, there are input units (input layer), hidden units (hidden layer), and output units (output layer). There is, furthermore, a single bias unit that is connected to each unit other than the input units. Neural networks are described in Duda et al., 2001, Pattern Classification, Second Edition, John Wiley & Sons, Inc., New York; and Hastie et al., 2001, The Elements of Statistical Leaning, Springer-Verlag, New York.
The basic approach to the use of neural networks is to start with an untrained network, present a training pattern to the input layer, and to pass signals through the net and determine the output at the output layer. These outputs are then compared to the target values; any difference corresponds to an error. This error or criterion function is some scalar function of the weights and is minimized when the network outputs match the desired outputs. Thus, the weights are adjusted to reduce this measure of error. For regression, this error can be sum-of squared errors. For classification, this error can be either squared error or cross-entropy (deviation). See, e.g., Hastie et al., 2001, The Elements of Statistical Leay~ning, Springer-Verlag, New York.
Three commonly used training protocols are stochastic, batch, and on-line. In stochastic training, patterns are chosen randomly from the training set and the network weights are updated for each pattern presentation. Multilayer nonlinear networks trained by gradient descent methods such as stochastic back-propagation perform a maximum-likelihood estimation of the weight values in the classifier defined by the network topology. In batch training, all patterns are presented to the network before learning takes place. Typically, in batch training, several passes are made through the training data. In online training, each pattern is presented once and only once to the net.

In some embodiments, consideration is given to starting values for weights. If the weights are near zero, then the operative part of the sigmoid commonly used in the hidden Iayer of a neural network (see, e.g., Hastie et al., 2001, The Elements of Statistical Learning, Springer-Verlag, New York) is roughly linear, and hence the neural network collapses into an approximately linear classifier. In some embodiments, starting values for weights are chosen to be random values near zero. Hence the classifier starts out nearly linear, and becomes nonlinear as the weights increase. Individual units localize to directions and introduce nonlinearities where needed. Use of exact zero weights leads to zero derivatives and perfect symmetry, and the algorithm never moves.
Alternatively, starting with large weights often leads to poor solutions.
Since the scaling of inputs determines the effective scaling of weights in the bottom layer, it can have a large effect on the quality of the final solution. Thus, in some embodiments, at the outset all expression values are standardized to have mean zero and a standard deviation of one. This ensures all inputs are treated equally in the regularization process, and allows one to choose a meaningful range for the random starting weights.
With standardization inputs, it is typical to take random uniform weights over the range [-0.7, +0.7].
A recurrent problem in the use of three-layer networks is the optimal number of hidden units to use in the network. The number of inputs and outputs of a three-layer network are determined by the problem to be solved. In embodiments of the present invention, the number of inputs for a given neural network can, in some embodiments, equal the number of molecular maxkers selected in the corresponding instance of step 214.
In other embodiments, for each input, two or more molecular markers will be selected (for example wherein ratios of genes (A/B) are utilized. The number of outputs for the neural network will typically be just one (ie wherein the ouput neuron is one dimensional e.g.
health vs. disease). If there are additional input dimensions, new additional output neurons may be created. In some embodiments more than one output is used so that more than just two states can be defined by the network. If too many hidden units are used in a neural network, the network will have too many degrees of freedom and is trained too long, there is a danger that the network will overfit the data. If there are too few hidden units, the training set cannot be learned. Generally speaking, however, it is better to have too many hidden units than too few. With too few hidden units, the classifier might not have enough flexibility to capture the nonlinearities in the data; with too many hidden units, the extra weight can be shrunk towards zero if appropriate regularization or pruning, as described below, is used. In typical embodiments, the number of hidden units in somewhere in the range of 5 to 100, with the number increasing with the number of inputs and number of training cases.
One general approach to determining the number of hidden units to use is to apply a regularization approach. In the regularization approach, a new criterion function is constructed that depends not only on the classical training error, but also on classifier complexity. Specifically, the new criterion function penalizes highly complex classifiers;
searching for the minimum in this criterion is to balance error on the training set with error on the training set plus a regularization term, which expresses constraints or desirable properties of solutions:
J = Jpat + 7v,J,.eg.
The parameter ~, is adjusted to impose the regularization more or less strongly. In other words, larger values for ~, will tend to shrink weights towards zero:
typically cross-validation with a validation set is used to estimate .~. This validation set can be obtained by setting aside a random subset of the population measured in step 202 of Fig.
2A. Other foi~rns of penalty have been proposed, for example the weight elimination penalty (see, e.g., Hastie et al., 2001, The Elements of Statistical Learning, Springer-Verlag, New York).
Another approach to determine the number of hidden units to use is to eliminate -prune - weights that are least needed. In one approach, the weights with the smallest magnitude are eliminated (set to zero). Such magnitude-based pruning can work, but is nonoptimal; sometimes weights with small magnitudes are important for learning and training data. In some embodiments, rather than using a magnitude-based pruning approach, Wald statistics are computed. The fundamental idea in Wald Statistics is that they can be used to estimate the importance of a hidden unit (weight) in a classifier. Then, hidden units having the least importance are eliminated (by setting their input and output weights to zero). Two algorithms in this regard are the Optimal Brain Damage (OBD) and the Optimal Brain Surgeon (OBS) algorithms that use second-order approximation to predict how the training error depends upon a weight, and eliminate the weight that leads to the smallest increase in training error.
Optimal Brain Damage and Optimal Brain Surgeon share the same basic approach of trainitng a network to local minimum error at weight w, and then pruning a weight that leads to the smallest increase in the training error. The predicted functional increase in the error for a change in full weight vector 8w is:

t 2 ~=C~~ .~+2~~.~z .~+o~,~~~3~
z where ~2 is the Hessian matrix. The first term vanishes because we are at a local minimum in error; third and higher order terms are ignored. The general solution for minimizing this function given the constraint of deleting one weight is:
z CSW-- H1N H_l.Zl9 L9~ 2_ W9 1l f l L ,99 ~d lH J99 Here, uq is the unit vector along the qth direction in weight space and Lq is approximation to the saliency of the weight q - the increase in training error if weight q is pruned and the other weights updated &w. These equations require the inverse of H. One method to calculate this inverse matrix is to start with a small value, Ho' =
a 1I, where a is a small parameter - effectively a weight constant. Next the matrix is updated with each pattern according to H-1 - H-1 _ Hm ~n:+l~m+1Hm m+I m + '~m+lHmlXm+1 am Eqn. 1 where the subscripts correspond to the pattern being presented and a", decreases with m. After the full training set has been presented, the inverse Hessian matrix is given by H-I = Hn I . In algorithmic form, the Optimal Brain Surgeon method is:
begin initialize nH, w, 8 train a reasonably large network to minimum error do compute H-1 by Eqn. 1 ~ E-- min w2 /(2~H-' ~ ) q ~'g 9 9 ~~ (saliency La) W.
~, ~- ~,- ~H~l ] . . H leg~ (saliency La) l qq until ,I(w) > 8 return w end The Optimal Brain Damage method is computationally simpler because the calculation of the inverse Hessian matrix in line 3 is particularly simple for a diagonal matrix. The above algorithm terminates when the error is greater than a criterion initialized to be 0. Another approach is to change line 6 to terminate when the change in J(w) due to elimination of a weight is greater than some criterion value.
In some embodiments, the back-propagation neural network (see, for example Abdi, 1994, "A neural network primer", J. Biol System. 2, 247-2~3) containing a single hidden layer of ten neurons (ten hidden units) found in EasyNN-Plus version 4.0g software package (Neural Planner Software Inc.) is used. In one specific example, parameter values within the EasyNN-Plus program were set as follows: learning parameter = 0.6, and momentum parameter = 0.~. In some embodiments in which the EasyNN-Plus version 4.0g software package is used, "outlier" samples are identified by performing twenty independently-seeded trials involving 20,000 learning cycles each.
5.14.3 CLUSTERTNG
In some embodiments, the expression values for select genes are used to.cluster a training set. For example, consider the case in which ten genes are used. Each member m of the training population will have expression values for each of the ten genes. Such values from a member m in the training population define the vector:
Xlm I X2m I X3m I X4m I XSm I X6m I X7m X8m X9m XlOm where X;m is the expression level of the ith gene in organism m. If there are m organisms in the training set, selection of i genes will define m vectors. Note that the methods of the present invention do not require that each expression value of every single gene used in the vectors be represented in every single vector m. In other words, data from a subject in which one of the ith genes is not found can still be used for clustering. In such instances, the missing expression value is assigned either a "zero" or some other normalized value.
In some embodiments, prior to clustering, the gene expression values are normalized to have a mean value of zero and unit variance.
Those members of the training population that exhibit similar expression patterns across the training group will tend to cluster together. A particular combination of genes of the present invention is considered to be a good classifier in this aspect of the invention when the vectors cluster into the trait groups found in the training population. For instance, if the training population includes patients with no osteoarthritis, mild osteoarthritis, moderate osteoarthritis, marked osteoarthritis, and severe osteoarthritis an ideal clustering classifier will cluster the population into five groups, with each group uniquely representing either absence or one of the four stages of osteoarthritis. In some embodiments, the clustering classifier simply clusters the population into a first subgroup (a first cluster) that does not have osteoarthritis and a second subgroup (a second cluster) that has osteoarthritis. In some embodiments, the classifier clusters the data into a first subgroup that has a particular stage of osteoarthritis (e.g., mild) and two or more subgroups that do not include subjects having the particular stage of osteoaxthritis represented in the first subgroup.
Clustering is described on pages 211-256 of Duda and Hart, Pattern Classification and Scene Analysis, 1973, John Wiley & Sons, Inc., New York. As described in Section 6.7 of Duda, the clustering problem is described as one of finding natural groupings in a dataset. To identify natural groupings, two issues are addressed. First, a way to measure similarity (or dissimilarity) between two samples is determined. This metric (similarity measure) is used to ensure that the samples in one cluster are more like one another than they are to samples in other clusters. Second, a mechanism for partitioning the data into clusters using the similarity measure is detemnined.
Similarity measures are discussed in Section 6.7 of Duda, where it is stated that one way to begin a clustering investigation is to define a distance function and to compute the matrix of distances between all pairs of samples in a dataset. If distance is a good measure of similarity, then the distance between samples in the same cluster will be significantly less than the distance between samples in different clusters. However, as stated on page 215 of Duda, clustering does not require the use of a distance metric. For example, a nonmetric similarity function s(x, x') can be used to compare two vectors x and x'.
Conventionally, s(x, x') is a symmetric function whose value is large when x and x' are somehow "similar". An example of a nonmetric similarity function s(x, x') is provided on page 216 of Duda.
Once a method for measuring "similarity" or "dissimilarity" between points in a dataset has been selected, clustering requires a criterion function that measures the clustering quality of any partition of the data. Partitions of the data set that extremize the criterion function are used to cluster the data. See page 217 of Duda.
Criterion functions are discussed in Section 6.~ of Duda.
More recently, Duda et al., Pattern Classification, 2nd edition, John Wiley 8i Sons, Inc. New York, has been published. Pages 537-563 describe clustering in detail. More information on clustering techniques can be found in Kaufman and Rousseeuw, 1990, Finding Groups in Data: An Introduction to Cluster Analysis, Wiley, New York, NY;
Everitt, 1993, Cluster analysis (3d ed.), Wiley, New York, NY; and Backer, 1995, Corraputer-Assisted Reasoning in Cluster Analysis, Prentice Hall, Upper Saddle River, New Jersey. Particular exemplary clustering techniques that can be used in the present invention include, but are not limited to, hierarchical clustering (agglomerative clustering using nearest-neighbor algorithm, farthest-neighbor algorithm, the average linkage algorithm, the centroid algorithm, or the sum-of squares algorithm), k-means clustering, fuzzy k-means clustering algorithm, and Jarvis-Patrick clustering.
5.14.4 PRINCIPAL COMPONENT ANALYSIS
Principal component analysis (PCA) has been proposed to analyze gene expression data. Principal component analysis is a classical technique to reduce the dimensionality of a data set by transforming the data to a new set of variable (principal components) that summarize the features of the data. See, for example, Jolliffe, 1986, Principal Component Analysis, Springer, New York. Principal components (PCs) are uncorrelate and are ordered such that the kth PC has the kth largest variance among PCs. The ktl' PC can be interpreted as the direction that maximizes the variation of the projections of the data points such that it is orthogonal to the first k - 1 PCs. The first few PCs capture most of the variation in the data set. In contrast, the last few PCs are often assumed to capture only the residual 'noise' in the data.
PCA can also be used to create a classifier in accordance with the present invention. In such an approach, vectors for the select genes described in the present invention can be constructed in the same manner described for clustering above. In fact, the set of vectors, where each vector represents the expression values for the select genes from a particular member of the training population, can be considered a matrix. In some embodiments, this matrix is represented in a Free-Wilson method of qualitative binary description of monomers (Kubinyi, 1990, 3D QSAR in drug design theory methods and applications, Pergamon Press, Oxford, pp 589-638), and distributed in a maximally compressed space using PCA so that the first principal component (PC) captures the largest amount of variance information possible, the second principal component (PC) captures the second largest amount of all variance information, and so forth until all variance information in the matrix has been accounted for.
Then, each of the vectors (where each vector represents a member of the training population) is plotted. Many different types of plots are possible. In some embodiments, a one-dimensional plot is made. In this one-dimensional plot, the value for the first principal component from each of the members of the training population is plotted. In this form of plot, the expectation is that members of a first subgroup (e. g. those subjects that do not have osteoarithritis) will cluster in one range of first principal component values and members of a second subgroup (e.g., those subjects that have osteoarthritis) will cluster in a second range of first principal component values.
In one ideal example, the training population comprises two subgroups:
"control"
and "patients with osteoarthritis." The first principal component is computed using the molecular marker expression values for the select genes of the present invention across the entire training population data set. Then, each member of the training set is plotted as a function of the value for the first principal component. In this ideal example, those members of the training population in which the first principal component is positive are the "responders" and those members of the training population in which the first principal component is negative are "patients with osteoarthritis."
In some embodiments, the members of the training population are plotted against more than one principal component. For example, in some embodiments, the members of the training population are plotted on a two-dimensional plot in which the first dimension is the first principal component and the second dimension is the second principal component. In such a two-dimensional plot, the expectation is that members of each subgroup represented in the training population will cluster into discrete groups. For example, a first cluster of members in the two-dimensional plot will represent subjects with mild osteoarthritis, a second cluster of members in the two-dimensional plot will represent subjects with moderate osteoarthritis, and so forth.
In some embodiments, the members of the training population are plotted against more than two principal components and a determination is made as to whether the members of the training population are clustering into groups that each uniquely represents a subgroup found in the training population. In some embodiments, principal component analysis is performed by using the R mva package (Anderson, 1973, Cluster Analysis for applications, Academic Press, New York 1973; Gordon, Classification, Second Edition, Chapman and Hall, CRC, 1999.). Principal component analysis is further described in Duda, Pattern Classification, Second Edition, 2001, John Wiley & Sons, Inc.
5.14.5 NEAREST NEIGHBOR CLASSIFIER ANALYSIS
Nearest neighbor classifiers are memory-based and require no classifier to be fit.
Given a query point xo, the k training points x~,.~, r, . . ., k closest in distance to xo are identified and then the point xo is classified using the k nearest neighbors.
Ties can be broken at random. In some embodiments, Euclidean distance in feature space is used to determine distance as:
dtl~ -~~xtl -x~~~
Typically, when the nearest neighbor algorithm is used, the expression data used to compute the linear discriminant is standardized to have mean zero and variance 1. In the present invention, the members of the training population are randomly divided into a training set and a test set. For example, in one embodiment, two thirds of the members of the training population are placed in the training set and one third of the members of the training population are placed in the test set. A select combination of genes described in the present invention represents the feature space into which members of the test set are plotted. Next, the ability of the training set to correctly characterize the members of the test set is computed. In some embodiments, nearest neighbor computation is performed several times for a given combination of genes of the present invention. In each iteration of the computation, the members of the training population are randomly assigned to the training set and the test set. Then, the quality of the combination of genes is taken as the average of each such iteration of the nearest neighbor computation.
The nearest neighbor rule can be refined to deal with issues of unequal class priors, differential misclassification costs, and feature selection. Many of these refinements involve some form of weighted voting for the neighbors. For more information on nearest neighbor analysis, see Duda, Pattern Classificatiov~, Second Edition, 2001, John Wiley ~
Sons, Inc; and Hastie, 2001, The Elements of Statistical Learning, Springer, New York.
5.14.6 LINEAR DISCRIMINANT ANALYSIS
Linear discriminant analysis (LDA) attempts to classify a subject into one of two categories based on certain object properties. In other words, LDA tests whether object attributes measured in an experiment predict categorization of the objects.
LDA typically requires continuous independent variables and a dichotomous categorical dependent variable. In the present invention, the expression values for the select combinations of genes described in the present invention across a subset of the training population serve as the requisite continuous independent variables. The trait subgroup classification of each of the members of the training population serves as the dichotomous categorical dependent variable.
13~

LDA seeks the linear combination of variables that maximizes the ratio of between-group variance and within-group variance by using the grouping information.
Implicitly, the linear weights used by LDA depend on how the expression of a molecular marker across the training set separates in the two groups (e.g., a group that has osteoarthritis and a group that does not have osteoarthritis) and how this gene expression correlates with the expression of other genes. In some embodiments, LDA is applied to the data matrix of the N members in the training sample by K genes in a combination of genes described in the present invention. Then, the linear discriminant of each member of the training population is plotted. Ideally, those members of the training population representing a first subgroup (e.g. those subjects that do not have osteoarthritis) will cluster into one range of linear discriminant values (e.g., negative) and those member of the training population representing a second subgroup (e.g. those subjects that have osteoarthritis) will cluster into a second range of linear discriminant values (e.g., positive).
The LDA is considered more successful when the separation between the clusters of discriminant values is larger. For more information on linear discriminant analysis, see Duda, Pattern Classification, Second Edition, 2001, John Wiley & Sons, Inc;
and Hastie, 2001, The Elements of Statistical Learning, Springer, New York; Venables &
Ripley, 1997, Modern Applied Statistics with s plus, Springer, New York.
5.14.7 QUADRATIC DISCRIMINANT ANALYSIS
Quadratic discriminant analysis (QDA) takes the same input parameters and returns the same results as LDA. QDA uses quadratic equations, rather than linear equations, to produce results. LDA and QDA are interchangeable, and which to use is a matter of preference and/or availability of software to support the analysis. Logistic regression takes the same input parameters and returns the same results as LDA and QDA.
5.14.8 SUPPORT VECTOR MACHINES
In some embodiments of the present invention, support vector machines (SVMs) are used to classify subjects using genesor genetic information. SVMs are a relatively new type of learning algorithm. See, for example, Cristianini and Shawe-Taylor, 2000, An Introduction to Support Vector Machines, Cambridge University Press, Cambridge, Boser et al., 1992, "A training algorithm for optimal margin classifiers, in Proceedings of the S'h Annual ACM Worlrshop on Computational Learning Theory, ACM Press, Pittsburgh, PA, pp. 142-152; Vapnik, 1998, Statistical Learning Theory, Wiley, New York. When used for classification, SVMs separate a given set of binary labeled training data with a hyper-plane that is [a maximal distance from each point using a fitting algorithm.
For cases in which no linear separation is possible, SVMs can work in combination with the technique of 'kernels', which automatically realizes a non-linear mapping to a feature space. The hyper-plane found by the SVM in feature space corresponds to a non-linear decision boundary in the input space.
In one approach, when a SVM is used, the gene expression data is standardized to have mean zero and unit variance and the members of a training population are randomly divided into a training set and a test set. For example, in one embodiment, two thirds of the members of the training population are placed in the training set and one third of the members of the training population are placed in the test set. The expression values for a combination of genes is used to train the SVM. Then the ability for the trained SVM to correctly classify members in the test set is determined. In some embodiments, this computation is performed several times for a given combination of molecular markers. In each iteration of the computation, the members of the training population are randomly assigned to the training set and the test set. Then, the quality of the combination of molecular markers is taken as the average of each such iteration of the SVM
computation.
For more information on SVMs, see Duda, Pattern Classification, Second Edition, 2001, John Wiley & Sons, Inc.; Hastie, 2001, The Elements of Statistical Learhihg, Springer, New York; and Furey et al., 2000, Bioinformatics 16, 906-914.
5.14.9 DECISION TREES
In some embodiments of the present invention, decision trees are used to classify subjects using expression data for combinations of genes. Decision tree algorithms belong to the class of supervised learning algorithms. The aim of a decision tree is to induce a classifier (a tree) from real-world example data. This tree can be used to classify unseen examples which have not been used to derive the decision tree.
A decision tree is derived from training data. An example contains values for the different attributes and what class the example belongs. In the present invention, the training data is expression data for a combination of genes across the training population.
The following algorithm describes a decision tree derivation:
Tree(Examples,Class,Attributes) Create a root node If all Examples have the same Class value, give the root this label Else if Attributes is empty label the root according to the most common value Else begin Calculate the information gain for each attribute Select the attribute A with highest information gain and make this the root attribute For each possible value, v, of this attribute Add a new branch below the root, corresponding to A = v Let Examples(v) be those examples with A = v If Examples(v) is empty, make the new branch a leaf node labeled with the most common value among Examples Else let the new branch be the tree created by Tree(Examples(v),Class,Attributes - {A}) end A more detailed description of the calculation of information gain is shown in the following. If the possible classes v; of the examples have probabilities P(v;) then the information content I of the actual answer is given by:
The I- value shows how much information we need in order to be able to describe the outcome of a classification for the specific dataset used. Supposing that the dataset contains p positive (e.g. has osteoarthritis) and n negative (e.g. healthy) examples (e.g.
individuals), the information contained in a correct answer is:
where loge is the logarithm using base two. By testing single attributes the amount of information needed to make a correct classification can be reduced. The remainder for a specific attribute A (e.g. a gene) shows how much the information that is needed can be reduced.
~ + xy. ~a R.~ TI ~~RIYP~~'7'(.t~) _ '~ ~ .I ( ' ;=I ~3 h- ~~ ~3x + i~p ~; + I~F
"v" is the number of unique attribute values for attribute A in a certain dataset, "i"
is a certain attribute value, "p;" is the number of examples for attribute A
where the classification is positive (e.g. cancer), "n;" is the number of examples for attribute A where the classification is negative (e.g. healthy).
The information gain of a specific attribute A is calculated as the difference between the information content for the classes and the remainder of attribute A:
+a~~~~~~~
The information gain is used to evaluate how important the different attributes are for the classification (how well they split up the examples), and the attribute with the highest information.
In general there are a number of different decision tree algorithms, many of which are described in Duda, Pattern Classification, Second Edition, 2001, John Wiley & Sons, Inc. Decision tree algorithms often require consideration of feature processing, impurity measure, stopping criterion, and pruning. Specific decision tree algorithms include, cut are not limited to classification and regression trees (CART), multivariate decision trees, ID3, and C4.5.
In one approach, when a decision tree is used, the gene expression data for a select combination of genes described in the present invention across a training population is standardized to have mean zero and unit variance. The members of the traitung population are randomly divided into a training set and a test set. For example, in one embodiment, two thirds of the members of the training population are placed in the training set and one third of the members of the training population are placed in the test set.
The expression values for a select combination of genes described in the present invention is used to construct the decision tree. Then, the ability for the decision tree to correctly classify members in the test set is determined. In some embodiments, this computation is performed several times for a given combination of molecular markers. In each iteration of the computation, the members of the training population are randomly assigned to the training set and the test set. Then, the quality of the combination of molecular markers is taken as the average of each such iteration of the decision tree computation.
5.14.10 EVOLUTIONARY METHODS
Inspired by the process of biological evolution, evolutionary methods of classifier design employ a stochastic search for an optimal classifier. In broad overview, such methods create several classifiers - a population - from a combination of genes described in the present invention. Each classifier varies somewhat from the other.
Next, the classifiers are scored on expression data across the training population. In keeping with the analogy with biological evolution, the resulting (scalar) score is sometimes called the f mess. The classifiers are ranked according to their score and the best classifiers are retained (some portion of the total population of classifiers). Again, in keeping with biological terminology, this is called survival of the fittest. The classifiers are stochastically altered in the next generation - the children or offspring.
Some offspring classifiers will have higher scores than their parent in the previous generation, some will have lower scores. The overall process is then repeated for the subsequent generation: The classifiers are scored and the best ones are retained, randomly altered to give yet another generation, and so on. In part, because of the ranking, each generation has, on average, a slightly higher score than the previous one. The process is halted when the single best classifier in a generation has a score that exceeds a desired criterion value.
More information on evolutionary methods is found in, for example, Duda, Pattern Classification, Second Edition, 2001, John Wiley & Sons, Inc.
5.14.11 BAGGING, BOOSTING AND THE RANDOM SUBSPACE
METHOD
Bagging, boosting and the random subspace method are combining techniques that can be used to improve weak classifiers. These techniques are designed for, and usually applied to, decision trees. In addition, Skurichina and Duin provide evidence to suggest that such techniques can also be useful in linear discriminant analysis.
In bagging, one samples the training set, generating random independent bootstrap replicates, constructs the classifier on each of these, and aggregates them by a simple majority vote in the final decision rule. See, for example, Breiman, 1996, Machine Learning 24, 123-140; and Efron & Tibshirani, An Introduction to Boostrap, Chapman &
Hall, New York, 1993.
In boosting, classifiers are constructed on weighted versions of the training set, which are dependent on previous classification results. Initially, all objects have equal weights, and the first classifier is constructed on this data set. Then, weights are changed according to the performance of the classifier. Erroneously classified objects (molecular markers in the data set) get larger weights, and the next classifier is boosted on the reweighted training set. In this way, a sequence of training sets and classifiers is obtained, which is then combined by simple majority voting or by weighted majority voting in the final decision. See, for example, Freund 8~ Schapire, "Experiments with a new boosting algorithm," Proceedings 13th International Conference on Machine Learning, 1996, 14~-156.
To illustrate boosting, consider the case where there are two trait extremes exhibited by the population under study, extreme phenotype 1 (e.g., severe osteoarthritis), and extreme phenotype 2 (e.g., no osteoarthritis). Given a vector of predictor molecular marker X selected in step 214, a classifier G(X) produces a prediction taking one of the type values in the two value set: {extreme phenotype 1, extreme phenotype 2}.
The error rate on the training sample is 1 "' err=-~1(Y~ ~G(xa)) N ;m where N is the number of subjects in the training set (the sum total of the subjects that have either extreme phenotype 1 or extreme phenotype 2). For example, if there are 49 organisms that have severe osteoarthritis and 72 organisms that have no osteoarthritis under study, N is 121.
A weak classifier is one chose error rate is only slightly better than random guessing. In the boosting algorithm, the weak classification algorithm is repeatedly applied to modified versions of the data, thereby producing a sequence of weak classifiers Gn,(x), m, = l, 2, ..., M. The predictions from all of the classifiers in this sequence are then combined through a weighted majority vote to produce the final prediction:
M
G(x) = SZ~ ~ amen, (x) m=1 Here al, a2, ,.,, au are computed by the boosting algorithm and their purpose is to weigh the contribution of each respective G,n(x). Their effect is to give higher influence to the more accurate classifiers in the sequence.
The data modifications at each boosting step consist of applying weights w1, w2, ..., wn to each of the training observations (x;, y;), i = 1, 2, ..., N. Initially all the weights are set to w; = 1/N, so that the first step simply trains the classifier on the data in the usual manner. For each successive iteration m = 2, 3, ..., M the observation weights are individually modified and the classification algorithm is reapplied to the weighted observations. At stem m, those observations that were misclassified by the classifier Gm_ 1(x) induced at the previous step have their weights increased, whereas the weights are decreased for those that were classified correctly. Thus as iterations proceed, observations that are difficult to correctly classify receive ever-increasing influence.
Each successive classifier is thereby forced to concentrate on those training observations that are missed by previous ones in the sequence.
The exemplary boosting algorithm is summarized as follows:
1. Initialize the observation weights w; = 1/N, i = 1, 2, ..., N.
2. For m = 1 to M:
(a) Fit a classifier Gm(x) to the training set using weights wt.
(b) Compute ~Nl wrI (Yr ~ G.., (xr )) errn~ = T/
~i-1 wi (c) Compute am log((1-errm)/errm).
(d) Set w; ~ w; ~ exp [am ~ I ( y; ~ Gm (x1 ))], i = 1, 2, . . . , N.
3. Output G(x) = sign ~ n 1 a",G", (x), In the algorithm, the current classifier Gm(x) is induced on the weighted observations at line 2a. The resulting weighted error rate is computed at line 2b. Line 2c calculates the weight am given to Gm(x) in producing the final classifier G(x) (line 3). The individual weights of each of the observations are updated for the next iteration at line 2d.
Observations misclassified by Gm(x) have their weights scaled by a factor exp(am), increasing their relative influence for inducing the next classifier Gm+i(x) in the sequence.
In some embodiments, modifications of the Freund and Schapire, 1997, Journal of Computer and System Sciences 55, pp. 119-139, boosting method are used. See, for example, Hasti et al., The Elements of Statistical Learning, 2001, Springer, New York, Chapter 10. In some embodiments, boosting or adaptive boosting methods are used.
In some embodiments, modifications of Freund and Schapire, 1997, Journal of Computer and System Sciences 55, pp. 119-139, are used. For example, in some embodiments, feature preselection is performed using a technique such as the nonparametric scoring methods of Park et al., 2002, Pac. Symp. Biocomput. 6, 52-63.
Feature preselection is a form of dimensionality reduction in which the genes that discriminate between classifications the best are selected for use in the classifier. Then, the LogitBoost procedure introduced by Friedman et al., 2000, Ann Stat 28, 337-407 is used rather than the boosting procedure of Freund and Schapire. In some embodiments, the boosting and other classification methods of Ben-Dor et al., 2000, Journal of Computational Biology 7, 559-583 are used in the present invention. In some embodiments, the boosting and other classification methods of Freund and Schapire, 1997, Journal of Computer and System Sciences 55, 119-139, are used.
In the random subspace method, classifiers are constructed in random subspaces of the data feature space. These classifiers are usually combined by simple majority voting in the final decision rule. See, for example, Ho, "The Random subspace method for constructing decision forests," IEEE Trans Pattern Analysis and Machine Intelligence, 1998; 20(8): 832-844.
5.14.12 OTHER MATHEMATICAL MODELS
The pattern classification and statistical techniques described above are merely examples of the types of classifiers that can be used to construct a classifier. Moreover, combinations of the techniques described above can be used. Some combinations, such as the use of the combination of decision trees and boosting, have been described. However, many other combinations are possible. In addition, other techniques in the art such as Projection Pursuit and Weighted Voting can be used to construct classifiers in instances of step 216.
5.15 IMPLEMENTING THE INVENTION
The present invention provides methods and systems for screening molecular markers to identify classifiers and/or for identifying classifiers for a trait and allows for the configuration of classifiers based on combinations of a large number of molecular markers.
The invention provides a selection process for reducing the potential large number of candidate molecular markers and/or combinations thereof down to a manageable number which can be evaluated in one or more mathematical models to derive one or more classifiers.
Some embodiments of the invention are preferably implemented on a computer system having a processor and a memory unit. The embodiments may be implemented as one or more software programs operating on a general purpose computer, such as a personal computer or workstation, or as dedicated special purpose hardware components.
The invention allows for an identification of classifiers while reducing the system requirements. It provides techniques for consideration of a large number of potential molecular markers in the classif er identification process with limited memory and computational requirements.
Some embodiments of the invention provide a data-driven selection of a subset of the candidate molecular markers based on their discrimination ability. Thus, it becomes possible to start out with a large group of potentially interesting molecular markers and to automatically prune the set of candidate molecular markers so that the computer system can handle the classifier identification process more efficiently, i.e. within less processing time and less memory space. This becomes more important as combinations of the molecular markers are generated in the classifier identification process and mathematical models are applied to each combination to derive classifiers, which may be a computationally expensive process, in particular when iterative techniques are applied, such as clustering, decision trees, neural networks, or evolutionary methods.
Since the possible number of combinations grows almost exponentially with the number of candidate molecular markers, the processing time for the classifier evaluation become a serious problem. The pruning of candidate molecular markers allows for the consideration of two, three, four and more combinations of candidate molecular markers as basis for the classifiers. It enables the evaluation of a large number of classifiers based on molecular markers showing a promising discrimination ability and supersedes a computer resource consuming evaluation of classifiers based on molecular markers which are likely not contributing to the final trait discrimination. Thus, the invention can be implemented on a computer having less computational power and memory while the quality of the derived classifiers is maintained. On the other hand, the invention allows fox the consideration of more molecular markers, which are potentially interesting for a given application (trait), with the same available system resources resulting in a possible higher classification accuracy for the derived classifiers.
The invention further provides a data-driven selection of the derived classifiers to remove classifiers and/or moleculax markers which do not significantly contribute to the trait discrimination in the later application phase. This automatic pruning step evaluates the discrimination power of the individual classifiers to reduce the necessary system requirements of a diagnostic system applying the selected classifiers for the wide variety of possible medical applications. Thus, a diagnostic system configured with the identified classifiers needs less computational power and memory and may be implemented on a smaller, less expensive device. This becomes more important when the applied mathematical models are more complex and powerful. Examples of complex classifiers are described in section 5.14 of this description and include, e.g. neural networks, nearest neighbor classifiers, decision tress, etc. The invention enables the operation of optimized combinations of complex classifiers on diagnostic devises with limited resources.
5.16 EMBODIMENTS OF THE INVENTION
6. EXAMPLES
Computer systems, computer program products, methods, and kits for providing health care have been disclosed. What follows are select examples that illustrate the utility and value of the present invention.
6.1 EVALUATING OSTEOARTHRITIS CLASSIFIERS USING ROC
CURVES
This example demonstrates the use of an embodiment of the invention to identify individuals with mild osteoarthritis. Osteoarthritis is a form of degenerative joint disease that involves the deterioration of and changes to the cartilage and bone. In response to inflammation in and about the joint, the body responds with bony recalcification around the joint structure. This process can be slow and gradual with minimal outward symptoms, or more rapidly progressive with significant pain and discomfort. Arthritic changes can occur in response to infection and injury of the joint as well.
Step 202 - generation of a training population.
Blood samples were taken from 44 test individuals not having any symptoms of osteoarthritis and 50 individuals having mild osteoarthritis using the methods described in Section 5.2. A molecular marker profile resulting in data for molecular marker products of the entire human genome was measured from each of these samples. This gene expression profile data together with knowledge of which subjects have osteoarthritis and which do not constitutes the training population 44. The 44 test individuals that do not have any symptoms of osteoarthritis constitute one trait subgroup within the training population and the 50 individuals having mild osteoarthritis constitute another trait subgroup within the training population.
Steps 204-218.
The training population collected in step 202 was used in order to identify combinations of genes that can serve as a classifier to differentiate mild osteoarthritis from non-osteoarthritis. Thus, the classifiers developed in this example are designed to yield a positive score when they predict that a subject has mild osteoarthritis and a negative score when they predict that the subject is in the control population. Using the approach described in Section 5.1, two specific classifiers were developed: 100000252 and 100000511. Classifier 100000252 comprises six genes and has the format:
SCORE = -1.839 + 0.8*HSPCA - 1.5525*IKBKAP + 1.10184*IL13RA1 +
0.78923*LAMCl - 1.3974*MAFB + 1.0602*PF4.
Classifier 100000511 comprises nine genes and has the format:
SCORE = -4.3754 + 0.10276*EGRl -1.1697*G2AN + 0.88767*HSPCA -0.55785*II~BKAP + 0.94015*IL13RA1 + 0.67515*LAMC1 - 1.5068*MAFB +
1.0798*PF4 + 0.4007*TNFAIP6.
Here, EGRl, G2AN, HSPCA, II~BI~AP, IL13RA1, LAMC1, MAFB, and TNFAIP6 are genes that were identified in step 204 and validated in step 208 (Section 5.1) for their ability to discriminate between subjects that have mild osteoarthritis and subjects that do not have osteoarthritis.
Step 2 ~0.
To judge which classifier is more suitable as a classifier for mild osteoarthritis, a ROC curve was computed for both classifiers using the gene expression data from the 44 test individuals not having any symptoms of osteoarthritis and the 50 individuals having mild osteoarthritis. The results of the ROC computation are illustrated in Fig. 8. The area under each ROC was computed. From this computation, it was determined that the area under the ROC curve corresponding to classifier 100000252 was 0.863 whereas the area under the ROC curve corresponding to classifier 100000511 was 0.8169.
Step 224.
In some embodiments, a classifier can be constructed that includes both classifiers 100000252 and 100000511 using the voting methods described in Section 5.1. In alternative embodiments, classifier 100000252 is selected to serve as a classifier for mild osteoarthritis because it generated a larger area under the ROC curve corresponding to the classifier when tested against the training population.
6.2 IDENTIFIED MOLECULAR MARKERS AND MOLECULAR
MARKER DATA MEASUREMENT TECHNIQUES
Molecular markers useful for input into one or more steps of the invention and techniques for measuring data values of such molecular markers, can be found in United States patent application serial No. 10/601,518, filed June 20, 2003, United States patent application serial No.lO/802,875, filed March 12, 2004, United States patent application serial No. 10/809,675, filed March 25, 2004, United States patent application serial No.
10/268,730, filed October 9, 2002, United States patent application serial No.
09/477,148, filed January 4, 2000, United States patent application serial No. 60/115,125, filed Jan. 6, 1999; and United States patent application serial No. 60/51,977, filed June 21, 2004 each of which is hereby incorporated herein by reference in its entirety.
6.3 CONSTRUCTION OF CLASSIFIERS FOR MANIC DEPRESSION
SYNDROME
This example demonstrates the use of the claimed invention to identify biomarkers to differentiate manic depression syndrome from non manic depression syndrome and use of same. As used herein, "manic depression syndrome" (MDS) refers to a mood disorder characterized by alternating mania and depression.
Step 202.
Blood samples were taken from patients who were diagnosed with manic depression as defined herein. In each case, the diagnosis of manic depression was corroborated by a skilled Board certified physician. Molecular marker data was measured for each of the molecular markers of the entire human genome using blood samples from individuals who were identified as having manic depression as described herein and ndividuals not having manic depression. Molecular marker data for both trait subgroups were compared and gene expression profiles for each trait subpopulation compared using commercially available GeneSpringTM softwares. Hybridizations to create the gene expression profiles were done using Affymetrix~ GeneChip~ platforms (U133A and U133 Plus 2.0) as described herein (data not shown). Samples from patients were clustered into two trait subgroups. The first trait subgroup included patients who have manic depression and the second trait subgroup included patients who do not have manic depression (i.e., control individuals).
Step 204.
The Wilcox Mann Whitney rank sum test was used to identify molecular marker data that could discriminate between the control and diseased trait subgroups with a p value of < 0.05.
Step 206.
Molecular markers were selected from those identified with p value of < 0.05 and the ability to discriminate between the control and diseased trait subgroups were confirmed using quantitative RT-PCR.
Steps 214-218.
Eight candidate moleculax markers were chosen and an exhaustive analysis of all possible combinations of said molecular markers were considered. Molecular marker data for each of the eight candidate molecular markers was obtained for each member of the training population and logistic regression applied to the molecular marker data so as to develop multiple classifiers. Each classifier was ranked on the basis of area under the curve and those classifiers with an ROC of greater than 0.9 chosen.
6.4 CONSTRUCTION OF CLASSIFIERS FOR PREDICTING
RESPONSE TO TREATMENT
This example demonstrates the use of an embodiment of the invention to identify a classifier for predicting the response of a subject to treatment.
Step 202.
Blood samples are taken from patients (for example patients with a disease) who are going to enter into treatment (for the disease), or who are already undergoing treatment, but at a timepoint before being able to determine how the patients will respond to treatment. In one embodiment, blood samples are taken from patientswho are about to enter into a clinical trial for a new treatment, or who are in the early stages of a clinical 1 S trial. Preferably blood samples are processed so as to preserve the RNA
and/or the protein products of the molecular markers. More preferably the blood samples are processed immediately, within 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 10 hours, 12, hours, 18 hours or 24 hours from having taken the blood samples from the patients.
Subsequent to when the blood samples are taken, patients continue to be monitored for response to treatment using traditional diagnostic methods and grouped into trait subgroups on the basis of the response to treatment. For example, trait subgroups can include patients with a positive response and no negative side effects, patients with a positive response and mild side effects, patients with a negative response, patients with a toxic response, and the like. In some embodiments, the evaluation of response to treatment can take days, weeks, months or years. In some embodiments, data as described in step 202 of Section 5.1 and Section 5.3 is obtained upon processing of the blood sample. In other embodiments, data of step 202 is obtained only after a determination of response to treatment has been made. In all cases, molecular marker data is obtained from a blood sample taken at a timepoint prior to being able to determine response to treatment. Once trait subgroups are identified on the basis of response to treatment, gene expression profiles of blood samples of each trait subgroup are compared using GeneSpringTM
software analysis. Hybridizations to create the gene expression profiles are done using Affymetrix~ GeneChip~ platforms (U133A and U133 Plus 2.0) platforms (U133A and U133 Plus 2.0) as described herein. Samples from patients are clustered into two trait subgroups on the basis of the response to treatment. For example, one trait subgroup demonstrates a positive response to treatment whereas the second trait subgroup demonstrates a toxic response to treatment.
Step 204.
The Wilcox Mann Whitney rank sum test is used to identify candidate molecular markers by identifying molecular marker data that discriminates between the response and non-response trait subgroups with a p value of < 0.05 to obtain candidate molecular markers.
Step 206.
Additional molecular marker data for the candidate molecular markers are obtained using quantitative RT-PCR. Some candidate molecular markers are removed at this point should the quantitative RT-PCR data not confirm the ability of each candidate molecular marker to discriminate as between the response trait subgroups.
Steps 214 - 2I8.
Candidate molecular marker combinations are chosen and an exhaustive analysis of all possible combinations of molecular markers are tested. To test all possible combinations of molecular markers, logistic regression is applied to the molecular marker data so as to develop multiple classifiers. Each classifier is ranked on the basis of area under the curve using the training population. Those classifiers ranking with an ROC area under curve of greater than 0.9 are further evaluated using a scoring population which is not the training population. Note that the blood samples used for the scoring population are obtained at the same time point as the blood samples used for the training population (e.g., at a time prior to being able to determine response to treatment).
6.5 CONSTRUCTION OF CLASSIFIERS FOR DETERMINING A
TRAIT OF INTEREST
6.5.1 This example demonstrates the selection of the composition of the training population and the trait subgroups of the training population so as to result in classifiers which are useful to predict disease.
Step 202.
In order to predict disease, blood samples are taken from patients at a time when said patients are disease free. Preferably blood samples are processed so as to preserve the RNA and/or the protein products of all molecular markers of the entire genome of said individual. More preferably the blood samples are processed immediately, within 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 10 hours, 12, hours, 18 hours or 24 hours from having taken the blood samples from the patients.
Subsequent to when the blood samples are taken, patients continue to be monitored for development of said disease using traditional diagnostic methods. At a given time point, two trait subgroups are identified, namely individuals who develop said disease of interest and individuals who do not develop said disease of interest. In some embodiments, the timepoint at which trait subgroups are identified can take days, weeks, months or years. In some embodiments, data as described in step 202 of Section 5.1 and Section 5.3 is obtained upon processing of the blood sample. In other embodiments, data of step 202 is obtained only after a determination of trait subgroups has been made. In all cases, data is obtained from a blood sample taken at a timepoint prior to being able to determine disease. Once trait subgroups are identified gene expression profiles of the molecular marker data of the molecular marker products from the blood samples of each trait subgroup are compared using GeneSpringTM software analysis.
Hybridizations to create the gene expression profiles are done using Affymetrix~ GeneChip~
platforms (U133A and U133 Plus 2.0) platforms (U133A and U133 Plus 2.0) as described herein and candidate molecular markers are identified where the molecular marker data is able to differentiate as between said two trait subgroups with a p value of <0.05.
Said candidate molecular markers are subsequently processed as described in steps 206 to 226 to identify classifiers and molecular markers capable of predicting disease.
6.5.~
This example demonstrates the selection of the composition of the training population and the trait subgroups of the training population so as to result in classifiers which are useful to determine treatment compliance.
Step 202.
In order to determine treatment compliance, blood samples axe taken from patients who are complying with said treatment of interest and patients who are not complying with said treatment. Preferably blood samples are processed so as to preserve the RNA and/or the protein corresponding to molecular markers of the entire genome of said individual.
More preferably the blood samples are processed immediately, within 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 10 hours, 12, hours, 18 hours or 24 hours from having taken the blood samples from the patients.

Molecular marker data is obtained as described in step 202 of Section 5.1 and Section 5.3 upon processing of the blood sample. Hybridizations to create the gene expression profiles are done using Affymetrix~ GeneChip~ platforms (U133A and Plus 2.0) platforms (U133A and U133 Plus 2.0) as described herein and candidate molecular markers are identified which differentiate as between patients who comply with said treatment of interest as compared with patients who do not comply with said treatment of interest with a p value of <0.05. Said candidate molecular markers are subsequently processed as described in steps 206 to 226 to identify classifiers and molecular markers capable of determining treatment compliance.
6.5.3 This example demonstrates the selection of the composition of the training population and the trait subgroups of the training population so as to result in classifiers which are useful to predict reoccurrence of disease.
Step 202.
In order to predict reoccurrence of disease, blood samples are taken from patients, all of whom have had a disease of interest, at a time when all of said patients are disease free. Preferably blood samples are processed so as to preserve the RNA and/or the protein corresponding to molecular markers of the entire genome of said individual.
More preferably the blood samples are processed immediately, within 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 10 hours, 12, hours, 18 hours or 24 hours from having taken the blood samples from the patients.
Subsequent to when the blood samples are taken, patients continue to be monitored for reoccurence of said disease using traditional diagnostic methods. At a given time point, two trait subgroups are identified, namely individuals who develop reoccurrence of said disease of interest and individuals who do not develop said disease of interest. In some embodiments, the timepoint at which trait subgroups are identified can take days, weeks, months or years. In some embodiments, data as described in step 202 of Section 5.1 and Section 5.3 is obtained upon processing of the blood sample. In other embodiments, data of step 202 is obtained only after a determination of trait subgroups has been made. In all cases, data is obtained from a blood sample taken at a timepoint prior to being able to determine reoccurrence of disease. Once trait subgroups are identified molecular marker data of the molecular marker products from the blood samples of each trait subgroup are compared using GeneSpringTM software analysis. Hy bridizations to create the gene expression profiles are done using Affymetrix~ GeneChip~
platforms (LTT'33A~~arid UI33 PIiis'2:(I) platt'orms (LT133A and U133 Plus 2.0) as described herein and candidate molecular markers are identified whose molecular marker data differentiates as between said two trait subgroups with a p value of <0.05. Said candidate molecular markers are subsequently processed as described in steps 206 to 226 to identify classifiers and molecular markers capable of predicting reoccurrence of disease.
6.5.4 This example demonstrates the use of classifiers used in series so as to diagnose a patient with a stage of disease, for example a specific stage of osteoarthritis.
We have identified previously four different stages of osteoarthritis; namely mild osteoarthritis, moderate osteoarthritis, marked osteoarthritis and severe osteoarthritis (see for example PCT patent application W002070737 Entitled "Compositions and Methods relating to osteoarthritis"
In some instances it is useful to determine which stage of osteoarthritis an individual has, and more importantly to confirm that said patient does not have any other stage of osteoarthritis. For example if an individual has changed their lifestyle and lost weight to determine whether the osteoarthritis has regressed.
Classifiers of the invention are able to differentiate as between two subgroups. As such - multiple classifiers are required to specifically stage an individual.
A first classifier is developed which differentiates as between osteoarthritis and non-osteoarthritis. As described, a training population is selected comprised of two trait subgroups where said first trait subgroup is comprised of individuals having osteoarthritis and the second trait subgroup comprised of individuals not having osteoarthritis.
Identification of candidate molecular markers which differentiate as between these two trait subgroups are identified as per step 202 and subsequently processed as described in steps 206 to 226 to identify classifiers and molecular markers capable of differentiating between osteoarthritis and non-osteoarthritis.
Similarly classifiers are identified which are capable of differentiating between (a) mild osteoarthritis and moderate osteoarthritis (b) moderate osteoarthritis and marked osteoarthritis (c) marked osteoarthritis and severe osteoarthritis.
In order to diagnose an individual as having marked osteoarthritis and not having mild osteoarthritis or severe osteoarthritis, a series of tests are applied.
First a classifier which determines whether said patient has osteoarthritis or not is applied.
Assuming said patient has osteoarthritis, a classifier is applied which determines whether said patient has either mild osteoarthritis or marked osteoarthritis. Assuming said patient has marked osteoartl~it"is;~a c'Tassi'1-ier is~applied to determine whether said patient has marked osteoarthritis or severe osteoarthritis. The result of these series of classifiers can determine that said patient has marked osteoarthritis and does not have any other stage of osteoarthritis.
6.6 CONSTRUCTION OF CLASSIFIERS FOR DETERMINING A
TRAIT OF INTEREST (OR DIFFERENTIATING BETWEEN TWO
TRAITS OF INTEREST) USING THE MOLECULAR MARKERS
IDENTIFIED IN ONE OF THE DISCLOSED TABLES
While the examples described below suggest selection of molecular markers prior to generating all combinations of classifiers from the disclosed Tables can be based on a specific measure of statistical significance (p value) as disclosed for each molecular marker (see Table G and Table H) it must be appreciated that the selection of molecular markers may also be based on any other method disclosed in this application, such as differential fold change or even a combination of selection ofp value and differential fold change. A skilled person in the art will recognize these other methods can also be used so as to permit the selection of subsets of the molecular markers to derive lists for which a reasonable number of combinations can be tested within the limits of the computer processing capacity. The skilled person will be able to transfer the details given in this section for examples based on a p value evaluation to carry out the selection based on other disclosed selection methods or selection methods known to him.
TABLE F
Selected Table Trait of InterestRecommended Training Population Key Trait of Key Trait of Members of TraitMembers of Trait Sub rou A Sub rou B

Table 1A Osteoarthritis Members have Members have and both Hypertension Osteoarthritis neither and Hypertension Osteoarthritis nor Hypertension Table 1 B Osteoarthritis Members have Members have and both Obesity Osteoarthritis neither and Obesity Osteoarthritis nor Obesity Table 1 C Osteoarthritis Members have Members have and both Allergies Osteoarthritis neither and Allergies Osteoarthritis nor Allergies Table 1D Osteoarthritis Members have Members have and both ''" "S'ystei'mic"SteroidsOsteoarthritis neither and are taking SystemicOsteoarthritis nor Steroids are taking Steroids Table 1 E Hypertension Members have Members do not H ertension have Hypertension Table 1F Obesity Members are ObeseMembers are not Obese Table 1 G Hypertension Members have Members do not Hypertension. have Hypertension Members have Members do not Hypertension have Osteoarthritis and Osteoarthritis.

Table 1H Hypertension Members have Members do not and Osteoarthritis Hypertension have either and Osteoarthritis Hypertension or Osteoarthritis Table 1I Obesity Members are ObeseMembers are not Obese Members have Members do not Obesity and have Osteoarthritis.

Osteoarthritis Table 1 J Obesity and Members have Members do not Osteoarthritis Osteoarthritis have either and are also Obese Osteoarthritis and are not Obese Table III Allergies Members have Members do not Allergies have Allergies Members have Members do not Allergies and have Osteoarthritis Osteoarthritis Table 1L Allergies and Members have Members do not Osteoarthritis Allergies and have either Allergies Osteoarthritis or Osteoarthritis Table 1 M Systemic SteroidsMembers have Members have been not taking Systemic been taking Steroids Systemic Steroids Members have Members do not been taking Systemic have Osteoarthritis.

Steroids and have Osteoarthritis Table 1N Systemic SteroidsMembers have Members do not and OsteoarthritisOsteoarthritis have Osteoarthritis and have been takingand have not been Systemic Steroidstaking systemic Steroids Table 1 O Taking Birth Members taking Members not taking Control Birth Control Birth Control Taking PrednisoneMembers taking Members not taking Prednisone Prednisone Taking Hormone Members taking Members not taking Replacement Hormone Hormone .. ;,,..Y --, ~.." ""~ ,yherapy ~.
,.... .,.". eplacement eplacement . .,.
.

Therapy Therapy Table 1P Type II DiabetesMembers have Members do not Type II Diabetes have Type II

Diabetes Table 1 Q Hyperlipidemis Members have Members do not Hyperlipidemia have Hyperlipidemia Table 1 R Lung Disease Members have Members do not Lung Disease have Lung Disease Table 1 S Bladder Cancer Members have Members do not Bladder Cancer have Bladder Cancer Table 1T Early Stage BladderMembers have Members do not Early Cancer Stage Bladder have Bladder Cancer Cancer Members have Members do not Early Stage Bladder have Early Stage Cancer Bladder Cancer Late Stage BladderMembers have Members do not Late Cancer Stage Bladder have Bladder Cancer Cancer Members have Members do not Late Stage Bladder have Late Stage Cancer Bladder Cancer Table 1 U Coronary Artery Members have Members do not CAD

Disease (CAD) have CAD

Table 1 V Rheumatoid Members have Members do not RA

Arthritis (RA) have RA

Table 1 W Rheumatoid Members have Members do not RA

Arthritis (R.A) have RA

Table 1X Depression Members have Members do not De ression have Depression Table 1Y Stage of Members have Members do not Mild Osteoarthritis OA have OA
- Mild Stage of Members have Members do not Osteoarthritis Moderate OA have OA
-Moderate Stage of Members have Members do not Osteoarthritis Marked OA have OA
-Marked Stage of Members have Members do not Osteoarthritis Severe OA have OA
-Severe Table 1 Z Liver Cancer Members have Members do not Liver Cancer have Liver Cancer Table 1Z(b) Liver Cancer Members have Members do not Liver Cancer have Liver Cancer Table lAA Schizophrenia Members have Members do not Schizophrenia have Schizo hrenia Table lAB Chagas Disease Members have Members do not Chagas Disease have Chagas Disease Table lAC Asthma Members have Members have OA

.. . , " . ., Asthma and OA
. .... ._.."
.. , Table 1 AD Asthma Members have Members do not Asthma have Asthma Table 1 AE Lung Caneer Members have Members do not Lung Cancer have Lung Cancer Table lAG Hypextension Members have Members do not Hypertension have Hypertension Table lAH Obesity Members have Members do not Obesity have Obesity Table lAI Ankylosing Members have Members do not Spondylitis Ankylosing have Ankylosing Spondylitis Spondylitis Table 2 Osteoarthritis Members have Members da not Osteoarthritis have Osteoarthritis Table 3A Schizophrernia Members have Members have or MDS

Manic DepressionSchizophrenia Syndrome (MDS) Table 3B Hepatitis or Members have Members have Liver Liver Cancer He atitis Cancer Table 3C Bladder Cancer Members have Members have or Liver Liver Cancer Bladder Cancer Cancer Table 3D Bladder Cancer Members have Members have or Testicular CancerBladder Cancer Testicular Cancer Table 3E Testicular CancerMembers have Members have or Kidney Cancer Testicular CancerKidney Cancer Table 3F Liver Cancer Members have Members have or Liver Stomach Cancer Cancer Stomach Cancer Table 3G Liver Cancer Members have Members have or Liver Colon Cancer Cancer Colon Cancer Table 3H Stomach Cancer Members have Members have or Colon Cancer Stomach Cancer Colon Cancer Table 3I Rheumatoid Members have Members have Arthritis or Rheumatoid Osteoarthritis Osteoarthritis Arthritis Table 3K Chagas Disease Members have Members have or Heart Heart Failure Chagas Disease Failure Table 3L Chagas Disease Members have Members have or CAD

Coronary Artery Chagas Disease Disease Table 3N Coronary Artery Members have Members have CAD Heart Disease or Heart Failure Failure Table 3P Asymptomatic Members have Members have Chagas or Asymptomatic Symptomatic Symptomatic Chagas Chagas Chagas Table 3Q Alzheimer's or Members have Members have Schizophrenia Alzheimer's Schizophrenia Table 3R Alzheimer's or Members have Members have Manic DepressionAlzheimer's Manic Depression Syndrome 'T'abl'e''4A" w-~SteoartTiritisMembers have Members do not ~"' _"..~ ..M=
..

Osteoarthritis have Osteoarthritis Table 4B Osteoarthritis Members have Members do not Osteoarthritis have Osteoarthritis Table 4C Mild OsteoarthritisMembers have Members do not Mild Osteoarthritis have Osteoarthritis Table 4D Mild OsteoarthritisMembers have Members do not Mild Osteoarthritis have Osteoarthritis Table 4E Moderate Members have Members do not Osteoarthritis Moderate have Osteoarthritis Osteoarthritis Table 4F Moderate Members have Members do not Osteoarthritis Moderate have Osteoarthritis Osteoarthritis Table 4G Marked Members have Members do not Osteoarthritis Marked have Osteoarthritis Osteoarthritis Table 4H Marked Members have Members do not Osteoarthritis Marked have Osteoarthritis Osteoarthritis Table 4I Severe OsteoarthritisMembers have Members do not Severe Osteoarthritishave Osteoarthritis Table 4J Severe OsteoarthritisMembers have Members do not Severe Osteoarthritishave Osteoarthritis Table 4I~ Mild OsteoarthritisMembers have Members have Mild or Moderate Osteoarthritis Moderate Osteoarthritis Osteoarthritis Table 4L Mild OsteoarthritisMembers have Members have Mild or Moderate Osteoarthritis Moderate Osteoarthritis Osteoarthritis Table 4M Mild OsteoarthritisMembers have Members have Mild or Marked Osteoarthritis Marked Osteoarthritis Osteoarthritis Table 4N Mild OsteoarthritisMembers have Members have Mild or Marked Osteoarthritis Marked Osteoarthritis Osteoarthritis Table 40 Mild OsteoarthritisMembers have Members have Mild or Severe Osteoarthritis Severe Osteoarthritis Osteoarthritis Table 4P Mild OsteoarthritisMembers have Members have Mild or Severe Osteoarthritis Severe Osteoarthritis Osteoarthritis Table 4Q Moderate Members have Members have Osteoarthritis Moderate Marked or Marked Osteoarthritis Osteoarthritis Osteoarthritis Table 4R Moderate Members have Members have Osteoarthritis Moderate Marked or Marked Osteoarthritis Osteoarthritis Osteoarthritis Table 4S Moderate Members have Members have ..
Osteoarthritis Moderate Severe Osteoarthritis or Severe OsteoarthritisOsteoarthritis Table 4T Moderate Members have Members have Osteoarthritis Moderate Severe Osteoarthritis or Severe OsteoarthritisOsteoarthritis Table 4U Marked Members have Members have Osteoarthritis Marked Severe Osteoarthritis or Severe OsteoarthritisOsteoarthritis Table 4V Marked Members have Members have Osteoarthritis Marked Severe Osteoarthritis or Severe OsteoarthritisOsteoarthritis Table SA Psoriasis Members have Members do not Psoriasis have Psoriasis Table SB Thyroid DisorderMembers have Members do not Thyroid Disorderhave Thyroid Disorder Table SC Irritable Bowel Members have Members do not Syndrome Irritable Bowelhave Irritable Bowel Syndrome Syndrome Table SD Osteoporosis Members have Members do not Osteoporosis have Osteoporosis Table SE Migraine HeadachesMembers have Members do not Migraine Headacheshave Migraine _,_ Headaches ~

Table SF Eczema Members have Members do not _ Eczema have Eczema Table SG NASH Members have Members do not NASH have NASH
~

Table SH Alzheimer's Members have Members do not Alzheimers' have Alzheimers' Table SI Manic DepressionMembers have Members do not Syndrome Manic Depressionhave Manic Syndrome Depression Syndrome Table SJ Crohn's Colitis Members have Members do not Crohn's Colitishave Crohn's Colitis Table SK Chronic CholecystisMembers have Members do not Chronic Cholecystishave Chronic Gholecystis Table SL Heart Failure Members have Members do not Heart Failure have Heart Failure Table SM Cervical Cancer Members have Members do not Cervical Cancerhave Cervical Cancer Table SN Stomach Cancer Members have Members do not Stomach Cancer have Stomach Cancer Table SO Kidney Cancer Members have Members do not Kidney Cancer have Kidney Cancer Table SP Testicular CancerMembers have Members do not Testicular Cancerhave Testicular " . ... ..... .... . Cancer ..~-~- ~~--_..,.

Table SQ Colon Cancer Members have Members do not Colon Cancer have Colon Cancer Table SR Hepatitis B Members have Members do not Hepatitis B have Hepatitis B

Table SS Pancreatic CancerMembers have Members do not Pancreatic Cancerhave Pancreatic Cancer Table ST Asymptomatic Members have Members do not Chagas Asymptomatic have Asymptomatic Chagas Chagas Table SU Symptomatic Members have Members do not Chagas Symptomatic have Symptomatic Chagas Chagas Table SV Bladder Cancer Members have Members do not Bladder Cancer have Bladder Cancer Table 6A Cancer Members have Members do not Cancer have Cancer Table 6B Cardiovascular Members have Members do not Disease Cardiovascular have Cardiovascular Disease Disease Table 6C Neurological Members have Members do not a Disorders Neurological have a Neurological Disorder Disorder Table 7A Celebrex~ or Members taking Members taking Other non Cox Inhibitor Celebrex~ Celebrex0 Cox Inhibitor Table 7B Celebrex~ Members taking Members not taking Celebrex~ Celebrex~

Table 7C Vioxx~ Members taking Members not taking Vioxx~ Vioxx~

Table 7D Vioxx~ or Other Members taking Members taking non Cox Inhibitor Vioxx~ Vioxx~ Cox Inhibitor Table 7E NSAIDS Members taking Members not taking NSAIDS NSAIDS

Table 7F Cortisone Members taking Members not taking Cortisone Cortisone Table 7G Visco SupplementMembers taking Members not taking Visco SupplementVisco Supplement Table 7H Lipitor~ Members taking Members not taking Lipitor~ Lipitor~

Table 7I Smokers Members are Members are not Smokers Smokers CONSTRUCTION OF CLASSIFIERS FOR DETERMINING A TRAIT OF INTEREST
Steps 202-20~
In order to identify useful classifiers for a trait of interest, for example mild osteoarthritis, one or more of the tables listed in Table F above which have the same recommended training population can be used. Thus for example, for mild osteoarthritis, one can select one or more of Tables 1Y; 4C; 4D. These molecular markers listed resulted from application of Steps 202-204 as outlined in Figure 2A as more fully described for each Table herein. Once one or more Tables have been selected, it is helpful to select a subset of the molecular markers in the Table or Tables before praceeding to step 206. For example, combining Tables 1Y, 4C and 4D and selecting molecular markers where the molecular marker data demonstrates an ability to differentiate as between the two trait subgroups with a p value of less than 0.0001 results in 212 molecular markers.
Note that p values resulting from the molecular marker data for each molecular marker identified in any of Tables 1A to 7I can be found in Tables 8A or Tables 8B below.
Table 8A identifies molecular markers via the Clone ID of the probe used to hybridize to the molecz!lar marker products. The Clone ID corresponds to the Clone ID
found in tables 1 A; 1 AC; 1 B; 1 C; 1 D; 1 E; 1 F; 1 G; 1 H; 1 I; 1 J; 1 K; 1 L; 1 M; 1 N; 1 O; 1 P;
1Q; 1R; 1V; 1X; 1Y; 1Z; 2; 4A; 4C; 4E; 4G; 4I; 4K; 4M; 40; 4Q; 4S; 4V; 7A; 7B;
7C;
7D; 7E; 7H; and 7I (ie those Tables generated using the ChondroChipTM as outlined herein). Table 8A then identifies the comesporiding Table in which the molecular marker is identified via said Clone ID. Finally the p value of the molecular marker data obt~.ir~ed using the ChondroCIlipTM is listed. Note that Table 8A is sorted first by Table number and then by p value.
'~fabie 8B identifies molecular markers via the Affymetrix~ Spot ID of the probe pair used to hybridize to the molecular marker products. The Affy Spot ID
corresponds to the Affy Spot ID found in tables Tables 1 AA; 1 AB; 1 AD; 1 AE; 1 AG; 1 AH; 1 AT; 1 S; 1 T;
1U; 1W; 1Z(b); 3A; 3B; 3C; 3D; 3E; 3F; 3G; 3H; 3I; 3K; 3L; 3P; 3Q; 3R; 4B; 4D;
4I1; 4H;
4J; 4L; 4N; 4P; 4R; 4T; 4V; SA; SB; SC; SD; SEE; SF; SG; SH; SI; SJ; SIB; Sh;
SM; SN;
50; SP ~Q; SR; jS; ST; SU; 5V; 6A; 6B; 6C; 7F; and 7G (ie those Tables generated using the AffymetrixTM Gene Chip as outlined herein). Table 8B then identifies the corresponding Table in which the molecular marker is identified via said Affymetrix~
Spot ID. Finally the p value of the molecular marker data obtained using the ChondroChipTM is listed. Note that Table 8B is sorted first by Table number and then by p value.
Step 206 For the 212 selected molecular markers - a training population is chosen having two trait subgroups where the two trait subgroups are outlined in Table F
above as corresponding to the Tables used to select the molecular markers, thus in this example, the first trait subgroup is members having mild osteoarthritis and the second trait subgroup is members not having osteoarthritis. A blood sample from each member of the training population is obtained and processed using techniques as described herein and mRNA
isolated. The resulting mRNA is reverse transcribed using ABI's High Capacity cDNA
S Archive Kit and the cDNA is then used for quantitative RT-PCR so as to collect molecular marker data fox application to a logistic regression model. Amplification primers are designed for each of the 212 molecular markers. Preferably primers are chosen which amplify across an intron junction. Quantitative Real Time PCR is performed using Qiagen's QuantiTectTMSybr Green RT-PCR kit and data corresponding to the level of RNA for each of the molecular markers Steps 214 - 218.
From the 212 candidate molecular markers selected, an exhaustive analysis of all possible combinations of molecular markers using the molecular marker data obtained using quantitative RT-PCR data is tested using logistic regression so as to develop multiple 1 S classifiers. Each classifier is ranked on the basis of area under the curve using the training population. Those classifiers ranking with an ROC area under curve of greater than 0.8 are further evaluated using a scoring population which is not the training population. Those classifiers resulting in an ROC area under the curve of greater than 0.7 as determined using thc~ scbrW g-popuianon are setectea. ~acn of ine-seiecrea c,ia~s~~~e~-~
~~~~mp~sed of a combination of molecular markers from one of Tables 1Y, 4C and 4D.
USE OF THE SELECTED CLASSIFIER TO DIAGNOSE AN INDIVIDUAL AS
HAVING MILD OSTEOARTHRITIS.
Any of the selected classifiers can be used to diagnose an individual as having mild osteoarthritis. A blood sample from a test individual is processed using techniques as 2S described herein to isolate mRNA. mRNA is reverse transcribed using ABI's High Capacity cDNA Archive Kit and the cDNA is then used for quantitative RT-PCR.
Amplification primers are designed for each of the molecular markers in the classifier selected which amplify across an intron junction. Quantitative Real Time PCR
is performed using Qiagen's QuantiTectTMSybr Green RT-PCR kit and data corresponding to the level of RNA for each of the molecular markers of the selected classifier obtained. The data is then used in conjunction with the logistic regression classifier so as to convert the data resulting from the Quantitative RT-PCR into a single number. If the number is greater than 0 - the test individual is diagnosed as having mild osteoarthritis is the number is less than 0 the test individual is diagnosed as not having osteoarthritis.

7. REFERENCES CITED
All references cited herein are incorporated herein by reference in their entirety and for all purposes to the same extent as if each individual publication or patent or patent application was specifically and individually indicated to be incorporated by reference in its entirety for all purposes.
The present invention can be implemented as a computer program product that comprises a computer program mechanism embedded in a computer readable storage medium. For instance, the computer program product could contain the program modules shown in Fig. 1. These program modules may be stored on a CD-ROM, DVD, magnetic disk storage product, or any other computer readable data or program storage product. The software modules in the computer program product can also be distributed electronically, via the Internet or otherwise, by transmission of a computer data signal (in which the software modules are embedded) on a carrier wave.
Many modifications and variations of this invention can be made without departing from its spirit and scope, as will be apparent to those skilled in the art.
The specific embodiments described herein are offered by way of example only, and the invention is to be limited only by the terms of the appended claims, along with the full scope of equivalents to which such claims are entitled.

T'a'ble'~~A
Clone TableNo.p-value Clone ID TableNO.
ID P
Value miob81431A 2.14e-05 fcrb4231 1A 3.784194e-03 ncrc58441A 1.1356e-04 seob0755 1A 3.784194e-03 hfcr31491A 1.30609e-04 seoc1023 1A 3.784194e-03 fcrb33301A 1.71826e-04 fcrb4995 1A 4.164516e-03 nerc97721A 1.7I826e-04 fcrb6031 1A 4.164516e-03 fcrb21621A 2.24449e-04 fcrb6896 1A 4.164516e-03 ncrb83431A 2.55864e-04 fcrc1689 1A 4.164516e-03 fcrb42261A 2.91185e-04 fcrc6228 1A 4.164516e-03 seoc61821A 2,91185e-04 ncr7672 1A 4.164516e-03 seoa34081A 3.75265e-04 ncrcl885 1A 4.164516e-03 fcrb61911A 4.80528e-04 ncrc3544 1A 4.164516e-03 mioc10281A 4.80528e-04 seoa2381 1A 4.164516e-03 seob29661A 5.4249e-04 seob9872 1A 4.164516e-03 seoc07751A 5.4249e-04 fcrb2090 IA 4.577725e-03 fcre18341A 7.73507e-04 hfcr0489 1A 4.577725e-03 ncrc32571A 7.88975e-04 miod1448 1A 4.577725e-03 ncrc24721A 8.68041e-04 ncr7813 1A 4.577725e-03 mioc39301A 1.216349e-03 ncrc9712 1A 4.577725e-03 ncr08471A 1.216349e-03 ncrc9855 IA 4.577725e-03 miob89471A 1.253946e-03 seoa0256 IA 4.577725e-03 fcrb52141A 1.357347e-03 seoa3555 IA 4.577725e-03 miob40371A 1.512638e-03 mioa0601 IA 4.9929e-03 miob89321A 1.512638e-03 fcrb2713 IA 5.026153e-03 fcrc01661A 1.683445e-03 mioa1704 1A 5.026153e-03 miob82491A 1.683445e-03 miob6419 IA 5.026153e-03 seob34851A 1.683445e-03 mioc2997 IA 5.026153e-03 seob79291A 1.683445e-03 mioc3127 lA 5.026153e-03 seob01541A 1.871074e-03 ncr7904 lA 5.026153e-03 fcrc63451A 2.076917e-03 ncrc5088 IA 5.026153e-03 mioa50591A 2.076917e-03 seoa2641 1A 5.026153e-03 fcrb43781A 2.302455e-03 seoa3359 1A 5.026253e-03 miod40661A 2.302455e-03 seob5069 1A 5.026153e-03 ncrc50911A 2.302455e-03 seob5213 1A 5.026153e-03 mioc01621A 2.452971e-03 ncr1948 1A 5.341175e-03 fcrc55161A 2.549264e-03 fcrb9324 1A 5.5I2248e-03 miob43081A 2.549264e-03 fcrc6990 1A 5.512248e-03 seob13191A 2.549264e-03 mioc1598 1A 5.512248e-03 fcr19971A 2.829017e-03 seob4752 1A 5.512248e-03 fcr44711A 2.819017e-03 mioa0528 1A 6,038584e-03 ncr06151A 2.819017e-03 mioc2074 1A 6,038584e-03 ncr95491A 2.819017e-03 miod5703 1A 6.038584e-03 ncrc07291A 2.819017e-03 ncrc3391 1A 6.038584e-03 fcrb4572IA 3.113486e-03 seob0168 1A 6,038584e-03 fcrb96551A 3.113486e-03 fcrb4345 1A 6.607863e-03 ncrb63201A 3.113486e-03 mioa2377 1A 6.607863e-03 seoa97091A 3.113486e-03 mioc1122 1A 6.607863e-03 seoa98701A 3.113486e-03 ncrc9637 1A 6.607863e-03 seob29591A 3.113486e-03 seob9152 IA 6.607863e-03 seoc10251A 3.113486e-03 fcrb1691 1A 7.2229I4e-03 fcr42141A 3.43455e-03 fcrb5928 1A 7.222914e-03 fcrb38971A 3.43455e-03 fcrc6010 1A 7.2229I4e-03 fcrb65081A 3.43455e-03 fcrc6566 1A 7.222914e-03 fcrb79441A 3.43455e-03 hfcr4477 1A 7.222914e-03 fcrc28071A 3.43455e-03 miob8691 1A 7.2229I4e-03 ncrc57801A 3.43455e-03 miod2330 IA 7.222914e-03 seoa04291A 3.43455e-03 ncr3037 1A 7.222914e-03 seoa19771A 3.43455e-03 ncrc9700 1A 7.222914e-03 fcr35591A 3.784194e-03 seoa9889 1A 7.222914e-03 t66 s'eob~~~~~1~A~ ~ :'~~2~~~14e-03miod7421 1A 0.

fcrb75841A 7.288639e-03 ncr0016 1A 0.01 fcrb52041A 7.886697e-03 ncrb8385 1A 0.01 fcrc56141A 7.886697e-03 ncrc5949 1A 0.01 mioc73311A 7.886697e-03 seoa2805 1A 0.01 mioc86821A 7.886697e-03 seoa6620 1A 0.01 ncr0212 1A 7.886697e-03 seob3141 1A 0.01 ncr7292 1A 7.886697e-03 fcr4639 1A 0.01 ncrb43311A 7.886697e-03 fcrb3237 1A 0.01 ncrb55951A 7.886697e-03 fcrb3244 1A 0.01 seoa66541A 7.886697e-03 fcrb4981 1A 0.01 seob45451A 7.886697e-03 fcrb5850 1A 0.01 seob83011A 7.886697e-03 fcrb9686 1A 0.01 seoc16641A 7.886697e-03 fcrb9959 1A 0.01 fcr1312 1A 8.602308e-03 hfcrl811 1A 0.01 fcrb20411A 8.602308e-03 miob3953 1A 0.01 mioa50851A 8.602308e-03 mioc0669 1A 0.01 ncrc88921A 8.602308e-03 ncrc2119 1A 0.01 seoa1584lA 8.602308e-03 ncrc4575 1A 0.01 seoa55771A 8.602308e-03 ncrc6127 1A 0.01 seob53191A 8.602308e-03 ncrc6382 1A 0.01 seob63861A 8.602308e-03 seoc0778 1A 0.01 seob87411A 8.602308e-03 seoc0924 1A 0.01 fcr1879 1A 9.372976e-03 ncrc1892 1A 0.01 fcrbl4571A 9.372976e-03 fcr0593 1A 0.01 fcrb36441A 9.372976e-03 fcrb3715 1A 0.01 fcrb43881A 9.372976e-03 mioc4420 1A 0.01 mioa88511A 9.372976e-03 miod3920 1A 0.01 mioa88521A 9.372976e-03 ncr6343 1A 0.01 mioc10601A 9.372976e-03 seob5209 1A 0.01 mioc19101A 9.372976e-03 ncr1876 lA 0.01 miod64371A 9.372976e-03 miod3417 1A 0.01 ncr6072 1.A 9.372976e-03 fcr4212 1A 0.01 seoa83991A 9.372976e-03 fcrb4413 1A 0.01 seob4197lA 9.372976e-03 fcrb5305 1A 0.01 seob90921A 9.372976e-03 fcrb9253 1A 0.01 fcrbl6891A 0.01 hfcr1189 1A 0.01 fcrb19901A 0.01 miod4464 1A 0.01 fcrb85421A 0.01 ncr3614 1A 0.01 seoa97051A 0.01 ncrc3536 1A 0.01 seob06881A 0.01 ncrc9562 1A 0.01 seob62791A 0.01 seoa3105 1A 0.01 seoc39651A 0.01 seob0085 1A 0.01 fcr3664 1A 0.01 seob3513 1A 0.01 fcr3717 1A 0.01 seob5044 1A 0.01 fcrb36861A 0.01 seob5336 1A 0.01 fcrb58131A 0.01 fcrb4985 1A 0.01 fcrb60121A 0.01 hfcr6634 1A 0.01 fcrb78031A 0.01 mioa5586 1A 0.01 fcrb89891A 0.01 mioa9831 1A 0.01 mioc69371A 0.01 miob6536 1A 0.01 ncrb84371A 0.01 miob7554 1A 0.01 ncrc56311A 0.01 mioc0728 1A 0.01 seoa24481A 0.01 mioc0904 1A 0.01 seoa85431A 0.01 mioc3040 1A 0.01 seob61311A 0.01 miod5627 1A 0.01 seob62721A 0.01 miod6731 1A 0.01 seob94061A 0.01 ncr2472 1A 0.01 seob98511A 0.01 ncrc3799 1A 0.01 fcr7042 1A 0.01 seob0879 1A 0.01 fcrb47991A 0.01 seob2797 1A 0.01 fcrc68771A 0.01 seob3112 1A 0.01 mioa96491A 0.01 seoc1189 1A 0.01 mioc11251A 0.01 fcr5625 1A 0.01 o n a...a fcrb2"~~~"1...r. mioa6913 1A 0.
.. 02 r....

'"
0"~'0~

hfcr65011A 0.01 miob5646 1A 0.02 miob67021A 0.01 mioc0528 1A 0.02 ncr3815 1A 0.01 miocl768 1A 0.02 seoa97771A 0.01 miod3592 1A 0.02 seob49251A 0.01 ncr1780 1A 0.02 seob83211A 0.01 ncrc6795 1A 0.02 mioc70841A 0.01 seoa6393 1A 0.02 fcrb44001A 0.01 seoa7126 1A 0.02 fcrb55501A 0.01 seob0497 1A 0.02 fcrb69681A 0.01 seob5004 1A 0.02 fcrc02411A 0.01 seoc2191 1A 0.02 fcrc17811A 0.01 hfcr3436 1A 0.02 mioa61351A 0.01 fcrb7240 1A 0.02 ncr8413 1A 0.01 fcrb3704 1A 0.02 seob80651A 0.01 fcrb3848 1A 0.02 seob98981A 0.01 fcrb5467 1A 0.02 seoc04991A 0.01 fcrb6734 1A 0.02 ncr0258 1A 0.02 fcrc4848 1A 0.02 ncr0179 1A 0.02 fcrc7228 1A 0.02 fcr4902 1A 0.02 hfcr4741 1A 0.02 fcr_b15291A 0.02 miob3845 1A 0.02 fcrb20441A 0.02 mioc0301 1A 0.02 fcrb38961A 0.02 mioc1354 1A 0.02 fcrb48901A 0.02 mioc5695 1A 0.02 fcrb60331A 0.02 miod0592 1A 0.02 fcrb64321A 0.02 ncr8156 1A 0.02 fcrc54021A 0.02 ncrc3377 1A 0.02 hfcr11411A 0.02 seoa5977 1A 0.02 mioa87781A 0.02 seoa7212 1A 0.02 miob73731A 0.02 seob1526 1A 0.02 miob85721A 0.02 seoc0945 1A 0.02 miob91301A 0.02 mioc0899 1A 0.02 miob97881A 0.02 ncr9003 1A 0.02 mioc12031A 0.02 seoa7530 1A 0.02 ncr0808 1A 0.02 fcrb2207 1A 0.02 ncrc96331A 0.02 fcrb6102 1A 0.02 seoa61721A 0.02 hfcr0501 1A 0.02 seoa97241A 0.02 mioa5447 1A 0.02 seob02531A 0.02 miob3809 1A 0.02 seob02881A 0.02 miob8531 1A 0.02 seob54781A 0.02 miob8802 1A 0.02 hfcr31601A 0.02 miob9671 1A 0.02 fcr3323 1A 0.02 mioc6156 1A 0.02 fcr5425 1A 0.02 ncrb4339 1A 0.02 fcr7295 1A 0.02 ncrb8451 1A 0.02 fcrb14281A 0.02 ncrc0217 1A 0.02 fcrb53891A 0.02 seoa0469 1A 0.02 fcrc20991A 0.02 seoa1100 1A 0.02 miob31311A 0.02 seoa4066 1A 0.02 mioc25921A 0.02 seoa9389 1A 0.02 mioc81531A 0.02 fcrb9671 1A 0.03 miod49381A 0.02 fcrc5577 1A 0.03 miod51221A 0.02 hfcr0285 1A 0.03 miod57851A 0.02 mioa2073 1A 0.03 seoa63641A 0.02 miob9020 1A 0.03 seoa91501A 0.02 mioc2546 1A 0.03 seob41921A 0.02 mioc3139 1A 0.03 seob67581A 0.02 mioc7986 1A 0.03 mioc08521A 0.02 miod1316 1A 0.03 fcrb25921A 0.02 miod7337 1A 0.03 fcrb96801A 0.02 ncr3811 1A 0.03 hfcr40071A 0.02 ncr9975 1A 0.03 mioa51311A 0.02 ncrb4474 1A 0.03 ' ricrc5653~lA ~0"':~~~3 fcrb2350 1A 0.
" 03 ncrc6264 1A 0.03 fcrb5202 1A 0.03 seoa4802 1A 0.03 fcrb9636 1A 0.03 seoa5554 1A 0.03 fcrc4916 1A 0.03 seoa6497 1A 0.03 hfcr0237 1A 0.03 seoa8851 1A 0.03 mioa4552 1A 0.03 seob0047 1A 0.03 mioa6442 1A 0.03 seob0885 1A 0.03 miob5675 1A 0.03 fcrb3288 1A 0.03 miob6372 1A 0.03 fcrb4270 1A 0.03 miob8532 1A 0.03 fcrb4415 1A 0.03 miob8657 1A 0.03 hfcr5905 1A 0.03 mioc3208 1A 0.03 mioa4229 1A 0.03 miod4083 1A 0.03 miob9065 1A 0.03 miod6029 1A 0.03 mioc1940 1A 0.03 ncr0097 1A 0.03 miod6685 1A 0.03 ncr2812 1A 0.03 ncr0836 1A 0.03 ncr3960 1A 0.03 ncr3112 1A 0.03 ncr5651 1A 0.03 ncr4647 1A 0.03 ncr9779 1A 0.03 ncr6408 1A 0.03 ncrc1871 1A 0.03 ncrb8105 1A 0.03 ncrc1889 1A 0.03 ncrc9642 1A 0.03 seoa5685 1A 0.03 ncrc9739 1A 0.03 seoa7223 1A 0.03 seoa1540 1A 0.03 seoa9959 1A 0.03 seoa3429 1A 0.03 seob7649 1A 0.03 seoa3515 1A 0.03 fcrb6359 1A 0.03 seoa4452 1A 0.03 fcrb7237 1A 0.03 seoa5787 1A 0.03 fcrb8485 lA 0.03 seoa6598 1A 0.03 fcrc1849 1A 0.03 seob1844 1A 0.03 fcrc2536 1A 0.03 seob2195 1A 0.03 fcrc3993 1A 0.03 s~ob3533 1A 0.03 hfcr2616 1A 0.03 seoc2246 1A 0.03 hfcr4423 1A 0.03 ncr8153 1A 0.03 mioa1388 1A 0.03 fcrb2133 1A 0.03 mioa9555 1A 0.03 fcrb2849 1A 0.03 mioa9581 1A 0.03 fcrb5929 1A 0.03 miob0931 1A 0.03 fcrb6201 1A 0.03 miob9495 1A 0.03 fcrc1181 1A 0.03 mioc1135 1A 0.03 fcrc1858 1A 0.03 miocl438 1A 0.03 fcrc2670 1A 0.03 mioc2451 1A 0.03 fcrc4033 1A 0.03 mioc3671 1A 0.03 mioa8747 1A 0.03 mioc4119 1A 0.03 miob8773 1A 0.03 miod5258 1A 0.03 miob9128 1A 0.03 ncr2995 1A 0.03 mioc0315 1A 0.03 ncrb3957 1A 0.03 mioc0472 1A 0.03 ncrc3735 1A 0.03 mioc2662 1A 0.03 seoa2385 1A 0.03 mioc4994 1A 0.03 seoa3670 1A 0.03 ncr6142 1A 0.03 seoa3811 1A 0.03 ncrc5508 1A 0.03 seob1572 1A 0.03 ncrc6423 1A 0.03 seob7946 1A 0.03 ncrc6888'1A 0.03 seob9334 1A 0.03 seoa5911 1A 0.03 fcrb4515 1A 0.04 seob3517 1A 0.03 fcrb6028 1A 0.04 seob6206 1A 0.03 fcrc4380 1A 0.04 seob7907 1A 0.03 fcrc5007 1A 0.04 seoc0149 1A 0.03 fcrc5850 1A 0.04 seoc0617 1A 0.03 fcrc5898 1A 0.04 fcrb1890 1A 0.03 mioa3331 1A 0.04 seob1956 1A 0.03 mioa3392 1A 0.04 miob2756 1A 0.03 mioa4076 1A 0.04 fcrb2198 1A 0.03 mioa4484 1A 0.04 fcrb2256 1A 0.03 miob3042 1A 0.04 "iniob3~~~~f""~ ''""~ 't~"."~'4'~seoc4316 1A 0 . 04 '~"~~
~A

miob72091A 0.04 fcrc5506 lAC 6.56e-05 mioc47311A 0.04 fcrb7072 lAC 1.6e-04 ncr2363 1A 0.04 fcrb9202 lAC 2.5e-04 ncr8628 1A 0.04 fcrb2252 lAC 4.03e-04 ncrb22471A 0.04 miob2601 lAC 4.21e-04 ncrb41821A 0.04 ncrc7085 lAC 4.84e-04 ncrc34161A 0.04 miob9073 lAC 5.8e-04 seoa15751A 0.04 seoc2248 lAC 6.37499e-04 seob07521A 0.04 seoc2682 1AC 8.42e-04 seob14231A 0.04 miob5885 lAC 8.48e-04 seob50991A 0.04 ncrb1186 lAC 1.13586e-03 seob73461A 0.04 fcrb9147 lAC 1.49537e-03 fcr1984 1A 0.04 seob0846 lAC 1.62274e-03 fcrb23341A 0.04 seoa1812 lAC 1.82148e-03 fcrb86971A 0.04 mioa5955 lAC 2.06855e-03 fcrb92021A 0.04 ncrc0672 1AC 2.23855e-03 fcrc63051A 0.04 fcrb3040 lAC 2.56056e-03 hfcr22501A 0.04 fcrb6188 lAC 2.56177e-03 hfcr23141A 0.04 fcrc4948 lAC 2.61225e-03 hfcr36601A 0.04 seocl631 lAC 2.71926e-03 mioa54041A 0.04 ncrc6994 lAC 2.82137e-03 miob2833lA 0.04 mioa1701 lAC 2.82574e-03 miod28321A 0.04 ncrc9343 1AC 2.83011e-03 ncr5621 1A 0.04 hfcr2930 lAC 3.28473e-03 ncrb42481A 0.04 seoa9740 lAC 3.80488e-03 ncrc35981A 0.04 ncrc5919 lAC 3.81434e-03 seoa43271A 0.04 fcrc7051 1AC 3.82487e-03 seoa70941A 0.04 seoa3322 lAC 3.82543e-03 seoa99161A 0.04 seob4216 1AC 4.4181e-03 seob07001A 0.04 mioc5664 lAC 4.42431e-03 seobl3451A 0.04 hfcr0521 1AC 4.76360e-03 seob41911A 0.04 mioa2475 lAC 4.76360e-03 seoc47201A 0.04 seob1660 lAC 4.76360e-03 ncr0291 1A 0.04 fcrb2054 1AC 4.90887e-03 fcr3936 1A 0.04 miob4441 1AC 5.11853e-03 fcr5190 1A 0.04 seob5778 1AC 5.14747e-03 fcrb13991A 0.04 miod0195 lAC 5.19208e-03 fcrb15231A 0.04 mioc1910 lAC 5.43168e-03 fcrb57021A 0.04 mioa8622 lAC 5.48183e-03 fcrb70511A 0.04 seob2085 lAC 5.48183e-03 fcrb94301A 0.04 miod3579 lAC 5.84188e-03 fcrc05971A 0.04 fcrb1697 lAC 5.88330e-03 fcrc18791A 0.04 mioa6585 1AC 5.89493e-03 fcrc46581A 0.04 seoa6495 lAC 5.89880e-03 hfcr31341A 0.04 mioa6991 lAC 5.9049e-03 miob09391A 0.04 ncrb5737 1AC 6.09961e-03 miob26561A 0.04 ncrc5663 lAC 6.16618e-03 miob36901A 0.04 ncrcl495 lAC 6.31851e-03 mioc86351A 0.04 seoa5838 lAC 6.31851e-03 ncr3834 1A 0.04 mioa9604 lAC 6.43571e-03 ncrb06021A 0.04 hfcr2963 lAC 6.54225e-03 ncrb22821A 0.04 fcrb3595 lAC 6.65596e-03 ncrc95571A 0.04 seoc3511 lAC 6.77289e-03 seoa16441A 0.04 miod5310 lAC 6.78152e-03 seoa37611A 0.04 mioa1417 1AC 7.24653e-03 seoa52141A 0.04 ncr3763 lAC 7.46504e-03 seoa56371A 0.04 seoc5209 lAC 7.50804e-03 seoa61751A 0.04 fcrc6989 lAC 7.55113e-03 seob01331A 0.04 seob7755 lAC 8.04089e-03 seob19081A 0.04 seob3163 lAC 8.19449e-03 seob33671A 0.04 miob1126 lAC 8.32195e-03 seob87421A 0.04 fcrb3012 1AC 8.87793e-03 seoc02841A 0.04 ncrc2919 lAC 8.87793e-03 17~

mioal3'~'7~""'""'~~1C ".~.'."v~'~9~v0e-03seob8873 lAC 0.01 seob43331AC 8.89910e-03 fcr1853 1AC 0.01 ncr2930 lAC 9.25216e-03 seoa5994 1AC 0.01 fcr3277 1AC 9.43613e-03 seoa5833 1AC 0.01 mioa40771AC 9.48849e-03 seoa5766 1AC 0.01 fcrb2457lAC 9.49961e-03 seoa0418 lAC 0.01 seob65721AC 9.54268e-03 mioa0132 1AC 0.01 fcrb2754lAC 9.56817e-03 fcrb9289 1AC 0.01 fcrb17671AC 9.66528e-03 seob2974 1AC 0.01 fcrb66981AC 0.01 ncrb3424 1AC 0.01 seoa8684lAC 0.01 fcrc4005 lAC 0.01 fcrb6460lAC 0.01 miob3174 1AC 0.01 ncrb6282lAC 0.01 seob0815 lAC 0.01 ncrc50391AC 0.01 fcrb5393 1AC 0.01 mioa3997lAC 0.01 ncrc3036 lAC 0.01 mioc2602lAC 0.01 ncrc3129 1AC 0.01 seoa30021AC 0.01 mioa6240 1AC 0.01 fcrc0456lAC 0.01 ncrc6480 lAC 0.01 fcrb70631AC 0.01 ncr3727 1AC 0.02 fcrb9909lAC 0.01 fcrb7650 1AC 0.02 seob83331AC 0.01 ncrb5837 lAC 0.02 miod3501lAC 0.01 fcrb9280 lAC 0.02 seoa09311AC 0.01 seob3105 lAC 0.02 fcrc04711AC 0.01 ncr4535 1AC 0.02 mioc1590lAC 0.01 miocl963 lAC 0.02 mioa9258lAC 0.01 fcrb6776 lAC 0.02 seob33171AC 0.01 mioa1015 1AC 0.02 fcrb3476lAC 0.01 seob3887 lAC 0.02 mioa3963lAC 0.01 seoa7647 lAC 0.02 seob93001AC 0.01 seoa6598 1AC 0.02 fcr2842 1AC 0.01 mioa3514 1AC 0.02 seoa2633lAC 0.01 seob0228 1AC 0.02 ncr0570 lAC 0.01 seobl818 lAC 0.02 miob7836lAC 0.01 miod1323 lAC 0.02 ncrb60731AC 0.01 mioc3603 1AC 0.02 seob63951AC 0.01 miob8578 1AC 0.02 miob27001AC 0.01 mioa8740 1AC 0.02 fcr7656 lAC 0.01 fcrc0921 lAC 0.02 ncrc4089lAC 0.01 miob9124 lAC 0.02 fcrb8505lAC 0.01 seoa0420 1AC 0.02 seoa6654lAC 0.01 fcr4357 lAC 0.02 seob48351AC 0.01 fcrb8664 lAC 0.02 fcrb42411AC 0.01 seob3191 1AC 0.02 fcrc15631AC 0.01 fcrb6779 1AC 0.02 fcrc40511AC 0.01 fcrb8940 1AC 0.02 fcrc7219lAC 0.01 hfcr2984 lAC 0.02 fcrb23211AC 0.01 seoa5933 1AC 0.02 seoc4470lAC 0.01 fcr3539 lAC 0.02 ncr2869 lAC 0.01 miob8454 lAC 0.02 fcrb5202lAC 0.01 mioc7904 lAC 0.02 ncrb8820lAC 0.01 ncr3397 lAC 0.02 seob3158lAC 0.01 seocl311 lAC 0.02 seoa5992lAC 0.01 seob4793 1AC 0.02 fcrb21981AC 0.01 miob4055 lAC 0.02 fcr4746 lAC 0.01 mioa4285 lAC 0.02 ncr8811 1AC 0.01 ncrc2796 lAC 0.02 mioa32821AC 0.01 hfcr0786 1AC 0.02 ncrb50601AC 0.01 miob2645 1AC 0.02 seob2185lAC 0.01 miob2968 lAC 0.02 fcrb1930lAC 0.01 fcr2218 lAC 0.02 fcrb7215lAC 0.01 mioc0940 lAC 0.02 fcrc08921AC 0.01 fcrc0355 1AC 0.02 fcrc43071AC 0.01 seoc1180 1AC 0.02 mioa5452lAC 0.01 fcrb8383 lAC 0.02 ;: ..,a E ;;.~a_ .,.. "..
"~~crc25'~6.~ .; fcrc5007 lAC 0.
,E '~ '" ~I~". 03 .: b2~~
SAC

ncr1967 lAC 0.02 seoa6151 1AC 0.03 ncr8357 lAC 0.02 mioc6902 lAC 0.03 ncrc5417lAC 0.02 seob3415 lAC 0.03 seob5493lAC 0.02 seob4150 1AC 0.03 ncr3483 lAC 0.02 fcrb8700 lAC 0.03 ncrc9727lAC 0.02 seoc0098 lAC 0.03 mioa47381AC 0.02 miob2858 lAC 0.03 seob25551AC 0.02 mioc2735 1AC 0.03 miocl3571AC 0.02 ncr0016 1AC 0.03 seob7424lAC 0.02 seoa2300 lAC 0.03 miob65981AC 0.02 seob8425 lAC 0.03 ncrc57241AC 0.02 seoa7917 1AC 0.03 fcrc04721AC 0.02 mioa8987 lAC 0.03 fcrb9659lAC 0.02 miob8199 lAC 0.03 mioc98811AC 0.02 miod4686 lAC 0.03 ncrc1393lAC 0.02 miob0795 lAC 0.03 mioa14271AC 0.02 miob8249 lAC 0.03 ncrc07441AC 0.02 mioc6055 lAC 0.03 fcr3559 lAC 0.02 seoc1159 1AC 0.03 hfcr3486lAC 0.02 fcrb4721 1AC 0.03 mioc00901AC 0.02 mioc7370 lAC 0.03 seob9970lAC 0.02 seoa4647 lAC 0.03 miob1139lAC 0.02 miob8274 1AC 0.03 miob37251AC 0.02 hfcr2653 lAC 0.03 seoa47831AC 0.02 mioc2750 lAC 0.03 ncrc8865lAC 0.03 seoa9656 lAC 0.03 ncr3793 1AC 0.03 mioa6621 1AC 0.03 ncrc11021AC 0.03 ncrc1643 1AC 0.03 seoc0698lAC 0.03 seoc2518 1AC 0.03 ncrc6778lAC 0.03 fcrb8504 lAC 0.03 fcrb6185lAC 0.03 ncrc0964 1AC 0.03 fcrco084lAC 0.03 ncr5046 lAC 0.03 miobl8331AC 0.03 miob7550 1AC 0.03 fcrb8236lAC 0.03 fcrb2301 lAC 0.03 miob64301AC 0.03 fcrc2099 1AC 0.03 fcrc22221AC 0.03 mioa2261 lAC 0.03 fcrb84651AC 0.03 miod6068 lAC 0.03 miod41421AC 0.03 fcrb8891 lAC 0.03 seob3090lAC 0.03 fcr2999 lAC 0.03 fcrb4480lAC 0.03 seoc2614 1AC 0.04 seob4002lAC 0.03 seob6696 1AC 0.04 fcrb28061AC 0.03 seoa0065 lAC 0.04 ncrb7329lAC 0.03 fcrb8080 lAC 0.04 seoa60781AC 0.03 mioa5681 1AC 0.04 fcrc05291AC 0.03 seob5147 1AC 0.04 ncrc15311AC 0.03 seob3520 lAC 0.04 fcrb95281AC 0.03 ncr3465 1AC 0.04 ncrb08111AC 0.03 fcrb1724 lAC 0.04 seoa61441AC 0.03 fcrb2784 1AC 0.04 fcrc25361AC 0.03 fcrb8495 1AC 0.04 miob3320lAC 0.03 mioa7976 lAC 0.04 ncrc0583lAC 0.03 ncrc9166 1AC 0.04 miob89891AC 0.03 seob8388 lAC 0.04 miob29181AC 0.03 ncrc6981 1AC 0.04 mioc6075lAC 0.03 miob9495 lAC 0.04 ncr1428 lAC 0.03 mioc7952 1AC 0.04 seob40761AC 0.03 miod6090 1AC 0.04 seob4001lAC 0.03 ncr9039 1AC 0.04 ncrc6127lAC 0.03 fcrb1503 1AC 0.04 miob94631AC 0.03 ncrc1031 lAC 0.04 fcrb8432lAC 0.03 seoc3836 lAC 0.04 ncr1876 lAC 0.03 mioc4190 1AC 0.04 seob0344lAC 0.03 ncrc2128 lAC 0.04 a.~ ~ ::
","rriiod018''71A~~::.,~ ,.";~0~. ncrc9772 1B 1. 92179e-04 ",", "."~ 7"' 04, miob7308 lAC 0.04 fcrb9655 1B 2.35144e-04 ncrb0060 1AC 0.04 seob0992 1B 3.16052e-04 seoc4888 1AC 0.04 mioc1122 1B 3.83188e-04 seoa4518 lAC 0.04 fcrb8542 1B 4.62953e-04 fcr0187 lAC 0.04 miod0592 1B 4.62953e-04 fcrb6211 1AC 0.04 fcrb4890 1B 5.57395e-04 fcrc2008 lAC 0.04 fcrc1834 1B 5.57395e-04 ncr3568 1AC 0.04 fcrb4378 1B 6.10836e-04 ncrc1367 1AC 0.04 fcrb6896 1B 6.68842e-04 mioa0647 1AC 0.04 seob6379 1B 7.31751e-04 ncrc9355 lAC 0.04 ncrc6888 1B 8.73736e-04 fcrc2007 lAC 0.04 ncr8975 1B 9.53596e-04 seoc0924 1AC 0.04 seoa3408 1B 9.53596e-04 mioc2596 lAC 0.04 ncrcl567 1B 1.039929e-03 fcrb5662 lAC 0.04 seob8301 1B 1.039929e-03 miob6188 1AC 0.04 seob1319 1B 1.133188e-03 fcrb2376 lAC 0.04 fcrb1689 1B 1.23385e-03 mioa9154 lAC 0.04 fcrb3654 1B 1.23385e-03 rniob7638 1AC 0.04 fcrc5516 1B 1.23385e-03 fcr7114 lAC 0.04 ncrc9739 1B 1.342418e-03 fcr0206 1AC 0.04 fcr1337 1B 1.459426e-03 fcr5571 1AC 0.04 mioa9935 1B 1.459426e-03 ncr6108 1AC 0.04 ncrc5091 1B 1.459426e-03 ncr6893 lAC 0.04 seoa3639 1B 1.459426e-03 ncr9779 lAC 0.04 seoc0778 1B 1.459426e-03 seob0572 1AC 0.04 ncr8588 1B 1.585433e-03 hfcr4497 1AC 0.04 seoc0775 1B 1.585433e-03 ncr6415 lAC 0.04 ncrc4135 1B 1.661059e-03 ncrc0424 1AC 0.04 hfcr4485 1B 1.721032e-03 fcrb6236 1AC 0.04 ncrc9855 1B 1.721032e-03 seoa6152 1AC 0.04 ncrc9899 1B 1.721032e-03 ncrc3068 lAC 0.04 ncrcl892 1B 1.809776e-03 seob0999 lAC 0.04 miob8143 1B 1.866843e-03 miob9710 1AC 0.04 seoc1025 1B 1.866843e-03 ncrb8518 1AC 0.04 mioc2074 1B 2.023522e-03 mioa5202 1AC 0.04 miob8932 1B 2.191755e-03 mioc2348 lAC 0.04 ncr3811 1B 2.191755e-03 ncrc9739 lAC 0.04 seoa3422 1B 2.191755e-03 seoa5552 lAC 0.04 mioa1388 1B 2.317067e-03 seoa6510 lAC 0.04 ncr1780 1B 2.372265e-03 seoa7373 1AC 0.04 ncrc9700 1B 2.372265e-03 seob3313 lAC 0.04 seoa8399 1B 2.372265e-03 fcrb7981 lAC 0.04 seoa8556 1B 2.372265e-03 miod0340 lAC 0.04 fcrb0193 1B 2.565809e-03 seob1906 lAC 0.04 fcrb5850 1B 2.565809e-03 mioc3300 lAC 0.04 fcrc1298 1B 2.773181e-03 seoc2272 lAC 0.04 miob2877 1B 2.773181e-03 fcr4012 lAC 0.04 ncr7813 1B 2.773181e-03 seob3493 1AC 0.04 seoa1653 1B 2.773181e-03 seoc6182 1AC 0.04 seoa4485 1B 2.773181e-03 fcrc0493 lAC 0.04 fcrb2437 1B 2.950074e-03 fcrc4848 lAC 0.04 fcrb9959 1B 2.995213e-03 seob0755 lAC 0.04 seoa3429 1B 2.995213e-03 mioc2443 lAC 0.04 mioa5059 1B 3.192597e-03 ncrc9704 lAC 0.04 ncrc9637 1B 3.232776e-03 mioa4064 1B 2.04e-05 seoa9870 1B 3.232776e-03 ncrc5844 1B 2.6e-05 seob0879 1B 3.232776e-03 mioc0899 1B 6.45e-05 seoc1023 1B 3.232776e-03 fcrc6228 1B 9.2e-05 fcrb8187 1B 3.48678e-03 seob4752 1B 9.2e-05 miob9614 1B 3.48678e-03 fcr1997 1B 1.26897e-04 ncrc3936 1B 3.48678e-03 ncrc9712 1B 1.56467e-04 ncrc4780 1B 3.48678e-03 ncr7904 1B 1.63741e-04 seoa9777 1B 3.48678e-03 Ii .% ~
' ..,.Ii ...Hi:u~~~~~ 3~.~59413e-03seoa3516 1B 7.669588e-03 it...f ..
~~~seoc22~~4~91B

mioc0206 1B 3.758178e-03 seoa9709 lB 7.669588e-03 seob2169 1B 3.758178e-03 seob1411 1B 7.669588e-03 miod4066 1B 4.047962e-03 fcr1060 1B 8.207892e-03 ncrb8203 1B 4.35717e-03 fcrb3518 1B 8.207892e-03 seoa9389 1B 4.35717e-03 miob9671 1B 8.207892e-03 fcrc1745 1B 4.686884e-03 seoa7212 1B 8.207892e-03 mioa8970 1B 4.686884e-03 seoa9889 1B 8.207892e-03 miob8249 1B 4.686884e-03 mioa8946 1B 8.355812e-03 mioc3906 1B 4.686884e-03 miob3696 1B 8.778576e-03 ncrc9491 1B 4.686884e-03 ncrc9877 1B 8.954837e-03 ncr4647 1B 5.038229e-03 fcr4214 1B 9.383215e-03 seob0089 1B 5.038229e-03 fcrb4543 1B 9.383215e-03 seob0154 1B 5.038229e-03 fcrb6028 1B 9.383215e-03 seob7946 1B 5.038229e-03 fcrb6432 1B 9.383215e-03 seob9092 1B 5.038229e-03 mioc6973 1B 9.383215e-03 ncr0212 1B 5.050847e-03 seoa0536 1B 9.383215e-03 ncrc3344 1B 5.050847e-03 seoa4452 1B 9.383215e-03 fcrb1503 1B 5.412379e-03 seoa5662 1B 9.383215e-03 fcrc0351 1B 5.412379e-03 seob0047 1B 9.383215e-03 mioc1910 1B 5.412379e-03 mioc1205 1B 9.969032e-03 ncrc0585 1B 5.412379e-03 fcrc6566 1B O.Ol ncrc3735 1B 5.412379e-03 mioa4229 1B 0.01 ncrc5653 1B 5.412379e-03 miob2466 1B 0.01 ncrc9704 1B 5.412379e-03 mioc1125 1B 0.01 seob2797 1B 5.438596e-03 ncr7532 1B 0.01 miob8947 1B 5.658293e-03 ncrc9401 1B 0.01 ncr76'72 1B 5.810553e-03 seob0755 1B 0.01 seoa3359 1B 5.810553e-03 seob4270 1B 0.01 mioa9831 1B 5.852052e-03 miod6437 1B 0.01 ncr0075 1B 5.89206e-03 ncr5488 1B 0.01 fcrb8877 1B 6.234019e-03 ncrc5072 1B 0.01 mioa2377 1B 6.234019e-03 ncr0836 1B 0.01 miob3953 1B 6.234019e-03 fcrb2704 1B 0.01 miob8657 1B 6.234019e-03 fcrb4727 1B 0.01 ncrc2080 1B 6.234019e-03 fcrc2651 1B 0.01 ncrc3520 1B 6.234019e-03 fcrc4161 lB 0.01 seoa1100 1B 6.234019e-03 hfcr0285 1B 0.01 ncrc0174 1B 6.292611e-03 mioa9555 1B 0.01 fcr3559 1B 6.684094e-03 mioc1028 1B 0.01 mioa0528 1B 6.684094e-03 ncrc3799 1B 0.01 ncrc3908 1B 6.684094e-03 ncrc5088 1B 0.01 ncrc9633 1B 6.684094e-03 seoa6314 1B 0.01 seoa0740 1B 6.684094e-03 seoa6393 1B 0.01 seoa2381 1B 6.684094e-03 seoc3876 1B 0.01 mioa8778 1B 6.761728e-03 fcrc0654 1B 0.01 fcrc5690 1B 7.162145e-03 fcr4471 1B 0.01 mioa2185 1B 7.162145e-03 fcrb4345 1B 0.01 ncrc0667 1B 7.162145e-03 fcrb5214 1B 0.01 ncrc5949 1B 7.162145e-03 fcrb7593 1B 0.01 ncrc8881 1B 7.162145e-03 fcrc2745 1B 0.01 seoa0469 1B 7.162145e-03 mioa1626 1B 0.01 seoa7917 1B 7.162145e-03 mioa9891 1B 0.01 miob3234 1B 7.260913e-03 ncr0615 1B 0.01 fcrb1990 1B 7.669588e-03 ncrc0807 1B 0.01 fcrc0637 1B 7.669588e-03 seoa5157 1B 0.01 fcrc4333 1B 7.669588e-03 seoa5685 1B 0.01 fcrc4360 1B 7.669588e-03 seoa5911 1B 0.01 mioa8851 1B 7.669588e-03 seoa9566 1B 0.01 miob9121 1B 7.669588e-03 seob7500 1B 0.01 mioc1438 1B 7.669588e-03 seoa1540 1B 0.01 ncrc0936 1B 7.669588e-03 fcr4743 1B 0.01 ncrc9867 1B 7.669588e-03 fcrb2334 1B 0.01 seoa0256 1B 7.669588e-03 fcrc6877 1B 0.01 'L..F ,....ti ..,.. %i...;:
a...il 0. 01 ~~~~~~seoa6620 1B 0.
..,..i O1 ..' ....
ii '~f miob9185 1B~~~

mioc3930 1B 0.01 seob2987 1B 0.01 ncrb8821 1B 0.01 seob3485 1B 0.01 ncrc3953 1B 0.01 seob4197 1B 0.01 ncrc5760 1B 0.01 seob7729 1B 0.01 seoa0045 1B 0.01 seob9485 1B 0.01 seoa6654 1B 0.01 seoc4316 1B 0.01 seob1399 1B 0.01 fcrc2014 1B 0.01 ncrc2685 1B 0.01 hfcr1189 1B 0.01 seobl345 1B 0.01 hfcr2275 1B 0.01 fcrb6359 1B 0.01 ncr4975 1B 0.01 hfcr4063 1B 0.01 ncrc5631 1B 0.01 mioa5586 1B 0.01 ncrc9944 1B 0.01 mioa6999 1B 0.01 seoa0429 1B 0.01 mioa9581 1B 0.01 seoa3555 1B 0.01 ncr4946 1B 0.01 seob1748 1B 0.01 seoa3717 1B 0.01 seoc3792 1B 0.01 seoa5554 1B 0.01 fcrb2380 1B 0.01 seoa5577 1B 0.01 fcrb5267 1B 0.01 seob0201 1B 0.01 fcrb7510 1B 0.01 seob9152 1B 0.01 fcrc1181 1B 0.01 ncr8153 1B 0.01 hfcr3209 1B 0.01 fcrb2849 1B 0.01 mioa3331 1B 0.01 mioa7239 1B 0.01 mioa6738 1B 0.01 miod3920 1B 0.01 miob7373 1B 0.01 ncrb1880 1B 0.01 miob9065 1B 0.01 ncrb8343 1B 0.01 ncr3034 1B 0.01 ncrc9557 1B 0.01 ncrc9159 1B 0.01 seob4192 1B 0.01 seoa1644 1B 0.01 seob5726 1B 0.01 seoa3105 1B 0.01 seob6206 1B 0.01 seoa3662 1B 0.01 fcr3654 1B 0.01 seoa3811 1B 0.01 fcrb2198 1B 0.01 seoa9916 1B 0.01 fcrb2713 1B 0.01 seob0497 1B 0.01 fcrc4848 1B 0.01 seob2959 1B 0.01 mioa9649 1B 0.01 seob9241 1B 0.01 mioc2173 1B 0.01 seoc3443 1B 0.01 ncr0847 1B 0.01 ncr2182 1B 0.02 ncr3614 1B 0.01 fcrb2162 1B 0.02 ncrb8385 1B 0.01 mioa6832 1B 0.02 ncrc9428 1B 0.01 miod5703 1B 0.02 seob1061 1B 0.01 mioc7084 1B 0.02 seob1808 1B 0.01 fcrbl898 1B 0.02 seoc5039 1B 0.01 fcrb2044 1B 0.02 fcrb1529 1B 0.01 fcrb4226 1B 0.02 fcrb5702 1B 0.01 fcrb6868 1B 0.02 mioa1097 1B 0.01 mioc6417 1B 0.02 miod6731 1B 0.01 ncrc2119 1B 0.02 seob4108 1B 0.01 ncrc4444 1B 0.02 seoc0924 1B 0.01 seoa5234 1B 0.02 mioa4552 1B 0.01 seobl322 1B 0.02 ncr0291 1B 0.01 seob4165 1B 0.02 fcrb4985 1B 0.01 seob6856 1B 0.02 fcrb6033 1B 0.01 seob7346 1B 0.02 fcrb9963 1B 0.01 seob9145 1B 0.02 miob8391 1B 0.01 seoc1406 1B 0.02 mioc0090 1B 0.01 hfcr3494 1B 0.02 mioc1126 1B 0.01 fcrb4067 1B 0.02 ncr3112 1B 0.01 fcrb5928 1B 0.02 ncrb3980 1B 0.01 fcrc1115 1B 0.02 ncrc3713 1B 0.01 fcrc5402 1B 0.02 ncrc4089 1B 0.01 hfcr0521 1B 0.02 ncrc6996 1B 0.01 miob2341 1B 0.02 seoa4040 1B 0.01 mioc2039 1B 0.02 mtioc5740F~~~~im~~~~lB~~~~~~~~~~~0~.~0~2 ncrc9044 1B 0.02 mioc57511B 0.02 seoa3556 1B 0.02 miod33061B 0.02 seoa4460 1B 0.02 ncr9140 1B 0.02 seob1538 1B 0.02 ncrc06401B 0.02 seob3493 1B 0.02 ncrc57621B 0.02 seob6386 1B 0.02 seob18441B 0.02 ncrc2463 1B 0.03 seob80651B 0.02 seob3317 1B 0.03 seob81291B 0.02 fcr5625 1B 0.03 fcrb74711B 0.02 fcrb3330 1B 0.03 fcrb23501B 0.02 fcrb6734 1B 0.03 fcrb37041B 0.02 fcrcl965 1B 0.03 fcrb42381B 0.02 fcrc2808 1B 0.03 hfcr59051B 0.02 fcrc6305 lB 0.03 mioa33671B 0.02 miob3308 1B 0.03 mioa98211B 0.02 miob6438 1B 0.03 miob43081B 0.02 ncrb0262 1B 0.03 mioc39621B 0.02 ncrc8873 1B 0.03 mioc73811B 0.02 ncrc9438 1B 0.03 ncrb86891B 0.02 seoa4324 1B 0.03 ncrc68711B 0.02 seob1268 1B 0.03 seob21951B 0.02 seob4075 1B 0.03 seob83111B 0.02 seob6131 1B 0.03 seoc11751B 0.02 ncrb0060 1B 0.03 miod06251B 0.02 fcr0997 1B 0.03 ncrc32571B 0.02 fcrb6031 1B 0.03 seoa28191B 0.02 fcrb6102 1B 0.03 fcrb81191B 0.02 fcrb7742 1B 0.03 miob63721B 0.02 fcrb8133 1B 0.03 mioc03711B 0.02 fcrc6382 1B 0.03 ncrc03831B 0.02 fcrc6997 1B 0.03 ncrc38551B 0.02 mioa6428 1B 0.03 ncrc64231B 0.02 miob2227 1B 0.03 ncrc94641B 0.02 miob3938 1B 0.03 seoa97051B 0.02 miob8802 1B 0.03 seob00851B 0.02 mioc0181 1B 0.03 seob35131B 0.02 mioc4603 1B 0.03 seob41911B 0.02 miod4083 1B 0.03 seob52091B 0.02 ncr3435 1B 0.03 seob66701B 0.02 ncrc3856 1B 0.03 fcrb60121B 0.02 ncrc4371 1B 0.03 fcrc58981B 0.02 seoa8543 1B 0.03 hfcr26161B 0.02 seob0265 1B 0.03 hfcr36601B 0.02 seob3367 1B 0.03 mioa61021B 0.02 seob5319 1B 0.03 mioa69131B 0.02 fcr1464 1B 0.03 mioc43181B 0.02 ncr1550 1B 0.03 miod70811B 0.02 fcr0135 1B 0.03 ncr3037 1B 0.02 fcrb3074 1B 0.03 ncrc29591B 0.02 fcrb8485 1B 0.03 seoa44361B 0.02 fcrc0166 1B 0.03 seoa59771B 0.02 fcrc4669 1B 0.03 seob36971B 0.02 fcrc5504 1B 0.03 seob88391B 0.02 fcrc6345 1B 0.03 miod34171B 0.02 hfcr0734 1B 0.03 fcr3620 1B 0.02 miob0496 1B 0.03 fcr5190 1B 0.02 miob1774 1B 0.03 fcrb38971B 0.02 miob3690 1B 0.03 fcrb44291B 0.02 mioc1763 1B 0.03 fcrb74951B 0.02 mioc3127 1B 0.03 hfcr00451B 0.02 mioc4731 1B 0.03 hfcr04891B 0.02 ncr0097 1B 0.03 mioa39451B 0.02 ncr5149 1B 0.03 ncrc45751B 0.02 ncr6343 1B 0.03 ricr~83'$~2~"..........1.~_"...,0..:Ø3..fcrb6279 1B 0.04 ncrcllll 1B0.03 hfcr6634 1B 0.04 ncrc2763 1B0.03 mioa0294 1B 0.04 ncrc3435 1B0.03 mioa9630 1B 0.04 ncrc4323 1B0.03 miob8259 1B 0.04 ncrc9793 1B0.03 miob9020 1B 0.04 seoa4587 1B0.03 ncr1428 1B 0.04 seoa9724 1B0.03 ncrb2247 1B 0.04 seob0288 1B0.03 ncrc2472 1B 0.04 seoc0866 1B0.03 ncrc3391 1B 0.04 ncrc4287 1B0.03 seoa2936 1B 0.04 ncr2273 1B0.03 seob6272 1B 0.04 hfcr0225 1B0.03 seoc0651 1B 0.04 fcrb2452 1B0.03 fcr0018 1B 0.04 fcrb4515 1B0.03 seoa7408 1B 0.04 fcrb6808 1B0.03 miod1846 1B 0.04 fcrb7803 1B0.03 fcrb3298 1B 0.04 fcrc6084 lB0.03 fcrb4799 1B 0.04 miob3042 1B0.03 fcrb5305 1B 0.04 mioc0301 1B0.03 fcrb5709 1B 0.04 mioc4280 1B0.03 hfcr2536 1B 0.04 mioc7331 1B0.03 miob2432 1B 0.04 mioc7910 1B0.03 miob2634 1B 0.04 ncr3834 1B0.03 miob4037 1B 0.04 ncrb8752 lB0.03 miod3592 1B 0.04 ncrc0423 1B0.03 ncr9108 1B 0.04 ncrc3596 1B0.03 ncrb6640 1B 0.04 ncrc3777 1B0.03 ncrb8437 1B 0.04 ncrc3895 1B0.03 ncrc4597 1B 0.04 seob0044 1B0.03 seoa2448 1B 0.04 seob3204 1B0.03 seob0938 1B 0.04 seob3699 1B0.03 seob3517 1B 0.04 seob7929 1.B0.03 seob7419 1B 0.04 seoc7281 1B0.03 seoc1593 1B 0.04 fcrb5867 1B0.03 ncrb0,602 1B 0.04 miob5675 1B0.03 ncrc4994 1B 0.04 seob7575 1B0.03 fcrb2592 1B 0.04 fcrb6436 1B0.03 fcrb5198 1B 0.04 fcrb7440 1B0.03 fcrb6191 lB 0.04 fcrc2670 1B0.03 fcrb9659 1B 0.04 fcrc5771 1B0.03 hfcr4741 1B 0.04 hfcr1141 1B0.03 mioa4542 1B 0.04 hfcr6043 1B0.03 mioa6093 1B 0.04 miob1506 1B0.03 mioa9294 1B 0.04 miob8531 1B0.03 miob2833 1B 0.04 miob9130 1B0.03 miod4686 1B 0.04 miocl783 1B0.03 miod5114 1B 0.04 mioc8016 1B0.03 miod7440 1B 0.04 miod2977 1B0.03 ncr0420 1B 0.04 ncr2160 1B0.03 ncrb4182 1B 0.04 ncr2288 1B0.03 ncrc3598 1B 0.04 ncr3960 1B0.03 ncrc4757 1B 0.04 ncrb8605 1B0.03 ncrc5232 1B 0.04 ncrc3171 1B0.03 seoa2765 1B 0.04 ncrc5230 1B0.03 seoa6377 1B 0.04 seoa0040 1B0.03 seoa7383 1B 0.04 seoa2641 1B0.03 seob8194 1B 0.04 seob0168 1B0.03 seob8333 1B 0.04 seob2283 1B0.03 seoc0843 1B 0.04 seoc2131 1B0.03 fcrb3868 1B 0.04 fcrb5204 1B0.03 fcrb0044 1B 0.04 seob5213 1B0.03 miod1532 1B 0.04 ncr6072 1B0.03 miod5301 1B 0.04 fcrb1562 1B0.04 fcrb5202 1B 0.04 ~~fcrb932~4.~~1B "D":'0'4 fcrc5418 lC 0.
' 01 fcrb96361B 0.04 mioa4674 1C 0.01 mioa89521B 0.04 mioc3906 1C 0.01 miob56461B 0.04 miod3421 1C 0.01 miob91631B 0.04 ncr7904 1C 0.01 miob93931B 0.04 ncrb1438 1C 0.01 mioc10601B 0.04 ncrc6697 IC 0.01 mioc19401B 0.04 seob8212 1C 0.01 mioc48881B 0.04 ncrb1677 1C 0.01 mioc49781B 0.04 hfcr0338 1C 0.01 ncr3598 1B 0.04 fcrb5926 1C 0.01 ncrc26751B 0.04 fcrc0456 1C 0.01 ncrc36241B 0.04 miob0167 1C 0.01 ncrc46541B 0.04 miob4570 1C 0.01 seoa01351B 0.04 miob8515 1C 0.01 seoa48021B 0.04 ncr4946 1C 0.01 seoa71261B 0.04 ncr5613 1C 0.01 seob13621B 0.04 ncrc6778 1C 0.01 seob50691B 0.04 seob1808 1C 0.01 seob75571B 0.04 fcrb8202 1C 0.01 ncr2995 1B 0.04 miob9901 1C 0.01 miod23231B 0.04 miod0777 1C 0,01 fcrb73241C 1.44645e-04 ncrc0421 1C 0.01 mioc74441C 1.162439e-03 ncrc7127 1C 0.01 miod34171C 1.503841e-03 seoa6364 1C 0.01 seoc42881C 1.503841e-03 seob2938 1C 0.01 ncrb33291C 1.928847e-03 seob5032 1C 0.01 miob30721C 2.453741e-03 seoc1508 1C 0.01 ncrc95571C 2.453741e-03 fcr4444 1C 0.01 seob95521C 2.453741e-03 fcrb2200 1C 0.01 seoa55771C 3.097092e-03 fcrb3074 1C 0.01 fcr5369 1C 3.879912e-03 fcrb7036 1C 0.01 .fcrc07751C 3.879912e-03 mioa1354 1C 0.01 hfcr02851C 3.879912e-03 mioa5059 1C 0.01 mioc02381C 3.879912e-03 mioa6034 1C 0.01 mioc74711C 3.879912e-03 mioa7957 1C 0.01 seoa28191C 3.879912e-03 miob5675 1C 0.01 seob29371C 3.879912e-03 mioc1425 1C 0.01 fcr2293 1C 4.825803e-03 ncr6072 1C 0.01 mioc25961C 4.825803e-03 ncrc6242 1C 0.01 mioc30401C 4.825803e-03 seoa8195 1C 0.01 ncr3869 1C 4.825803e-03 seob1757 1C 0.01 ncrc67561C 4.825803e-03 miod6947 1C 0.01 seob54581C 4.825803e-03 ncrb6327 1C 0.01 seob56581C 4.825803e-03 fcr4328 1C 0.02 fcrb50161C 5.961096e-03 fcrb2252 1C 0.02 mioa97921C 5.961096e-03 fcrb3476 1C 0.02 fcr2276 IC 7.314953e-03 fcrb7340 1C 0.02 fcrb57231C 7.314953e-03 hfcr2367 1C 0.02 miod53491C 7.314953e-03 mioa6476 1C 0.02 ncrc57801C 7.314953e-03 mioa6734 1C 0.02 seob01331C 7.314953e-03 mioa9630 1C 0.02 seoc03941C 7.314953e-03 miob0764 1C 0.02 miod53011C 7.453402e-03 miob3320 1C 0.02 seob98711C 7.453402e-03 mioc1524 1C 0.02 miob24481C 8.919456e-03 mioc3962 1C 0.02 miob60871C 8.919456e-03 miod0456 1C 0.02 mioc14401C 8.919456e-03 miod5682 1C 0.02 ncr1387 1C 8.919456e-03 ncr0212 1C 0.02 seoa71781C 8.919456e-03 ncr8588 1C 0.02 seoa99971C 8.919456e-03 ncrb2092 1C 0.02 seob05641C 9.170707e-03 ncrb4339 1C 0.02 fcrb62111C 0.01 ncrb5704 1C 0.02 fcrb64641C 0.01 ncrcl665 lC 0.02 17~

ncrc68711C 0-.02 fcrb9680 1C 0.03 seoa38631C 0.02 fcrc0180 1C 0.03 seoa43271C 0.02 fcrc2082 lC 0.03 seoa76471C 0.02 fcrc4734 1C 0.03 seob31821C 0.02 fcrc6228 1C 0.03 seob77471C 0.02 fcrc6888 1C 0.03 seoc61~91C 0.02 fcrc6970 1C 0.03 fcrb14201C 0.02 mioa4196 1C 0.03 fcrb82361C 0.02 mioa6832 1C 0.03 fcrb9450lC 0.02 mioa9033 1C 0.03 mioa45521C 0.02 mioa9891 1C 0.03 mioa89121C 0.02 miob2918 1C 0.03 miob33581C 0.02 miob3953 lC 0.03 mioc02761C 0.02 miob9336 1C 0.03 mioc2561lC 0.02 miod6938 1C 0.03 mioc25921C 0.02 ncr2581 1C 0.03 mioc56331C 0.02 ncr7292 1C 0.03 miod06251C 0.02 ncr8234 1C 0.03 miod25701C 0.02 ncrb6394 1C 0.03 ncr3163 1C 0.02 ncrc0413 1C 0.03 ncr4656 1C 0.02 ncrc5959 1C 0.03 ncrb2400lC 0.02 ncrc6795 1C 0.03 ncrb83851C 0.02 ncrc9642 1C 0.03 ncrc91661C 0.02 seoa1615 1C 0.03 seoa61061C 0.02 seoa2042 1C 0.03 seobll441C 0.02 seoa4802 1C 0.03 seob39041C 0.02 seoa6497 1C 0.03 seob43631C 0.02 seoa8501 1C 0.03 seob78861C 0.02 seob0321 1C 0.03 fcr1098 1C 0.03 seob4160 1C 0.03 fcr1855 1C 0.03 seob4782 1C 0.03 fcr2088 1C 0.03 seob5711 1C 0.03 fcrb23801C 0.03 seob6836 1C 0.03 fcrb27541C 0.03 seob9241 1C 0.03 fcrb75931C 0.03 fcrb5087 1C 0.03 fcrc47221C 0.03 fcr4503 1C 0.04 fcrc48161C 0.03 fcrb3024 1C 0.04 fcrc48411C 0.03 fcrb3217 1C 0.04 hfcr28951C 0.03 fcrb3920 1C 0.04 hfcr34861C 0.03 fcrc4968 1C 0.04 hfcr35141C 0.03 fcrc5160 1C 0.04 miob02021C 0.03 fcrc6609 1C 0.04 miob22271C 0.03 hfcr2314 lC 0.04 miob92841C 0.03 mioa2475 1C 0.04 miob97141C 0.03 mioa3290 1C 0.04 mioc11071C 0.03 miob8711 1C 0.04 mioc34921C 0.03 miob9052 1C 0.04 miod13581C 0.03 mioc3413 1C 0.04 miod22321C 0.03 ncr0808 1C 0.04 miod69881C 0.03 ncr2370 1C 0.04 ncr0075 1C 0.03 ncr2994 1C 0.04 ncrc16531C 0.03 ncr3718 1C 0.04 ncrc23771C 0.03 ncr8481 2C 0.04 ncrc69201C 0.03 ncrb0328 1C 0.04 seoa60701C 0.03 ncrb6557 1C 0.04 seob13991C 0.03 ncrc0393 1C 0.04 seob26611C 0.03 ncrc0442 1C 0.04 seob29501C 0.03 ncrc0576 1C 0.04 seob60201C 0.03 ncrcll93 1C 0.04 seob90011C 0.03 ncrc6407 1C 0.04 fcrb87621C 0.03 seoa0913 1C 0.04 fcr4566 1C 0.03 seoa1856 1C 0.04 fcrb42751C 0.03 seoa2178 1C 0.04 fcrb42871C 0.03 seoa3007 1C 0.04 seoa31061C 0.04 seoc6182 1D 1.703326e-03 seoa33591C 0.04 fcr3559 1D 2.051133e-03 seoa83991C 0.04 fcrc2014 1D 2.051133e-03 seob31511C 0.04 ncr7904 1D 2.051133e-03 seob35031C 0.04 ncrcl049 1D 2.459013e-03 seob55741C 0.04 seoa3847 1D 2.459013e-03 seoc25041C 0.04 seob4270 1D 2.459013e-03 seoc44161C 0.04 seoc0775 1D 2.459013e-03 ncr1876 1C 0.04 mioc7444 1D 2.935338e-03 ncr6401 1C 0.04 ncrb8665 1D 2.935338e-03 fcrb30161C 0.04 seoa4056 1D 2.935338e-03 fcrb81211C 0.04 seoa6620 1D 2.935338e-03 fcrb95831C 0.04 fcrb2633 1D 3.489334e-03 fcrc43601C 0.04 hfcr6370 1D 3.489334e-03 hfcr26861C 0.04 mioa6913 1D 3.489334e-03 mioa06011C 0.04 ncrc0544 1D 3.489334e-03 mioa40771C 0.04 seob0755 1D 3.489334e-03 mioa44841C 0.04 seob9871 1D 4.12035e-03 mioa47531C 0.04 ncrb1420 1D 4.131123e-03 mioa89521C 0.04 ncrc8909 1D 4.131123e-03 mioa90671C 0.04 seoa3&39 1D 4.131123e-03 miob32571C 0.04 seoa5444 1D 4.131123e-03 miob35521C 0.04 seoa9817 1D 4.131123e-03 miob41261C 0.04 ncr1387 1D 4.414279e-03 miob50161C 0.04 fcrb6005 1D 4.871761e-03 miob94411C 0.04 hfcr6501 1D 4.871761e-03 mioc01811C 0.04 mioc1783 1D 4.871761e-03 mioc35231C 0.04 ncr3843 1D 4.871761e-03 mioc43311C 0.04 ncrc1003 1D 4.871761e-03 mioc87501C 0.04 ncrc6623 1D 4.871761e-03 ncr0335 1C 0.04 fcrb3169 lD 5.723282e-03 ncr2408 1C 0.04 fcrb3618 1D 5.723282e-03 ncr3141 1C 0.04 hfcr2314 1D 5.723282e-03 ncr5027 1C 0.04 mioa4845 1D 5.723282e-03 ncr5065 1C 0.04 miob9905 1D 5.723282e-03 ncr8975 1C 0.04 mioc1940 1D 5.723282e-03 ncr9378 1C 0.04 mioc4190 1D 5.723282e-03 ncrc04451C 0.04 mioc6320 1D 5.723282e-03 ncrcl3611C 0.04 ncr6141 1D 5.723282e-03 ncrc15671C 0.04 ncrc6811 1D 5.723282e-03 ncrc51501C 0.04 fcr1997 1D 6.698726e-03 seoa05011C 0.04 fcrb1420 1D 6.698726e-03 seoa14601C 0.04 seoa2641 1D 6.698726e-03 seoa57421C 0.04 seoa7669 1D 6.698726e-03 seoa75301C 0.04 seoa8543 1D 6.698726e-03 seob00381C 0.04 seob0154 1D 6.698726e-03 seob02001C 0.04 fcrb9655 1D 7.812173e-03 seob41271C 0.04 hfcr0285 1D 7.812173e-03 seob6017lC 0.04 mioa4552 1D 7.812173e-03 seob91521C 0.04 mioa9821 1D 7.812173e-03 seoc16281C 0.04 mioc5270 1D 7.812173e-03 hfcr05211D 1.21e-05 ncrb4319 1D 7.812173e-03 miod44641D 3.22085e-04 fcrb2452 1D 9.078758e-03 seoc50391D 4.04275e-04 fcrb6939 1D 9.078758e-03 fcrb42751D 6.25978e-04 hfcr3209 1D 9.078758e-03 seob00891D 9.48395e-04 miod0777 1D 9.078758e-03 ncrc08071D 1.158413e-03 ncr4648 1D 9.078758e-03 seoa61521D 1.158413e-03 seoa2428 1D 9.078758e-03 fcrb55501D 1.408023e-03 seoa3852 1D 9.078758e-03 ncrb83851D 1.408023e-03 fcr1879 1D 0.01 ncrc96371D 1.408023e-03 fcrb6896 1D 0.01 ncrc99101D 1.408023e-03 mioa0252 1D 0.01 seoa55771D 1.703326e-03 mioa6476 1D 0.01 seob43631D 1.703326e-03 miod3591 1D 0.01 ncrb8605~~~1D~ '~":d~~ ncrc9286 1D 0.01 ncrc20801D 0.01 seoa3863 lD 0.01 seoa98701D 0.01 seoa4053 1D 0.01 ncr3237 1D 0.01 seob0442 1D 0.01 fcr3861 1D 0.01 seob0650 1D 0.01 fcr1562 1D 0.01 seob6279 1D 0.01 fcrc64701D 0.01 seoc0924 1D 0.01 mioa08201D 0.01 seoc4161 1D 0.01 mioa43181D 0.01 seoa8993 1D 0.01 mioc03021D 0.01 ncr0527 1D 0.02 miod74291D 0.01 fcrb7780 1D 0.02 ncrc02621D 0.01 mioa0494 1D 0.02 ncrc30451D 0.01 mioa3945 1D 0.02 seoa40401D 0.01 mioa5511 1D 0.02 seoa75461D 0.01 mioc2750 1D 0.02 seob01681D 0.01 mioc3066 1D 0.02 seobl5261D 0.01 seoa6377 1D 0.02 seob63791D 0.01 seoa7897 1D 0.02 miod19091D 0.01 seoa8556 1D 0.02 fcrb15031D 0.01 seob1285 1D 0.02 fcrb57261D 0.01 seob5219 1D 0.02 fcrc10431D 0.01 seob7765 1D 0.02 fcrc55161D 0.01 seoc0149 1D 0.02 mioa39631D 0.01 fcrb3592 1D 0.02 mioc15901D 0.01 fcrb5070 1D 0.02 mioc29971D 0.01 fcrc4360 1D 0.02 mioc52031D 0.01 hfcr2629 1D 0.02 miod41291D 0.01 mioc0999 1D 0.02 ncrb87211D 0.01 miod4629 1D 0.02 ncrc05391D 0.01 miod6467 1D 0.02 ncrc23941D 0.01 ncr3189 1D 0.02 seoa31021D 0.01 seoa1439 10 0.02 seoa48021D 0.01 seoa4518 1D 0.02 seoa6~.Z81D 0.01 seoa5253 1D 0.02 seob98721D 0.01 seoa8401 1D 0.02 seoc27231D 0.01 seob6272 1D 0.02 fcrb43211D 0.01 ncr0451 1D 0.02 fcrbl7311D 0.01 fcrb3192 1D 0.02 fcrb47881D 0.01 fcrb9856 1D 0.02 fcrb60121D 0.01 hfcr2890 1D 0.02 hfcr18111D 0.01 mioa3471 1D 0.02 mioa21851D 0.01 mioa8852 1D 0.02 mioc73311D 0.01 miob0764 1D 0.02 ncr0213 1D 0.01 miob9130 1D 0.02 ncr1523 1D 0.01 miod6324 1D 0.02 ncr4140 1D 0.01 ncr0335 1D 0.02 ncr4522 1D 0.01 ncr4946 1D 0.02 ncr6072 1D 0.01 ncrb3314 1D 0.02 ncr7532 1D 0.01 ncrb6394 1D 0.02 seoa31051D 0.01 ncrc4597 1D 0.02 seoa83001D 0.01 seoa8486 1D 0.02 seob07521D 0.01 seob0085 1D 0.02 seob53191D 0.01 seob1319 1D 0.02 seob62061D 0.01 seob1345 1D 0.02 seob83011D 0.01 seob5528 1D 0.02 fcr0997 1D 0.01 seob6703 1D 0.02 fcr2195 1D 0.01 mioc0899 1D 0.02 fcrb15801D 0.01 ncrc2529 1D 0.02 fcrb57961D 0.01 fcr5836 1D 0.03 mioa46741D 0.01 fcrb6715 1D 0.03 miod69381D 0.01 fcrb6890 1D 0.03 ncr4647 1D 0.01 fcrb9401 1D 0.03 ncrc08631D 0.01 fcrc0597 1D 0.03 ncrc37771D 0.01 fcrc2096 1D 0.03 1~1 ~fcrc61~0~8~~1D~ 0.03 ~~ ncrc0576 1D 0.03 hfcr33751D 0.03 ncrc6846 1D 0.03 hfcr66341D 0.03 ncrc9232 1D 0.03 mioa36201D 0.03 seoa5691 1D 0.03 mioa40641D 0.03 seob0265 1D 0.03 miob33081D 0.03 seob5213 1D 0.03 miob87021D 0.03 seoc1484 1D 0.03 mioc30421D 0.03 seocl664 1D 0.03 mioc37461D 0.03 miod7351 1D 0.04 mioc40641D 0.03 ncr6401 1D 0.04 miod07081D 0.03 fcr1346 1D 0.04 miod40661D 0.03 fcr2218 1D 0.04 ncr1476 1D 0.03 fcrb2208 1D 0.04 ncr3713 1D 0.03 fcrb3134 1D 0.04 ncrb12241D 0.03 fcrb4271 1D 0.04 ncrc21101D 0.03 fcrb5219 1D 0.04 ncrc33441D 0.03 fcrc0637 1D 0.04 seob27971D 0.03 fcrcl974 1D 0.04 seob29531D 0.03 hfcr0400 1D 0.04 seob43331D 0.03 mioa3528 1D 0.04 seob68531D 0.03 mioa4196 1D 0.04 seoc07801D 0.03 mioa8773 1D 0.04 seoc25891D 0.03 mioa8851 1D 0.04 seoc42881D 0.03 mioa9492 1D 0.04 seoc50101D 0.03 mioa9604 1D 0.04 mioc32961D 0.03 miob5646 1D 0.04 seoc34871D 0.03 mioc0090 1D 0.04 fcrb23501D 0.03 mioc1028 1D 0.04 fcrb36291D 0.03 mioc1088 1D 0.04 fcrb51811D 0.03 mioc3671 1D 0.04 fcrb97961D 0.03 miod4083 1D 0.04 miob32341D 0.03 ncr8588 1D 0.04 mioc43181D 0.03 ncrb6581 1D 0.04 mioc48881D 0.03 ncrc0413 1D 0.04 mioc57401D 0.03 ncrc1765 1D 0.04 mioc63741D 0.03 seoa3230 1D 0.04 mioc74211D 0.03 seoa7373 1D 0.04 miod05921D 0.03 seob0782 1D 0.04 miod48951D 0.03 seob3493 1D 0.04 ncr3297 1D 0.03 seob5478 1D 0.04 ncr3316 1D 0.03 seob6131 1D 0.04 ncrb21311D 0.03 seoc2131 1D 0.04 ncrc06461D 0.03 seoc2510 1D 0.04 ncrc52071D 0.03 seoc7281 1D 0.04 ncrc62421D 0.03 mioc1205 1D 0.04 ncrc95571D 0.03 ncr1526 1D 0.04 seoa11171D 0.03 seoc2249 1D 0.04 seoa27441D 0.03 fcr3664 1D 0.04 seoa37011D 0.03 fcr5350 1D 0.04 seoa41321D 0.03 fcrb1399 1D 0.04 seob65841D 0.03 fcrb5850 1D 0.04 seoc14801D 0.03 fcrc0795 1D 0.04 fcrb38571D 0.03 fcrc2346 1D 0.04 fcrb66391D 0.03 hfcr6611 1D 0.04 fcrb72371D 0.03 mioa1882 1D 0.04 fcrb73211D 0.03 miob2668 1D 0.04 mioa50851D 0.03 miob3531 1D 0.04 mioa67311D 0.03 miob5016 1D 0.04 miob87111D 0.03 miob8418 1D 0.04 mioc49941D 0.03 mioc0902 1D 0.04 ncr0509 1D 0.03 mioc2662 1D 0.04 ncrb20921D 0.03 mioc3663 1D 0.04 ncrb41821D 0.03 mioc8063 1D 0.04 ncrb75161D 0.03 miod3327 1D 0.04 miod6731ZL ~0~.04~~ miod7461 1E 3.68556e-04 -ncr0212 1D 0.04 fcr4699 1E 4.06485e-04 ncr0673 1D 0.04 fcrb3237 1E 4.06485e-04 ncrb83321D 0.04 fcrb6431 1E 4.06485e-04 ncrc08491D 0.04 fcrc5142 1E 4.06485e-04 ncrc66971D 0.04 ncr4545 1E 4.9315e-04 ncrc70401D 0.04 seoa5253 1E 4.9315e-04 ncrc89491D 0.04 fcrb3244 1E 5.42475e-04 seoa32451D 0.04 fcrb4345 1E 5.42475e-04 seoa59111D 0.04 fcrc0529 1E 5.42475e-04 seob83861D 0.04 hfcr3149 1E 5.42475e-04 seob86391D 0.04 ncr3527 1E 5.96221e-04 seoc10091D 0.04 ncrc2273 1E 5.96221e-04 seoc25181D 0.04 seob8311 1E 5.96221e-04 seoc52091D 0.04 seob9872 1E 5.96221e-04 fcr5509 1E 5.52e-06 fcr1328 1E 6.54736e-04 miob97481E 6.23e-06 fcrb2993 1E 6.54736e-04 ncr4140 1E 8.29e-06 fcrc6976 1E 6.54736e-04 miob24921E 1.23e-05 mioa6091 1E 6.54736e-04 miodl6751E 1.81e-05 seoa5234 1E 6.54736e-04 mioc11071E 2.05e-05 mioa4564 1E 7.18387e-04 fcrc06721E 2.63e-05 miod7414 1E 7.18387e-04 ncrb20911E 2.97e-05 seob2195 1E 7.18387e-04 seob53791E 3.35e-05 fcrb3870 1E 7.8757e-04 miod56511E 5.38e-05 hfcr3922 1E 7.8757e-04 fcrb47991E 6.03e-05 miobl115 1E 7.8757e-04 fcrc08391E 6.03e-05 ncrc2670 1E 7.8757e-04 seobl7661E 6.03e-05 fcrb3135 1E 8.62701e-04 fcr4084 1E 7.58e-05 fcrb6225 IE 8.62701e-04 fcrb95691E 7.58e-05 fcrb6808 1E 8.62701e-04 mioc28721E 8.48e-05 mioc2152 1E 8.62701e-04 miod74401E 8.48e-05 mioc2667 1E 8.62701e-04 seoa48021E 8.48e-05 miod3347 1E 8.62701e-04 ncr5168 1E 1.31557e-04 seob6446 1E 8.62701e-04 miob39821E 1.46458e-04 fcrb6502 1E 9.44225e-04 miod69471E 1.55187e-04 miod3854 1E 9.44225e-04 miob46681E 1.62887e-04 ncr3306 1E 9.44225e-04 ncr5055 1E 1.62887e-04 fcr4128 1E 1.032616e-03 seob79291E 1.62887e-04 fcrb3868 1E 1.032616e-03 ncrc38951E 1.80984e-04 fcrb6181 1E 1.032616e-03 seoc07801E 1.84e-04 mioc0824 1E 1.032616e-03 miod11951E 2.00899e-04 ncr8177 1E 1.032616e-03 ncrb19561E 2.00899e-04 fcrb4003 1E 1.128372e-03 fcrb71131E 2.05387e-04 ncrb0364 1E 1.128372e-03 fcrc10801E 2.22795e-04 fcrb3461 1E 1.232026e-03 miod72701E 2.22795e-04 miob6438 1E 1.232026e-03 ncrb03841E 2.22795e-04 mioc3492 1E 1.232026e-03 ncrc45311E 2.22795e-04 ncrc3842 1E 1.232026e-03 fer5618 1E 2.35e-04 ncrc9428 1E 1.232026e-03 fcrb32171E 2.46844e-04 seoa4066 1E 1.232026e-03 fcrb46561E 2.46844e-04 seoa7249 1E 1.232026e-03 fcrc70571E 2.46844e-04 seoa7295 1E 1.232026e-03 mioc14401E 2.46844e-04 seob9282 1E 1.232026e-03 fcrb53391E 2.73236e-04 fcrb1691 1E 1.34414e-03 ncrc35201E 2.73236e-04 mioc1438 1E 1.34414e-03 ncre48851E 2.73236e-04 ncrc3690 1E 1.34414e-03 ncrc97361E 2.73236e-04 seob1153 1E 1.34414e-03 nerc07281E 3.02172e-04 fcr1748 1E 1.465309e-03 ncrc25071E 3.02172e-04 fcrb4988 1E 1.465309e-03 seoc04161E 3.02172e-04 hfcr2616 1E 1.465309e-03 fcrc04871E 3.33867e-04 seoa2641 1E 1.465309e-03 seob41171E 3.33867e-04 seob3088 1E 1.465309e-03 seoa69301E 3.59091e-04 seob6882 1E 1.465309e-03 fcrb43911E 3.68556e-04 fcrb2704 1E 1.596162e-03 fcrb4077~~" .-~~~ ~1~. 596162e-03fcrb3704 1E 3. 079469e-03 ~~~ ~1E
~

fcrc07751E 1.596162e-03 mioa2791 1E 3.079469e-03 fcrc58461E 1.596162e-03 mioa5586 1E 3.079469e-03 ncr4946 1E 1.596162e-03 mioc7620 1E 3.079469e-03 seoa11001E 1.596162e-03 ncr0615 1E 3.079469e-03 fcrb31341E 1.737364e-03 ncr5613 1E 3.079469e-03 fcrb61871E 1.737364e-03 ncrb7102 1E 3.079469e-03 mioa86791E 1.737364e-03 ncrc2387 1E 3.079469e-03 ncrc33911E 1.737364e-03 ncrc3434 1E 3.079469e-03 seoa94211E 1.737364e-03 seoa8268 1E 3.079469e-03 fcrc46581E 1.889615e-03 seob6217 1E 3.132386e-03 hfcr39211E 1.889615e-03 fcrb5087 1E 3.151829e-03 mioc40891E 1.889615e-03 fcrb6432 1E 3.332111e-03 ncrc69811E 1.889615e-03 fcrb9655 1E 3.332111e-03 seoa98141E 1.889615e-03 fcrc5290 1E 3.332111e-03 fcrb28511E 2.053657e-03 fcrc6888 1E 3.332111e-03 fcrc62821E 2.053657e-03 miob3938 1E 3.332111e-03 mioa20731E 2.053657e-03 miob6099 1E 3.332111e-03 mioa76171E 2.053657e-03 ncrc3377 1E 3.332111e-03 ncr3034 1E 2.053657e-03 ncrc6417 1E 3,332111e-03 ncrb83831E 2.053657e-03 seoa0302 1E 3.332111e-03 seoa39891E 2.053657e-03 mioc7170 1E 3.415744e-03 seob76491E 2.053657e-03 seob3090 1E 3.433659e-03 seob79411E 2.053657e-03 ncrb0060 1E 3.505099e-03 mioc08991E 2.084443e-03 fcrb5164 1E 3.602913e-03 fcr1312 1E 2.230269e-03 hfcr2895 1E 3.602913e-03 fcr4803 1E 2.230269e-03 miob8143 1E 3.602913e-03 fcrb53051E 2.230269e-03 ncr0847 1E 3.602913e-03 fcrc01121E 2.230269e-03 ncrb6903 1E 3.602913e-03 mioa84841E 2.230269e-03 ncrc6459 1E 3.602913e-03 miob79221E 2,230269e-03 ncrc6953 1E 3.602913e-03 ncr4673 1E 2.230269e-03 ncrc9159 1E 3.602913e-03 ncrb34981E 2.230269e-03 ncrc9709 1E 3.602913e-03 ncrc28881E 2,230269e-03 seoa1567 1E 3.602913e-03 seob00471E 2.230269e-03 seob7505 1E 3.602913e-03 miob80961E 2.267647e-03 fcr3936 1E 3.892973e-03 fcrb52531E 2.36428e-03 fcrb5928 1E 3.892973e-03 mioa42411E 2.42027e-03 fcrb9856 1E 3.892973e-03 mioa67311E 2.42027e-03 hfcr4349 1E 3.892973e-03 seoa15401E 2.42027e-03 mioa0891 1E 3.892973e-03 fcrc69161E 2.465032e-03 ncr3112 1E 3.892973e-03 fcrb79511E 2.474444e-03 ncr8827 1E 3.892973e-03 fcrb75841E 2.624525e-03 ncrc2701 1E 3.892973e-03 fcrc01661E 2.624525e-03 ncrc3436 1E 3.892973e-03 mioa30801E 2.624525e-03 seoa0464 1E 3,892973e-03 miob97141E 2.624525e-03 seoa5094 1E 3.892973e-03 miod15741E 2.624525e-03 seoa7250 1E 3.892973e-03 miod53691E 2.624525e-03 seob3462 1E 3.892973e-03 ncr5568 1E 2.624525e-03 seob7978 1E 3.892973e-03 ncrb00541E 2.624525e-03 fcr6748 1E 4.203441e-03 ncrc02491E 2.624525e-03 hfcr0501 1E 4.203441e-03 ncrc37991E 2.624525e-03 hfcr4423 1E 4.203441e-03 ncrc67561E 2.624525e-03 ncr2288 1E 4.203441e-03 ncrc98551E 2.624525e-03 ncr8594 1E 4.203441e-03 fcrc39981E 2.701854e-03 ncrb8392 1E 4.203441e-03 fcrb77001E 2.796352e-03 ncrc1595 1E 4.203441e-03 fcrb33421E 2.843941e-03 ncrc9228 1E 4.203441e-03 fcrb80801E 2.843941e-03 seoa1559 1E 4.203441e-03 miob81461E 2.843941e-03 seoa2734 1E 4.203441e-03 mioc20191E 2.843941e-03 seoa4167 1E 4.203441e-03 miod51841E 2.843941e-03 seob3517 1E 4.203441e-03 ncrc57441E 2.843941e-03 fcrb1890 1E 4.328333e-03 seoa99591E 2.843941e-03 hfcr2890 1E 4.366058e-03 fcrb50911E 2.938134e-03 seob3076 1E 4.392091e-03 seob18441E 4.448804e-03 seoa3415 1E 5.674225e-03 seob40901E 4.507008e-03 seoa4107 1E 5.674225e-03 fcrb47211E 4.535517e-03 seob0639 1E 5.674225e-03 fcrb66671E 4.535517e-03 seob7229 1E 5.674225e-03 fcrb78291E 4.535517e-03 seoa4608 1E 5.687664e-03 fcrc19711E 4.535517e-03 fcrb1801 1E 6.105844e-03 miob26341E 4.535517e-03 fcrb2080 1E 6.105844e-03 mioc03371E 4.535517e-03 fcrb2866 1E 6.105844e-03 ncrb86931E 4.535517e-03 fcrb3782 1E 6.105844e-03 ncrc37771E 4.535517e-03 fcrb5841 1E 6.105844e-03 ncrc49851E 4.535517e-03 fcrb8877 1E 6.105844e-03 ncrc90041E 4.535517e-03 fcrc2429 1E 6.105844e-03 ncrc96421E 4.535517e-03 fcrc5233 1E 6.105844e-03 ncrc97291E 4.535517e-03 hfcr0489 1E 6.105844e-03 seob73461E 4.535517e-03 mioa7140 1E 6.105844e-03 seoc40931E 4.535517e-03 miob3042 1E 6.105844e-03 seob47661E 4,631477e-03 miob6518 1E 6.105844e-03 nCrc49201E 4.676963e-03 miob9671 1E 6.105844e-03 seob77471E 4.676963e-03 mioc4319 1E 6.105844e-03 fcrb21131E 4.890457e-03 miod4784 1E 6.105844e-03 fCrb51181E 4.890457e-03 ncr3960 1E 6.105844e-03 fCrb58671E 4.890457e-03 ncrc2600 1E 6.105844e-03 fcrb64841E 4.890457e-03 ncrc5844 1E 6.105844e-03 fcrb89491E 4.890457e-03 ncrc9712 1E 6.105844e-03 fcrc10141E 4.890457e-03 seob4752 1E 6.105844e-03 miob28691E 4.890457e-03 seob8999 1E 6.105844e-03 miob57081E 4.890457e-03 seoc0705 1E 6.373843e-03 miod45181E 4.890457e-03 fcrb9096 1E 6.543015e-03 ncr4485 1E 4.890457e-03 fcrb1769 1E 6.565913e-03 ncr6755 1E 4.890457e-03 fcrb2256 1E 6.565913e-03 nerc20801E 4.890457e-03 fcrb2713 1E 6.565913e-03 seoa82761E 4.890457e-03 fcrb3330 1E 6.565913e-03 seob13191E 4.890457e-03 fcrb6785 1E 6.565913e-03 seob46761E 4.890457e-03 fcrb8910 1E 6.565913e-03 ncr9108 1E 4.911008e-03 mioa2261 1E 6.565913e-03 fcrb28591E 5.26957e-03 mioa7957 1E 6.565913e-03 fcrb66351E 5.26957e-03 miob8572 1E 6.565913e-03 fcrc04811E 5.26957e-03 ncr3435 1E 6.565913e-03 fcrc17401E 5.26957e-03 ncre1349 1E 6.565913e-03 fcrc63451E 5.26957e-03 ncrc2472 1E 6.565913e-03 fcrc68681E 5.26957e-03 ncrc5363 1E 6.565913e-03 hfcr05601E 5.26957e-03 ncrc9412 1E 6.565913e-03 mioc25611E 5.26957e-03 seoa9389 1E 6.565913e-03 ncrb44741E 5.26957e-03 seob6015 1E 6.565913e-03 ncrc55531E 5.26957e-03 fcr1068 1E 6.596445e-03 ncrc89491E 5.26957e-03 seoc0068 1E 6.651586e-03 seoa42021E 5.26957e-03 mioa7169 1E 6.962202e-03 seoa90941E 5.26957e-03 fcrb1539 1E 7.055974e-03 seob20771E 5.26957e-03 fcrb3314 1E 7.055974e-03 seob83001E 5.26957e-03 hfcr3134 1E 7.055974e-03 fcr1633 1E 5.674225e-03 ncr3811 1E 7.055974e-03 fcr4469 1E 5.674225e-03 ncr4975 1E 7.055974e-03 fer4965 1E 5.674225e-03 ncrc3887 1E 7.055974e-03 fcrbl4111E 5.674225e-03 seoa5157 1E 7.055974e-03 fcrb21971E 5.674225e-03 seoa9873 1E 7.055974e-03 fcrb61711E 5.674225e-03 seob0817 1E 7.055974e-03 fcrc49711E 5.674225e-03 seob2966 1E 7.055974e-03 mioc01621E 5.674225e-03 seob9485 1E 7.055974e-03 ncr8337 1E 5.674225e-03 seoc1023 1E 7.055974e-03 ncr8413 1E 5.674225e-03 fcrb3227 1E 7.08625e-03 ncrc35261E 5.674225e-03 fcrc2775 1E 7.131424e-03 ncrc37351E 5.674225e-03 ncrc2227 1E 7.298626e-03 ncrc50721E 5.674225e-03 fcrb7812 1E 7.483981e-03 seoa02561E 5.674225e-03 seob0783 1E 7.547751e-03 fcr1388 1E 7.577631e-03 mioc3045 1E 9.349172e-03 fcr7419 1E 7.577631e-03 mioc5633 1E 9.349172e-03 fcrb14921E 7.577631e-03 ncr0420 1E 9.349172e-03 fcrb21961E 7.577631e-03 ncr3163 1E 9.349172e-03 fcrb26331E 7.577631e-03 ncrb8352 1E 9.349172e-03 fcrb44001E 7.577631e-03 ncrc3598 1E 9.349172e-03 fcrb53891E 7.577631e-03 seob3594 1E 9.349172e-03 fcrb79441E 7.577631e-03 seoc0018 1E 9.349172e-03 fcrc25731E 7.577631e-03 fcrc0430 1E 9.763205e-03 mioa15701E 7.577631e-03 hfcr3494 1E 9.763205e-03 miob72761E 7.577631e-03 fcrc6138 1E 9.895207e-03 miob85831E 7.577631e-03 miod2837 1E 9.96009e-03 ncr9378 1E 7.577631e-03 fcr1337 1E 0.01 ncrb40251E 7.577631e-03 fcr2303 1E 0.01 ncrc35961E 7.577631e-03 fcr4212 1E 0.01 seoa15751E 7.57763Ie-03 fcr4224 1E 0.01 seob35131E 7.577631e-03 fcr6708 1E 0,01 seob55561E 7.577631e-03 fcrc4151 1E 0.01 seob67581E 7.577631e-03 mioa2374 1E 0.01 seob74191E 7,577631e-03 mioc1524 1E 0.01 hfcr64061E 7.634442e-03 miod7442 1E 0.01 ncrc38401E 7.795327e-03 ncr3963 1E 0.01 fcrb48681E 7.885744e-03 ncrc3544 1E 0.01 mioa15321E 7.885744e-03 ncrc4384 1E 0.01 fcrb12021E 8.132553e-03 ncrc6332 1E 0.01 fcrb22081E 8.132553e-03 miob2533 1E 0.01 fcrb63511E 8.132553e-03 miab9559 1E 0.01 fcrc06371E 8.132553e-03 fcrb7240 1E 0.01 fcrc58501E 8.132553e-03 seob5751 1E 0.01 ncrc04571E 8.132553e-03 seoa0536 1E 0,01 ncrc32571E 8.132553e-03 hfcr0676 1E 0.01 ncrc38021E 8.132553e-03 fcrb6481 1E 0.01 ncrc89881E 8.132553e-03 ncr0927 1E 0.01 ncrc94691E 8.132553e-03 seob6835 1E 0,01 seob12131E 8.132553e-03 fcrb4238 1E 0.01 seob80291E 8.132553e-03 fcrc1115 1E 0.01 seob93331E 8.132553e-03 fcrc6877 1E 0.01 miocl9951E 8.415065e-03 mioa0252 1E 0.01 miod52181E 8.473284e-03 mioa1388 1E 0.01 ncrb20851E 8.473284e-03 miob3427 lE 0,01 fcr1305 1E 8.722468e-03 miob8609 1E 0.01 fcrb27841E 8.722468e-03 miob8830 1E 0.01 fcrb36151E 8.722468e-03 mioc1125 1E 0.01 fcrb49261E 8.722468e-03 mioc5226 1E 0.01 fcrb60311E 8.722468e-03 miod1236 1E 0.01 fcrb71271E 8.722468e-03 ncrc2196 1E 0.01 fcrb92221E 8.722468e-03 ncrc2868 1E 0.01 fcrc24711E 8.722468e-03 ncrc5631 1E 0.01 hfcr57371E 8.722468e-03 seoa3408 1E 0,01 mioc39301E 8.722468e-03 seoa7126 1E 0.01 mioc42701E 8.722468e-03 seob4584 1E 0.01 ncr2995 1E 8.722468e-03 seoc0220 1E 0.01 seoa04691E 8.722468e-03 seoc1876 1E 0.01 seoa14191E 8.722468e-03 miod1358 1E 0.01 seoa27651E 8.722468e-03 fcr0111 1E 0.01 seoa99301E 8.722468e-03 fcr3528 1E 0.01 seob54781E 8.722468e-03 fcrb1720 1E O,Ol seoc24471E 8.722468e-03 fcrb2639 1E 0.01 miod63641E 9.069698e-03 fcrb3896 1E 0,01 fcr6630 1E 9.098454e-03 fcrb4985 1E 0,01 fcrbl5781E 9.349172e-03 fcrb9611 1E 0.01 fcrb37021E 9.349172e-03 fcrc2082 1E 0.01 mioa42451E 9.349172e-03 fcrc6419 1E 0.01 miob65971E 9.349172e-03 mioa6428 1E 0.01 miob3411--lE ~~0.01 fcrc4307 1E 0.01 miob87071E O.Ol mioa7239 1E 0.01 ncrb2558lE 0.01 miob7550 1E 0.01 ncrc03831E 0.01 mioc4366 1E O.Ol ncrc69411E 0.01 ncr3404 1E 0.01 ncrc96121E 0.01 ncr7924 1E 0.01 seoa36331E 0.01 ncrc2730 1E 0.01 seobl8911E 0.01 ncrc3171 1E 0.01 seob19811E 0.01 ncrc4135 1E 0.01 miob92481E 0.01 ncrc5123 1E 0.01 fcrb87621E 0.01 seoa7542 1E 0.01 ~

ncrb01641E 0.01 seoa7647 1E 0.01 fcrb39331E 0.01 seob0409 lE 0.01 fcrb42811E 0.01 seob0928 1E O.Ol fcrb44131E 0.01 mioc2726 1E 0.01 fcrb47121E 0.01 fcrb3001 1E 0.01 fcrb47211E 0.01 fcrb5202 1E 0.01 fcrb54401E 0.01 fcrcl689 1E 0.01 fcrb60621E 0.01 fcrc5086 1E 0.01 fcrb62361E 0.01 fcrc6990 1E 0.01 fcrb67501E 0.01 hfcr6033 lE 0.01 fcrb99631E 0.01 ncr4118 1E 0.01 fcrc4876lE 0.01 ncrb5244 1E 0.01 mioal89l1E 0.01 ncrc9401 1E 0.01 miob88251E 0.01 seoa4485 1E 0.01 mioc22041E 0.01 seob0918 1E 0.01 mioc86821E 0.01 seob2959 1E 0.01 ncr2996 1E 0.01 seob4165 1E 0.01 ncr3843 1E 0.01 miod5793 1E 0.01 ncrb74821E 0.01 seoc2029 1E 0.01 ncrc50251E 0.01 seob2953 1E 0.01 ncrc95191E 0.01 fcrc6382 1E 0.01 ncrc97391E 0.01 mioc4667 1E 0.01 seob00631E 0.01 fcr0027 1E 0.01 seob10091E 0.01 fcrb1856 1E 0.01 seob89111E 0.01 fcrb2254 1E 0.01 seoc10251E 0.01 fcrb4729 1E 0.01 seoc21361E 0.01 fcrb6033 1E 0.01 seoc51251E 0.01 fcrb6084 1E 0.01 ncrc06631E 0.01 fcrb7036 1E 0.01 ncrb82391E O.Ol fcrb7380 1E 0.01 fcrb23181E 0.01 fcrc4180 1E 0.01 miob89471E 0.01 fcrc6010 1E 0.01 seoa44641E 0.01 fcrc6228 1E 0.01 mioc75611E 0.01 mioa4229 1E 0.01 fcrb16841E 0.01 miob2386 1E 0.01 fcrb19691E 0.01 miob4370 1E 0.01 fcrb77031E 0.01 miob5119 1E 0.01 fcrc43601E 0.01 miob8704 1E 0.01 mioa49751E 0.01 miod5672 1E 0.01 mioa53261E O.Ol ncr3339 1E 0.01 mioa90251E 0.01 ncr5713 1E 0.01 miob23411E 0.01 ncrc1615 1E 0.01 ncr1768 1E 0.01 ncrc2928 1E 0.01 ncr4545 1E 0.01 ncrc5559 1E 0.01 ncrc66971E 0.01 ncrc5653 1E 0.01 seob02011E 0.01 seoa2381 1E 0.01 seob98511E 0.01 seoa4460 1E 0.01 seoa91291E 0.01 seob1196 1E 0.01 seob29381E 0.01 seob5645 1E 0.01 fcrb93641E 0.01 seob6415 1E 0.01 mioc13571E 0.01 seoc1175 1E 0.01 seoc29041E 0.01 seoa5433 1E 0.01 fcrb15731E 0.01 ncr3118 1E 0.01 1~7 ncrb760-01E O~.b1 hfcr2821 1E 0.01 mfob10591E 0.01 hfcr3834 1E O.Ol fcr0253 1E 0.01 mioa7955 1E 0.01 fcr3717 1E 0.01 miod3241 1E 0.01 fcrb31651E 0.01 ncrc9947 1E 0.01 fcrb55271E 0.01 seoa1041 1E 0.01 fcrb67171E 0.01 seoa1653 1E 0.01 fcrc22541E 0.01 seoa5741 1E 0.01 fcrc42771E 0.01 seob2156 1E 0.01 hfcr12101E 0.01 seob8839 1E 0.01 hfcr56201E 0.01 seob9617 1E 0.01 mioc28911E 0.01 seoc2264 1E 0.01 ncr5557 1E 0.01 seoc6182 IE 0.01 ncr9469 1E 0.01 fcrb9017 1E 0.01 ncrc43731E 0.01 miob4333 1E 0.01 ncrc48631E 0.01 mioc7986 1E 0.02 ncrc54921E 0.01 fcr3282 1E 0.02 ncrc88411E 0.01 fcrb3808 1E 0.02 seoa00541E 0.01 fcrb5675 1E 0.02 seoa16721E 0.01 fcrc6460 1E 0.02 seob60841E 0.01 hfcr5959 1E 0.02 seoc10091E 0.01 miob7443 1E 0.02 seoc35191E 0.01 mioc3042 1E 0.02 fcrc56901E 0.01 mioc7509 1E 0.02 mioc62121E 0.01 miodl333 1E 0.02 seoa45871E 0.01 miod6437 1E 0.02 fcrb27651E 0.01 ncr3412 1E 0.02 hfcr31601E 0.01 ncrb5972 1E 0.02 fcr435'71E 0.01 ncrc2776 1E 0.02 mioc12031E 0.01 ncrc4539 1E 0.02 ncrb82011E 0.01 seoa0044 1E 0.02 fcrc60431E 0.01 seoa0221 1E 0.02 fcrb11901E 0.01 seoa4457 1E 0.02 fcrb16901E 0.01 seoa9445 1E 0.02 fcrb18771E 0.01 seoa9690 1E 0.02 fcrb19921E 0.01 seob1423 1E 0.02 fcrb24241E 0.01 seob1947 1E 0.02 fcrb38971E 0.01 seob6781 1E 0.02 fcrb82362E 0.01 seoc0651 1E 0.02 fcrb84221E 0.01 miod3160 1E 0.02 fcrc6970lE 0.01 seoc4900 1E 0.02 fcrc6989lE 0.01 seob8065 1E 0.02 mioa50971E 0.01 ncrc4772 1E 0.02 miod08941E 0.01 fcr0748 1E 0.02 miod53491E 0.01 fcrb1689 1E 0.02 ncr3975 1E 0.01 fcrb2592 1E 0.02 ncrc02881E 0.01 fcrb2719 1E 0.02 ncrc34531E 0.01 fcrb2749 1E 0.02 ncrc39531E 0.01 fcrb3288 1E 0.02 ncrc54231E 0.01 fcrb4572 1E 0.02 ncrc97931E 0.01 fcrb6870 1E 0.02 seoa22191E 0.01 miob4570 1E 0.02 seoa47831E 0.01 miod4521 1E 0.02 seoa55801E 0.01 ncr715,1 1E 0.02 seoa64531E 0.01 ncr7960 1E 0.02 seob41921E 0.01 ncrcl643 1E 0.02 seoc16421E 0.01 ncrc5039 1E 0.02 seoa17361E 0.01 seoa1410 1E 0.02 fcr2106 1E 0.01 seoa1615 1E 0.02 fcrb35731E 0.01 seoa2775 1E 0.02 fcrb58921E 0.01 seob0787 1E 0.02 fcrb62021E 0.01 seob0857 1E 0.02 fcrb66431E 0.01 seoc1180 1E 0.02 fcrc50601E 0.01 mioc0019 1E 0.02 ncrb71111E 0.02 seob9970 1E 0.02 fcrb20511E 0.02 fcrb2166 lE 0.02 fcrb27451E 0.02 fcrb2813 1E 0.02 fcrb31811E 0.02 fcrb4760 1E 0.02 fcrb36541E 0.02 fcrb5092 1E 0.02 fcrb43651E 0.02 fcrb9481 1E 0.02 fcrb59121E 0.02 mioc4204 1E 0.02 fcrb65491E 0.02 seoa4436 1E 0.02 fcrb76161E 0.02 seob1860 1E 0.02 fcrb93381E 0.02 seob5558 1E 0.02 fcrc54711E 0.02 seob7981 1E 0.02 mioa21561E 0.02 seoc1495 1E 0.02 miob26641E 0.02 ncr8357 1E 0.02 mioc20741E 0.02 mioc4100 1E 0.02 mioc31111E 0.02 seob5551 1E 0.02 ncrb83431E 0.02 seob2336 1E 0.02 ncrcl6311E 0.02 seob1052 1E 0.02 ncrc57751E 0.02 ncrc9354 1E 0.02 ncrc94381E 0.02 fcrb1337 1E 0.02 ncrc97721E 0.02 fcrb1380 1E 0.02 seoa55771E 0.02 fcrb1930 1E 0.02 seoa70771E 0.02 fcrb2094 1E 0.02 seoa98171E 0.02 fcrb4781 1E 0.02 seob08151E 0.02 fcrb5037 1E 0.02 seob09711E 0.02 fcrb5181 1E 0.02 seob31121E 0.02 fcrb5326 1E 0.02 seob56241E 0.02 fcrb6699 1E 0.02 hfcr25471E 0.02 fcrc0810 . 0.02 seoa41321E 0.02 fcrc7219 1E 0.02 seoc45851E 0.02 miob0636 1E 0.02 fcrb23341E 0.02 miob9652 1E 0.02 miob93401E 0.02 mioc3105 1E 0.02 mioa33671E 0.02 mioc3994 1E 0.02 fcr4695 1E 0.02 ncr2363 1E 0.02 fcrb20591E 0.02 ncrc0640 1E 0.02 fcrb45421E 0.02 ncrc3435 1E 0.02 fcrb52141E 0.02 ncrc3464 1E 0.02 fcrb63221E 0.02 ncrc4798 1E 0.02 fcrb67381E 0.02 ncrc5724 1E 0.02 fcrb86141E 0.02 seoa1263 1E 0.02 fcrc40671E 0.02 seoa1910 1E 0.02 hfcr30461E 0.02 seob0228 1E 0.02 mioa16871E 0.02 seoc0616 1E 0.02 mioa24231E 0.02 seoc2682 1E 0.02 miob75311E 0.02 seoa3895 1E 0.02 mioc40221E 0.02 seoc3469 1E 0.02 mioc45041E 0.02 fcrb5926 1E 0.02 miod49981E 0.02 ncrc4555 1E 0.02 miod74861E 0.02 hfcr0366 1E 0.02 ncr4700 1E 0.02 miob9010 1E 0.02 ncrb88211E 0.02 fcrb1296 1E 0.02 ncrc97661E 0.02 fcrb1809 1E 0.02 seoa31641E 0.02 fcrb4570 1E 0.02 seoa58431E 0.02 fcrb5123 1E 0.02 seob12731E 0.02 fcrb5723 1E 0.02 seob34851E 0.02 fcrc0363 1E 0.02 seob52131E 0.02 fcrc5379 1E 0.02 ncrb02991E 0.02 fcrc7087 1E 0.02 seoc27221E 0.02 mioa3428 1E 0.02 ncr0721 1E 0.02 mioa3629 1E 0.02 miod37761E 0.02 miob0986 1E 0.02 fcrb42711E 0.02 miod6234 1E 0.02 hfcr30101E 0.02 ncr0808 1E 0.02 ncrb02651E 0.02 ncrc7049 1E 0.02 1~9 seoa432~4~~~~~~"~~~ "~~~ G~~: ncrc3593 1E 0.
~~ ~1E- 02 ~ 03 seoa54941E 0.02 ncrc4323 1E 0.03 seob97561E 0.02 ncrc5079 1E 0.03 seob97641E 0.02 ncrc5230 1E 0.03 seob98981E 0.02 seoa2179 1E 0.03 ncrb75571E 0.02 seoa6621 1E 0.03 seob60171E 0.03 seoa7543 1E 0.03 fcr1844 1E 0.03 miob8320 1E 0.03 fcrb34831E 0.03 miob9185 1E 0.03 fcrb54671E 0.03 seoc4132 1E 0.03 fcrb96341E 0.03 fcrb1369 1E 0.03 fcrc11811E 0.03 fcrc0232 1E 0.03 fcrc47221E 0.03 seob7030 1E 0.03 mioa80321E 0.03 fcr2442 1E 0.03 miob98201E 0.03 fcrb1876 1E 0.03 mioc03011E 0.03 fcrb1990 1E 0.03 mioc17681E 0.03 fcrb3717 1E 0.03 mioc19101E 0.03 fcrb4067 1E 0.03 mioc20971E 0.03 fcrb5564 1E 0.03 mioc4929lE 0.03 fcrc0148 1E 0.03 miod49741E 0.03 fcrc2040 1E 0.03 ncr2994 1E 0.03 hfcr5691 1E 0.03 ncr3189 1E 0.03 mioa5836 1E 0.03 ncr8075 1E 0.03 miob6598 1E 0.03 ncrc12591E 0.03 miod4540 1E 0.03 ncrc21821E 0.03 ncr0097 1E 0.03 ncrc36241E 0.03 ncr3357 1E 0.03 ncrc47281E 0.03 ncr3402 1E 0.03 ncrc89321E 0.03 ncrb2027 1E 0.03 ncrc98671E 0.03 ncrc0284 1E 0.03 seoa17391E 0.03 ncrc2613 1E 0.03 seoa44521E 0.03 ncrc2859 1E 0.03 seoa46401E 0.03 ncrc9284 1E 0.03 seoa66541E 0.03 seoal747 1E 0.03 seoa66971E 0.03 seoa3761 1E 0.03 seob00461E 0.03 seoa8822 1E 0.03 seob02191E 0.03 seob2717 1E 0.03 seob11551E 0.03 fcrc0559 1E 0.03 seob21391E 0.03 fcrc7222 1E 0.03 seob31891E 0.03 ncrb3980 1E 0.03 seoc16311E 0.03 seoal582 1E 0.03 seoc21311E 0.03 mioc7952 1E 0.03 mioa59511E 0.03 seoc4038 1E 0.03 seob50641E 0.03 miob1774 1E 0.03 ncrc90551E 0.03 mioc1030 1E 0.03 seoc81151E 0.03 seob7039 1E 0.03 seob01821E 0.03 ncr8314 1E 0.03 mioc23361E 0.03 fcr2079 1E 0.03 seoc13071E 0.03 fcr2798 1E 0.03 fcr4471 1E 0.03 fcr3269 1E 0.03 fcrb13811E 0.03 fcr4846 1E 0.03 fcrb31201E 0.03 fcrb4470 1E 0.03 fcrb47821E 0.03 fcrb7852 1E 0.03 fcrc03761E 0.03 fcrb8133 1E 0.03 fcrc39931E 0.03 fcrc5705 1E 0.03 fcrc51341E 0.03 hfcr0412 1E 0.03 mioa01921E 0.03 hfcr2629 1E 0.03 mioa16601E 0.03 hfcr6043 1E 0.03 mioa91541E 0.03 mioa1380 1E 0.03 mioa98311E 0.03 mioa6913 1E 0.03 mioc15601E 0.03 miob3072 1E 0.03 miod23691E 0.03 miod4066 1E 0.03 miod32541E 0.03 ncr3141 1E 0.03 ncrc05291E 0.03 ncr7267 1E 0.03 ncrc3713T'E "0:U3 miob4673 1E 0.04 seoa15521E 0.03 mioc3066 1E 0.04 seoa16611E 0.03 mioc7421 1E 0.04 seoa17201E 0.03 ncr0110 1E 0.04 seoa59111E 0.03 ncr1055 1E 0.04 seoa9777lE 0.03 ncr7904 1E 0.04 seob49251E 0.03 ncr9934 1E 0.04 seob8817lE 0.03 ncrb8569 1E 0.04 fcrb50511E 0.03 ncrcl374 1E 0.04 mioc09101E 0.03 ncrc2290 1E 0.04 miod07731E 0.03 seoa1770 1E 0.04 seob5562lE 0.03 seoa7587 1E 0.04 hfcr25241E 0.03 seob0065 1E 0.04 fcr4468 1E 0.03 seob5882 1E 0.04 fcrb57251E 0.03 seoa9042 1E 0.04 fcrb89811E 0.03 seoc1957 1E 0.04 fcrb93241E 0.03 mioa0601 1E 0.04 fcrc5614lE 0.03 miod5771 1E 0.04 fcrc60111E 0.03 seob1839 1E 0.04 fcrc65511E 0.03 fcrb2591 1E 0.04 fcrc70141E 0.03 miob1830 1E 0.04 mioa33921E 0.03 fcrb4981 1E 0.04 mioa35141E 0.03 hfcr5514 1E 0.04 mioa54471E 0.03 fcrb2715 1E 0.04 mioa89461E 0.03 fcrb2792 1E 0.04 miob71171E 0.03 fcrb2796 1E 0.04 miob9679IE 0.03 fcrb2979 1E 0.04 mioc47661E 0.03 fcrb6817 1E 0.04 ncr0836 1E 0.03 fcrb9007 1E 0.04 ncr7286 1E 0.03 f_crb9225 1E 0.04 ncrb43191E 0.03 mioal944 1E 0.04 ncrc03931E 0.03 mioa5202 1E 0.04 ncrc08491E 0.03 miob2656 1E 0.04 ncrc35371E 0.03 mioc5736 1E 0.04 ncrc50191E 0.03 mioc6878 1E 0.04 ncrc65871E 0.03 ncr6326 1E 0.04 seoa01351E 0.03 ncr7341 1E 0.04 seoa25281E 0.03 ncr9371 1E 0.04 seoa83521E 0.03 ncrb8641 1E 0.04 seoa94821E 0.03 ncrc3417 1E 0.04 seob02001E 0.03 ncrc7131 1E 0.04 seob50541E 0.03 seoa4327 1E 0.04 seob6l331E 0.03 seob2735 1E 0.04 miod72121E 0.03 seob7729 1E 0.04 ncr2182 1E 0.03 ncr0612 1E 0.04 seoa64661E 0.03 ncr3834 1E 0.04 fcrb25541E 0.03 mioc8766 1E 0.04 seoa99971E 0.04 ncr0251 1E 0.04 hfcr21481E 0.04 ncrc2463 1E 0.04 mioa70691E 0.04 fcrb5756 1E 0.04 miob27091E 0.04 miob2116 1E 0.04 fcr4494 1E 0.04 ncr3718 1E 0.04 fcrb15801E 0.04 fcrb8449 1E 0.04 fcrb1633lE 0.04 ncrc5804 1E 0.04 fcrb24991E 0.04 fcr2276 1E 0.04 fcrb28491E 0.04 miod2334 1E 0.04 fcrb32191E 0.04 seob3139 1E 0.04 fcrb35151E 0.04 seoc1881 1E 0.04 fcrb84451E 0.04 seoa8960 1E 0.04 fcrb96361E 0.04 fcrb2569 1E 0.04 fcrc51371E 0.04 fcrb6472 1E 0.04 fcrc53281E 0.04 fcrc0351 1E 0.04 mioa03641E 0.04 fcrc0661 1E 0.04 mioa28511E 0.04 fcrc2647 1E 0.04 fer~c63$1~T~' d":"(~'4' seob0547 lE 0.04 hfcr11891E 0.04 seob1906 1E 0.04 hfcr27701E 0.04 seob5395 1E 0.04 miob58I01E 0.04 hfcr5865 1E 0.04 miod60251E 0.04 fcrc5402 1E 0.04 ncrb48691E 0.04 miob9228 1E 0.04 ncrb85381E 0.04 hfcr5970 1E 0.04 ncrc02171E 0.04 ncrc4079 1E 0.04 ncrc07471E 0.04 hfcr4046 1E 0.04 ncrc11681E 0.04 seob3513 1F 1.44e-05 ncrc26071E 0.04 miob9652 1F 9.27e-05 ncrc6416lE 0.04 mioa2290 1F 1.00455e-04 ncrc89031E 0.04 mioa3987 1F 1.01e-04 seoa18021E 0.04 hfcr4489 1F 1.17e-04 seoa35551E 0.04 ncr2363 1F 1.17e-04 seoa54471E 0.04 mioc5633 1F 1.43481e-04 seob19081E 0.04 fcrb4383 1F 1.45e-04 seoc12301E 0.04 mioa2204 1F 1.61204e-04 mioa1077lE 0.04 seoa3516 1F 1.61204e-04 fcrc39581E 0.04 hfcr6129 1F 1.84e-04 mioc10021E 0.04 ncr0132 1F 1.89e-04 ncrc94401E 0.04 mioa8439 1F 2.36e-04 seoc3554lE 0.04 ncr3598 1F 2.41e-04 fcrc0725lE 0.04 fcr3936 1F 2.51e-04 seob97341E 0.04 mioa8864 1F 2.53933e-04 ncr2583 lE 0.04 fcrb1296 1F 2.69e-04 fcrb1775lE 0.04 hfcr6033 1F 2.73e-04 fcrb19161E 0.04 ncrc3257 1F 2.77e-04 fcrb88721E 0.04 fcr4214 1F 2.91e-04 mioa8318lE 0.04 hfcr2041 1F 2.92e-04 miob04871E 0.04 fcr6748 1F 3.0e-04 seoa48291E 0.04 mioa3321 1F 3.02e-04 seoc05131E 0.04 hfcr5991 1F 3.11e-04 fcrc23761E 0.04 seoa5554 1F 3.36e-04 fcrc27371E 0.04 mioa3392 1F 3.68e-04 fcrc49481E 0.04 mioa2213 1F 3.85e-04 ncr0451 1E 0.04 fcrb2784 1F 3.9297e-04 ncr3177 1E 0.04 fcrb8196 1F 3.9297e-04 ncrc48821E 0.04 ncrb8530 1F 4.1e-04 seoa36701E 0.04 ncr3614 1F 4.14e-04 seob04831E 0.04 seoc1025 1F 4.3e-04 seob57621E 0.04 ncr4975 1F 4.61e-04 fcr5425 1E 0.04 seob4075 1F 4.85728e-04 fcrb43881E 0.04 ncr7839 1F 5.15e-04 fcrb78851E 0.04 ncr3815 1F 5.17e-04 fcrc48481E 0.04 hfcr5942 1F 5.23e-04 fcrc50071E 0.04 ncr8843 1F 5.4e-04 mioa11651E 0.04 fcr4212 1F 5.61e-04 mioa39871E 0.04 fcrb2334 1F 5.71e-04 miob20671E 0.04 hfcr2275 1F 5.9789e-04 miob27561E 0.04 mioa7465 1F 6.04e-04 miob73731E 0.04 miob5810 1F 6.13e-04 mioc33321E 0.04 hfcr5864 1F 6.22e-04 mioc75421E 0.04 fcrc6542 1F 6.5e-04 mioc80551E 0.04 fcr6708 1F 6.58e-04 miod08781E 0.04 mioa0311 1F 6.62324e-04 ncrc05581E 0.04 mioc7381 1F 6.62324e-04 ncrc11031E 0.04 ncrb8419 1F 6.62324e-04 ncrc12471E 0.04 seob4105 1F 6.64e-04 ncrc38561E 0.04 miob1545 1F 7.28e-04 seoa23851E 0.04 fcrb1633 1F 7.31e-04 seoa47081E 0.04 mioa1276 1F 7.32964e-04 seoa85561E 0.04 seoa5849 1F 7.39e-04 seoa97051E 0.04 fcrb5467 1F 7.45e-04 s-eoa23851F "''~T:'f3e'-0'4' seob3697 1F 1. 914664e-03 -seoa04181F 7.66e-04 fcrb2763 1F 1.928684e-03 seob81741F 7.87e-04 hfcr5220 1F 1.93844e-03 mioc48351F 8.0e-04 seob0755 1F 2.00392e-03 miob36961F 8.10329e-04 ncrc3650 1F 2.009196e-03 hfcr56201F 8.11e-04 miod3558 1F 2.053453e-03 fcr4414 1F 8.14e-04 fcrb7l13 1F 2.056425e-03 fcrb43451F 8.23e-04 fcrb3234 1F 2.126705e-03 seob83681F 8.38e-04 seoc4260 1F 2.16146e-03 seob34681F 8.52e-04 mioa8150 1F 2.221472e-03 mioa17041F 8.62e-04 ncrb3369 1F 2.227846e-03 hfcr59051F 8.63e-04 hfcr6141 1F 2.247103e-03 seoa33921F 8.67e-04 seoc0068 1F 2.284571e-03 ncr2967 1F 8.69e-04 ncr3037 1F 2.287966e-03 seob41921F 8.69e-04 miob5888 1F 2.294368e-03 fcr4695 1F 8.71e-04 ncr6108 1F 2.294368e-03 mioa09721F 8.94977e-04 seob2685 1F 2.294368e-03 mioa21561F 9.04e-04 miob3982 1F 2.333313e-03 ncr0847 1F 9.04e-04 ncrb0513 1F 2.334478e-03 fcrr_68771F 9.07e-04 ncr9789 1F 2.417674e-03 fcrb64651F 9.31e-04 ncrb0460 1F 2.498431e-03 hfcr34671F 9.62e-04 seoa5577 1F 2.502273e-03 mioa03641F 9.72e-04 ncr9933 1F 2.508241e-03 fcrc51361F 9.87503e-04 mioc0950 1F 2.588257e-03 ncrb59721F 1.013254e-03 seob4140 1F 2.589542e-03 ncrb80561F 1.025404e-03 mioc5603 1F 2.598809e-03 ncrb18801F 1.06995e-03 ncrb8437 1F 2.624852e-03 mioa46741F 1.08854e-03 seoc2904 1F 2.671857e-03 fcrb73241F 1.132955e-03 mioa1868 1F 2.739644e-03 ncrb52331F 1.138168e-03 mioc641'7 1F 2.739644e-03 mioa09091F 1.189492e-03 ncr3960 1F 2.739644e-03 fcr0680 1F 1.198768e-03 ncrc3908 1F 2.739644e-03 fcrb2C541F 1.198768e-03 seob0409 1F 2.739644e-03 miob96711F 1.198768e-03 seob6872 1F 2.739644e-03 fcrc62281F 1.250123e-03 fcrb5645 1F 2.768103e-03 mioc06991F 1.295875e-03 seoc2633 1F 2.782919e-03 hfcr54991F 1.299893e-03 seob6015 1F 2.805059e-03 fcrb22461F 1.302022e-03 mioc7471 1F 2.877428e-03 miodl3891F 1.318907e-03 seoal749 1F 2.894471e-03 fcr4902 1F 1.361678e-03 seob7465 1F 2.982853e-03 fcrb22541F 1.375926e-03 ncr4572 1F 2.989794e-03 seoa0353lF 1.396441e-03 ncrc6871 1F 2.989794e-03 fcrc18031F 1.449724e-03 fcrb6715 1F 2.992521e-03 miob37571F 1.449724e-03 ncrc1411 1F 3.058649e-03 fcrb21621F 1.456955e-03 hfcr2930 1F 3.059881e-03 mioa36931F 1.472788e-03 fcr3717 1F 3.061351e-03 fcrb27441F 1.519135e-03 miod4019 1F 3.069637e-03 ncr6137 1F 1.552827e-03 fcrb2080 1F 3.191254e-03 fcrb65081F 1.580622e-03 ncr9175 1F 3.251549e-03 fcrb13971F 1.592036e-03 hfcr6509 1F 3.257391e-03 mioa90621F 1.656379e-03 mioa2292 1F 3.259973e-03 hfcr65151F 1.683036e-03 ncrc9850 1F 3.259973e-03 ncrc0856lF 1.694702e-03 seob2283 1F 3.259973e-03 ncr0097 1F 1.703212e-03 seoa4066 1F 3.271732e-03 mioa98911F 1.746709e-03 fcrb1950 1F 3.335219e-03 seoa05361F 1.746709e-03 seoa9690 1F 3.339303e-03 ncr3402 1F 1.757133e-03 ncrc4985 1F 3.345513e-03 miob18301F 1.762715e-03 ncr7945 1F 3.346416e-03 fcrb23211F 1.829751e-03 hfcr6043 1F 3.358515e-03 fcrc02871F 1.883587e-03 seob6486 1F 3.422066e-03 fcrb68681F 1.914664e-03 mioa0607 1F 3.45556e-03 fcrc47221F 1.914664e-03 fcrc5604 1F 3.457081e-03 mioa44841F 1.914664e-03 seob8311 1F 3.529644e-03 miob71171F 1.914664e-03 miob0167 1F 3.551533e-03 ncr4503 1F 3.551533e-03 ncrc0849 1F 4.961124e-03 seoa16891F 3.551533e-03 seob1808 1F 4.961124e-03 seob70391F 3.55I533e-03 seob5673 1F 4.961124e-03 fcrb23461F 3,631263e-03 seoa7605 1F 5.014096e-03 miob83731F 3.653412e-03 seob8483 1F 5.052725e-03 hfcr54731F 3.664884e-03 mioc4842 1F 5.15722e-03 ncr4009 1F 3.745829e-03 hfcr5522 1F 5.165867e-03 ncr8177 1F 3.751419e-03 hfcr6687 1F 5.175533e-03 miob10061F 3,759831e-03 fcrc2126 1F 5.191177e-03 ncr0609 1F 3,81321e-03 mioa5452 1F 5.191277e-03 seoa48291F 3,817245e-03 miob2093 1F 5.218484e-03 mioa23431F 3.845825e-03 ncrc0017 1F 5.259195e-03 ncrc69531F 3.845825e-03 nerc3457 1F 5.259195e-03 fcr5425 1F 3.859974e-03 mioa8820 1F 5.382406e-03 ncr4485 1F 3.863747e-03 miod0456 1F 5.382406e-03 fcrb56391F 3.865895e-03 ncr3763 1F 5.382406e-03 mioa47531F 3,865895e-03 ncrb1670 1F 5.382406e-03 mioa90611F 3.865895e-03 fcrbl922 1F 5.52751e-03 miob01801F 3.865895e-03 fcr3269 1F 5,600028e-03 miob01891F 3.865895e-03 fcr2417 1F 5.638591e-03 miod01871F 3.865895e-03 seoa5785 1F 5.696732e-03 seoa62381F 3.865895e-03 seoa1089 1F 5.709807e-03 seob11331F 3.865895e-03 fcrb2510 1F 5.834736e-03 ncrc30301F 3.933089e-03 miob2227 1F 5.834736e-03 seoa85471F 3,94569e-03 mioc2561 1F 5.834736e-03 fcr3743 1F 3,964466e-03 seoa5691 1F 5.834736e-03 mioa89841F 3.964466e-03 seoc2220 1F 5.834736e-03 hfcr63751F 4,00567e-03 ncrc7127 1F 5.867274e-03 hfcr59561F 4,064802e-03 mioc2019 1F 5.891644e-03 ncr3189 1F 4.13181e-03 mioa6969 1F 5.906016e-03 seob6U841F 4.203369e-03 seob7575 1F 6.097516e-03 fcrb68701F 4.204555e-03 ncrc3593 1F 6.113977e-03 mioc39581F 4.204555e-03 ncrb4390 1F 6.170009e-03 ncrb74821F 4,204555e-03 fcrb6896 1F 6.199739e-03 ncrc27801F 4.204555e-03 hfcr5232 1F 6.232511e-03 seoc09571F 4.204555e-03 ncrc5019 1F 6.25157e-03 mioc05671F 4.232828e-03 mioc5194 1F 6.294222e-03 fcrb51941F 4,288194e-03 mioa3940 1F 6.302655e-03 ncr3419 1F 4,297798e-03 fcrb2315 1F 6.316326e-03 fcrc54711F 4,303567e-03 fcrc6374 1F 6.320005e-03 fcrc39421F 4,385521e-03 mioa6721 1F 6.320005e-03 seob67011F 4,441698e-03 miob2466 1F 6.320005e-03 seoa4056IF 4.458253e-03 ncr8481 1F 6.320005e-03 seob84891F 4.541888e-03 ncrb8605 1F 6.320005e-03 fcrb23051F 4,556834e-03 seoa0099 1F 6.320005e-03 mioa21851F 4.556834e-03 seoa0388 1F 6.320005e-03 fcr0821 1F 4.569082e-03 seocl348 1F 6.320005e-03 fcr3823 1F 4,569082e-03 miob6099 1F 6.326637e-03 hfcr22951F 4.569082e-03 ncrc5079 1F 6.338658e-03 hfcr05221F 4,585758e-03 seob0586 1F 6.373817e-03 mioc70731F 4,595129e-03 seoa9889 1F 6.418044e-03 mioc20941F 4.690236e-03 fcr4699 1F 6.424174e-03 mioc25461F 4.717485e-03 fcr5019 1F 6.499398e-03 miob43071F 4.736268e-03 miod4269 1F 6.541189e-03 fcr7656 1F 4.778951e-03 ncrb0571 1F 6.562046e-03 ncrc98551F 4.823976e-03 seob7419 1F 6.640174e-03 hfcr36741F 4,87135e-03 mioa2774 1F 6.667082e-03 fcrb55271F 4.893773e-03 ncr6401 1F 6.682138e-03 fcrb22071F 4.928943e-03 seoa7266 1F 6.701766e-03 mioc61561F 4.9342e-03 seob3699 1F 6.753484e-03 fcr7095 1F 4.961124e-03 fcrb2356 1F 6.825088e-03 mioa91471F 4.961124e-03 seob6380 1F 6.825088e-03 miob93361F 4.961124e-03 fcrc2050 1F 6.840188e-03 ncrb63851F 4.961124e-03 hfcr5237 1F 6.840188e-03 mioa9033~ ~'~~-~~F' """'6":'8'40~1~88e-03ncrc6846 1F 8. 631293e-03 .~~~ """

mioa96301F 6.840188e-03 seoa5303 1F $,631293e-03 miob02131F 6.840188e-03 fcrb2l33 1F 8.647588e-03 ncrc0427lF 6.840188e-03 hfcr3149 1F 8.667419e-03 seoa60781F 6.840188e-03 mioc3206 1F 8.811712e-03 fcr66911F 6.901382e-03 ncrc3733 1F 8.814606e-03 fcrb52971F 6.946937e-03 miod5703 1F 8,861302e-03 fcrc39981F 6.962613e-03 ncrb1861 1F 8.882816e-03 ncrc97001F 6.962613e-03 mioc4420 1F 8.897141e-03 seob52601F 6.964546e-03 fcr5470 1F 8.900727e-03 ncrc42261F 6.994012e-03 miod5505 1F 8,920428e-03 seob43631F 7.032122e-03 fcrb2126 1F 9.199582e-03 ncr95021F 7.122344e-03 ncrc1578 1F 9.205415e-03 seoa04691F 7.129312e-03 hfcr3224 1F 9.219579e-03 mioa88511F 7.216452e-03 miod2232 1F 9.250174e-03 mioa56811F 7.217865e-03 seoa5578 1F 9.281899e-03 miob89471F 7.23621e-03 mioa6991 1F 9.31265e-03 seoa51511F 7.267712e-03 ncr7570 1F 9.31265e-03 fcr07511F 7.378041e-03 seoa0420 1F 9.31265e-03 fcrc39931F 7.389258e-03 seoa8750 1F 9.31265e-03 fcrb66201F 7.397348e-03 seob1526 1F 9.31265e-03 fcrc68921F 7.397348e-03 seob3322 1F 9,31265e-03 hfcr00111F 7.397348e-03 miob2668 1F 9.456576e-03 miob79701F 7.397348e-03 seoc1078 1F 9.467094e-03 ncrb02621F 7.397348e-03 seob2163 1F 9.524093e-03 ncrc11921F 7.397348e-03 ncr3040 1F 9.576948e-03 seoa73691F 7.397348e-03 ncrb3449 1F 9.579048e-03 seoc16311F 7.397348e-03 ncr9429 1F 9.603911e-03 mioa14171F 7.397472e-03 ncrc4323 1F 9,65164e-03 fcrb30791F 7.406491e-03 fcrc4307 1F 9.674174e-03 seoc07781F 7.492808e-03 seoa0137 1F 9.678908e-03 ncrb32381F 7.636087e-03 mioa2330 1F 9.685757e-03 mioc76621F 7.706037e-03 fcrb2306 1F 9.694983e-03 seoa34151F 7.706037e-03 hfcr5228 1F 9.697553e-03 seob04421F 7.706037e-03 fcrb7700 1F 9.721976e-03 ncr59751F 7.721088e-03 hfcr5181 1F 9.747239e-03 seoa04291F 7.723786e-03 ncrc3544 1F 9.74833e-03 fcrb61951F 7.742707e-03 ncrb8518 1F 9,75683e-03 ncr86931F 7.754322e-03 fcrb3192 1F 9,87471e-03 fcr14041F 7.795263e-03 fcr5006 1F 9.917803e-03 nerc18711F 7.818883e-03 fcr5199 1F 9.921817e-03 mi.oa13041F 7.968902e-03 ncrb0200 1F 9,990439e-03 fcrb09791F 7.993636e-03 fcrb1477 1F 0.01 mioc35801F 7.993636e-03 fcrb1494 1F 0.01 seoa04861F 7.993636e-03 mioc7444 1F 0,01 seoc56271F 7.993636e-03 ncr8272 1F 0.01 fcrb65741F 8.038684e-03 ncrc2831 1F 0.01 mioc01611F 8.075165e-03 ncrc4757 1F 0,01 miod35911F 8.085647e-03 seoa0145 1F 0.01 hfcr58601F 8.106115e-03 seob0949 1F 0.01 ncrc34161F 8.126637e-03 seob1100 1F 0,01 miob22871F 8.142773e-03 seob4793 1F 0.01 seob14191F 8.163409e-03 seob5812 1F 0.01 fcrc09101F 8.3518e-03 miod2412 1F 0.01 seob22211F 8.390552e-03 hfcr5969 1F 0.01 seoc17851F 8.437195e-03 miod5622 1F 0.01 ncr3276IF 8.480045e-03 hfcr5719 1F O,OI

seoa8902IF 8.536355e-03 seoa8814 1F 0.01 mioa87131F 8.544193e-03 seoa3555 1F 0.01 fcr02241F 8.631293e-03 ncrb6742 IF 0.01 fcr45031F 8.631293e-03 miob2905 1F 0.01 mioa60931F 8.631293e-03 fcrb6650 1F 0,01 mioa97921F 8.631293e-03 fcrb1689 1F 0.01 ncr54261F 8.631293e-03 ncrb2092 1F 0.01 " fcrc0405IF ""'"D~:~-0~1 ncr3965 1F 0.

fcrc2014 1F 0.01 miod6467 1F 0.01 ncr5529 1F 0.01 ncrc9343 1F 0.01 seoa8754 1F 0.01 miod0807 1F 0.01 ncrc4444 1F 0.01 ncrb0328 1F 0.01 fcrb0673 1F 0.01 seob0061 1F 0.01 seoc3426 1F 0.01 miod7011 1F 0.01 ncrc0863 1F 0.01 hfcr2693 1F 0.01 mioc4843 1F 0.01 ncrc9228 1F 0.01 fcrb2713 1F 0.01 ncrc4903 1F 0.01 ncr8995 1F 0.01 seob9480 1F 0.01 hfcr1083 1F 0.01 fcrb6791 1F 0.01 ncrc1904 1F 0.01 mioa2038 1F 0.01 fcrc6855 1F 0.01 hfcr0383 1F 0.01 mioc7731 1F 0.01 fcrc5190 1F 0.01 seob5099 1F 0.01 fcrc5250 1F 0.01 ncrc1999 1F 0.01 hfcr1201 1F 0.01 seoc5911 1F 0.01 mioa9007 1F 0.01 miod4539 1F 0.01 mioa9935 1F 0.01 fcr0018 1F 0.01 mioc3906 1F 0.01 fcr0990 1F 0.01 ncr9487 1F 0.01 hfcr4275 1.F 0.01 ncrc0249 1F 0.01 mioa8767 1F 0.01 seoa0045 1F 0.01 miob3426 1F 0.01 seoa6358 1F 0.01 miob4756 1F 0.01 miob5495 1F 0.01 ncr2269 lF 0.01 mioa9649 1F 0.01 ncrc3971 1F 0.01 miod2369 1F 0.01 seoa0470 1F 0.01 ncr3934 1F 0.01 seob2994 1F 0.01 hfcr5865 1F 0.01 seoc6666 1F 0.01 fcrc6476 1F 0.01 ncr3971 1F 0.01 fcrc0471 1F 0.01 seob5562 1F 0.01 seob1145 1F 0.01 seoc~888 1F 0.01 fcrc2131 1F 0.01 ncr8357 1F 0.01 mioal293 1F 0.01 seob5579 1F 0.01 ncrc5631 1F 0.01 fcrb3017 1F 0.01 mioc3045 1F 0.01 fcrc0686 1F 0.01 seob6535 1F 0.01 mioc1086 1F 0.01 seob0375 1F 0.01 mioa1763 1F 0.01 fcrb8187 1F 0.01 seob6879 1F 0.01 ncr6343 1F 0.01 ncr0898 1F 0.01 ncr3483 1F 0.01 ncrc3313 1F 0.01 ncrc5363 1F 0.01 fcrb2256 1F 0.01 miob4090 1F 0.01 seoa2162 1F 0.01 fcrb7443 1F 0.01 ncrc4798 1F 0.01 miod4759 1F 0.01 seob2797 1F 0.01 fcr1855 1F 0.01 ncrc9867 1F 0.01 mioc2381 1F 0.01 fcrb6106 1F 0.01 fcr5536 1F 0.01 fcrb7703 1F 0.01 fcrc0775 1F 0.01 fcrc0042 1F 0.01 hfcr3089 1F 0.01 hfcr1189 1F 0.01 hfcr6628 1F 0.01 mioa8778 1F 0.01 miob2285 1F O.Ol mioa8970 1F 0.01 ncrc2600 1F 0.01 miob7290 1F 0.01 seoa4174 1F 0.01 miob9248 1F 0.01 seoa7478 1F 0.01 mioc2039 1F 0.01 seob9241 1F 0.01 miod2641 1F 0.01 seoc7498 1F 0.01 ncrb5940 1F 0.01 seoa9724 1F 0.01 ncrb8319 1F 0.01 seob7309 1F 0.01 seob2689 1F 0.01 miob8812 1F 0.01 seob5032 1F 0.01 seob1844 1F 0.01 seob7946 1F 0.01 seob5899 1F 0.01 ncr2926 1F 0.01 ncr4180 1F 0.01 mioa3331 1F 0.01 seob6882 1F 0.01 "~mioa2~~~5~'~~F " """'0':'0'7.miod4521 1F 0 "~" ' .

fcrb6301 1F 0.01 ncr2381 1F 0.01 fcr0608 1F 0.01 ncr4790 1F 0.01 fcr5354 1F 0.01 ncrb8102 1F 0.01 ncr3496 1F 0.01 ncrc2685 1F 0.01 mioc6204 1F 0.01 ncrc5150 1F 0.01 ncrb6327 1F 0.01 seoal736 1F 0.01 ncrcl665 1F 0.01 seob0547 1F 0.01 fcrb7380 1F 0.01 seob7500 1F 0.01 fcrb3135 1F 0.01 seob7722 1F 0.01 ncrc0743 1F 0.01 seoc2131 1F 0.01 seob2959 1F 0.01 miob8565 1F 0.01 hfcr1115 1F 0.01 fcr6057 1F 0.01 ncr3148 1F 0.01 fcrc5233 1F 0.01 mioa6580 1F 0.01 mioa8919 1F 0.01 seoa7926 1F 0.01 fcrb5901 1F 0.01 seoc0355 1F 0.01 ncrb0384 1F 0.01 mioa6552 1F 0.01 fcrb2792 1F 0.01 ncrc4539 1F 0.01 seob6279 1F 0.01 seoa5302 1F 0.01 mioc4730 1F 0.01 seocl791 1F 0.01 seoa2209 1F 0.01 miod4564 1F 0.01 seoc0780 1F 0.01 mioc3682 1F 0.01 seob0534 1F 0.01 fcrb2318 1F 0.01 mioa6135 1F 0.01 fcrc4456 1F 0.01 seoc0924 1F 0.01 mioc3716 1F 0.01 seoa1644 1F 0.01 ncr0420 1F 0.01 seoa1263 1F 0.01 ncrb4428 1F 0.01 mioc4022 1F 0,01 miob4352 1F 0.01 seoa6867 1F 0.01 seob8261 1F 0.01 mioa2970 1F 0.01 miobo442 1F 0.01 fcrb6102 1F 0.01 hfcr2657 1F 0.01 mioal388 1F 0.01 miod'74211F 0.01 mioa1921 1F 0.01 mioc7895 1F 0.01 fcrc1090 1F 0.01 mioc2110 1F 0.01 mioc4100 1F 0.01 ncrc0696 IF 0.01 seob0008 1F 0.01 fcrb8321 IF 0.01 fcrb5343 1F 0.01 seoc3854 1F 0.01 fcrb6279 1F 0.01 fcr5392 IF 0.01 miod4176 1F 0.01 miod2334 lF 0.01 fcrc6335 1F 0.01 miod4355 1F 0.01 fcr6279 1F 0.01 mioa3668 1F 0.01 fcrb1618 1F 0.01 fcrc7047 IF 0.01 fcrb6107 1F 0.02 miob3084 IF 0.01 mioa2173 1F 0.01 seob9898 1F 0.01 mioa5508 1F 0.01 mioc4575 IF 0.01 mioc3669 1F 0.01 seoa5658 IF 0.01 mioc3994 1F 0.01 seob3734 1F 0.01 mioc5039 1F 0.01 fcrb2596 1F 0.01 mioc6973 1F 0.01 mioa4318 1F 0.01 miod7440 1F 0.01 seoa3408 1F 0.01 ncrb8693 1F 0.01 miod1021 1F 0.01 ncrc9217 1F 0.01 ncrc3068 1F 0.01 seoa0014 IF 0.01 seob8204 IF 0.01 seob4676 1F 0.01 mioa2537 1F 0.01 seob7569 1F 0.01 fcrb3651 IF 0.01 seob9368 1F 0.01 fcrb2208 1F 0.01 fcr2303 1F 0.01 hfcr5023 1F 0.01 ncrc0393 1F 0.01 fcr1182 1F 0.01 mioa6035 1F 0.01 fcrb8949 1F 0.01 mioc3576 1F 0.01 hfcr2984 IF 0.01 fcrb8516 1F 0.01 hfcr5987 1F 0.01 mioa5836 1F 0.01 miob1561 1F 0.01 mioc4032 1F 0.01 mioc8437 1F 0.01 seoc4093 1F 0.01 'hfcr551~4' ""'""'l7':"0'1seob0992 1F 0.02 "" ""~TF' ~

mioc75091F 0.01 seob8562 1F 0.02 seoc68621F 0.01 seoc2589 1F 0.02 fcrc53441F 0.01 fcr2306 1F 0.02 mioc78181F 0.01 miod3197 1F 0.02 hfcr27701F 0.01 ncr3197 1F 0.02 ncrb37021F 0.01 seoc6030 1F 0.02 fcrb34661F 0.01 mioc1707 1F 0.02 ncrb41821F 0.01 miob4760 1F 0.02 hfcr37381F 0.01 miob3953 1F 0.02 fcr7318 1F 0.01 seoc8175 1F 0.02 ncrb75611F 0.01 seoa5784 1F 0.02 ncrc02881F 0.01 ncrc4732 1F 0.02 ncrc45081F 0.01 mioa4975 1F 0.02 seoc62661F 0.01 ncrb6261 1F 0.02 ncrc48081F 0.01 ncrc1402 1F 0.02 ncr7666 1F 0.01 fcrb6028 1F 0.02 seob34191F 0.01 fcrc0357 1F 0.02 fcrc05291F 0.01 seoa7530 1F 0.02 mioa73611F 0.01 fcr5571 1F 0.02 fcrcl7451F 0.01 fcrc0415 1F 0.02 miod38271F 0.01 miod5123 1F 0.02 ncr6925 1F 0.01 seoa7546 1F 0.02 seoa71571F 0.01 fcr1844 1F 0.02 seoc52281F 0.01 fcrb7458 1F 0.02 ncrb60731F 0.01 fcrc3600 1F 0.02 ncrc40151F 0.01 seoa3665 1F 0.02 ncr0806 1F 0.01 fcr0768 1F 0.02 fcrc40541F 0.01 hfcr5045 1F 0.02 fcrc53511F 0.01 fcrb8333 1F 0.02 ncrc46201F 0.01 fcrc5290 1F 0.02 seoc66961F 0.01 hfcr0521 1F 0.02 seob93021F 0.01 mioa0192 1F 0.02 fcrc60891F 0.01 miob4574 1F 0.02 ncrb42481F 0.01 miob6029 1F 0.02 ncr5473 1F 0.01 ncrb0323 1F 0.02 fcrb60621F 0.01 ncrb0487 1F 0.02 fcr41&0 1F 0.01 ncrb7516 1F 0.02 fcrc42441F 0.01 ncrc3121 1F 0.02 miod12651F 0.02 ncrc4089 1F 0.02 mioa88581F 0.02 seoa2051 1F 0.02 seob21881F 0.02 seoa7126 1F 0.02 fcrb56441F 0.02 seob0185 1F 0.02 seob97501F 0.02 seob9420 1F 0.02 seob37471F 0.02 ncrc2472 1F 0.02 fcrb11201F 0.02 seob5652 1F 0.02 fcrb72071F 0.02 mioa1055 1F 0.02 miod74291F 0.02 mioa1891 1F 0.02 mioc76861F 0.02 ncrc0259 1F 0.02 fcr1557 1F 0.02 fcrc0569 1F 0.02 fcr3005 1F 0.02 fcrb1901 1F 0.02 fcrbl6781F 0.02 miob8515 1F 0.02 fcrc61961F 0.02 seoc1490 1F 0.02 hfcr40071F 0.02 mioa1392 1F 0.02 miob71351F 0.02 miod1718 1F 0.02 mioc27281F 0.02 miod6090 1F 0.02 ncr0213 1F 0.02 miob9519 1F 0.02 ncrc24151F 0.02 ncr3165 1F 0.02 seoa16161F 0.02 miob4864 1F 0.02 seoa26901F 0.02 ncr7967 1F 0.02 seoa46471F 0.02 fcrc1209 1F 0.02 seoa54731F 0.02 fcrb6084 1F 0.02 seoa71781F 0.02 seoa4783 1F 0.02 seoa84011F 0.02 mioc0090 1F 0.02 19~

mioc729--6' """'0":"0'2 ncrc5025 1F 0.02 TF

mioal632 1F 0.02 seoa4012 1F 0.02 fcrb4109 1F 0.02 fcr7424 1F 0.02 ncrc0445 1F 0.02 fcrb7036 1F 0.02 miocl249 lF 0,02 mioa3239 lF 0.02 fcr4782 1F 0.02 miob2355 1F 0.02 ncr7973 1F 0.02 miob9073 1F 0.02 ncrb0782 1F 0.02 miocl524 1F 0.02 ncrc2273 1F 0.02 miod2570 1F 0.02 mioc4667 1F 0.02 miod3347 1F 0.02 hfcr0501 1F 0,02 ncrb0864 1F 0.02 seoc0502 1F 0,02 ncrb2400 1F 0.02 mioc7904 1F 0.02 ncrc2807 1F 0.02 fcrb2933 1F 0,02 seoa5253 1F 0.02 fcrb6968 1F 0.02 seob2938 1F 0.02 fcrb7780 1F 0.02 seoc5218 1F 0.02 fcrc5139 1F 0.02 ncrc6861 1F 0.02 fcrc5763 1F 0.02 ncr3045 1F 0.02 mioa1979 1F 0.02 mioc4161 1F 0.02 mioa6252 1F 0.02 fcrb6651 1F 0.02 miob6536 1F 0.02 fcrc1014 1F 0.02 miob9441 1F 0.02 ncrc2018 lF 0.02 mioc3683 1F 0.02 fcrc1207 1F 0.02 mioc4319 1F 0.02 fcrb6236 1F 0.02 miod5030 1F 0,02 mioc7441 1F 0.02 ncr0083 1F 0.02 ncr4648 1F 0.02 ncrc0798 1F 0.02 fcrc0096 1F 0.02 ncrc4759 1F 0.02 mioc8682 1F 0.02 seob2661 1F 0.02 seoal789 1F 0.02 seob7404 1F 0.02 ncr5484 1F 0.02 seoc1476 1F 0.02 fcr4804 1F 0.02 mioa4667 1F 0.02 fcrc5060 1F 0.02 mioc3139 1F 0,02 ncrc9284 1F 0.02 ncr8686 2F 0,02 fcr5663 1F 0.02 seoa1041 1F 0.02 miob8707 1F 0.02 seoa5731 1F 0.02 ncr3397 1F 0.02 fcrb2156 1F 0.02 seob5894 1F 0.02 miob7391 1F 0.02 miod2834 1F 0.02 mioc2394 1F 0.02 fcrb5812 1F 0.02 hfcr2890 1F 0.02 ncr0210 1F 0.02 mioa0647 1F 0.02 ncr3149 1F 0.02 seoa3109 1F 0.02 miod3027 1F 0.02 mioc7774 1F 0.02 fcrb8430 1F 0.02 fcrb9671 1F 0.02 miod6134 1F 0.02 ncr6415 1F 0.02 seoc2173 1F 0.02 ncrb6557 1F 0.02 seoa5800 1F 0.02 seob8501 1F 0.02 fcrc5547 1F 0.02 seob4057 1F 0.02 fcrb2150 1F 0.02 ncrc1595 1F 0.02 fcrb4988 1F 0.02 fcr7646 1F 0.02 fcrb6663 1F 0.02 miob9325 1F 0.02 fcrcS107 1F 0.02 ncrc9978 1F 0.02 fcrc6970 1F 0.02 fcrb2396 1F 0.02 mioa2528 1F 0.02 fcrb5351 1F 0.02 miob9284 1F 0.02 fcrb5534 1F 0.02 miob9817 1F 0.02 seob3594 1F 0.02 mioc2828 1F 0.02 miob1873 1F 0.02 miod0455 1F 0.02 mioc6878 1F 0.02 miod2977 1F 0.02 seoa9501 1F 0.02 ncr3843 1F 0.02 seob9292 1F 0.02 ncr9934 1F 0.02 fcrb6986 1F 0.02 ncrc1102 1F 0.02 seob5004 1F 0.02 seob5767 1F 0.02 fcrb8674 1F 0.02 seob7278 1F 0.02 fcrcl298 1F 0.02 seob9092 1F 0.02 seocl7641F 0.-02 mioc1992 1F 0.02 mioa16211F 0.02 seoa0913 1F 0.02 fcr3181 1F 0.02 seoa7296 1F 0.02 miob80961F 0.02 seoc0098 1F 0.02 hfcr26701F 0.02 seoa4518 1F 0.02 mioc89341F 0.02 ncrc6841 1F 0.02 hfcr59591F 0.02 fcrbl562 1F 0.02 fcr5625 1F 0.02 miob6702 1F 0.02 seoa14391F 0.02 mioc6274 1F 0.03 fcrb56031F 0.02 ncrb6675 1F 0.03 miob09741F 0.02 fcrb4232 1F 0.03 miob63551F 0.02 fcrb5164 1F 0.03 fcrb50021F 0.02 mioa1025 1F 0.03 mioc35651F 0.02 miob0942 1F 0.03 ncrc27011F 0.02 miob4956 1F 0.03 hfcr29551F 0.02 mioc2619 1F 0.03 hfcr11771F 0.02 ncr1055 1F 0.03 fcrb00441F 0.02 ncr2695 1F 0.03 seoc56621F 0.02 ncrb4025 1F 0.03 mioc52401F 0.02 ncrc3104 1F 0.03 hfcr12381F 0.02 ncrc5088 1F 0.03 fcr1347 1F 0.02 ncrc5947 1F 0.03 ncr6920 1F 0.02 ncrc9704 1F 0.03 seoa11171F 0.02 seoa2900 1F 0.03 hfcr25471F 0.02 seoa7129 1F 0.03 ncrb25581F 0.02 seob0263 1F 0.03 miod68011F 0.02 ncrb6949 1F 0.03 fcrb61411F 0.02 fcrb8447 1F 0.03 hfcr11411F 0.02 seoa3514 1F 0.03 hfcr53831F 0.02 seoc1642 1F 0.03 mioa45481F 0.02 seoc4756 1F 0.03 miob34741F 0.02 seoa0024 1F 0.03 miob4~731F 0.02 fcrb5438 1F 0.03 miob65981F 0.02 seoc4909 1F 0.03 mioc16401F 0.02 miob1563 1F 0.03 ncr3123 1F 0.02 ncrc2463 1F 0.03 ncr3782 1F 0.02 seoa7078 1F 0.03 ncrc13261F 0.02 seob6467 1F 0.03 seoa10831F 0.02 mioc6075 1F 0.03 seoa24021F 0.02 mioc4077 1F 0.03 seoa35441F 0.02 seob9730 1F 0.03 seoa97091F 0.02 fcrb7237 1F 0.03 seob27751F 0.02 fcrc7010 1F 0.03 seob95521F 0.02 ncrb4015 1F 0.03 seoc00561F 0.02 fcrc5372 1F 0.03 fcrb27601F 0.02 seoa8738 1F 0.03 fcrc52971F 0.02 seob6198 1F 0.03 mioa73171F 0.02 miod4370 1F 0.03 seoa15011F 0.02 fcr4900 1F 0.03 miod27861F 0.02 seob6028 1F 0.03 fcrb45421F 0.02 seob4087 1F 0.03 ncrc08831F 0.02 fcr3528 1F 0.03 hfcr64061F 0.02 ncrc3141 1F 0.03 ncrc55081F 0.02 ncrb8404 1F 0.03 ncrc56081F 0.02 ncrc5327 1F 0.03 ncrb00901F 0.02 miob9678 1F 0.03 seob50441F 0.02 fcrc5379 1F 0.03 seoc51341F 0.02 ncr4378 1F 0.03 fcr5472 1F 0.02 miob1178 1F 0.03 mioa50971F 0.02 mioal532 1F 0.03 ncrc34531F 0.02 seob6386 1F 0.03 miob85861F 0.02 fcr4846 1F 0.03 miod48571F 0.02 fcrb4504 1F 0.03 mioc06491F 0.02 mioa0840 1F 0.03 mi o '"""1""' iT': -0'3n c r c 5 8 4 1 0 a 2 F~ - 4 F .
3 2"7 0 ""~ 3 mioa47701F 0.03 seoal480 1F 0.03 mioc31271F 0.03 seoa4670 1F 0.03 mioc39621F 0.03 seoa4675 1F 0.03 mioc63601F 0.03 seoa5094 1F 0.03 miod11571F 0.03 seoa8642 1F 0.03 miod15421F 0.03 seob0063 1F 0.03 ncr1780 1F 0.03 seob6177 1F 0.03 ncrc62641F 0.03 miod0228 1F 0.03 ncrc90441F 0.03 fcrb6185 1F 0.03 ncrc92371F 0.03 ncrb2053 1F 0.03 seoa28011F 0.03 fcrb2430 1F 0.03 seob12681F 0.03 ncrb2027 1F 0.03 seob60871F 0.03 fcr6386 1F 0.03 fcr4722 1F 0.03 ncrc0715 1F 0.03 fcrb93241F 0.03 seob6670 1F 0.03 mioa34671F 0.03 fcrb3283 1F 0.03 ncrb73291F 0.03 seoa0064 1F 0.03 fcr4795 1F 0.03 mioa3588 1F 0.03 fcr5141 1F 0.03 seoa1615 1F 0.03 seob65851F 0.03 ncrb7386 1F 0.03 ncr3690 1F 0.03 mioc4895 1F 0.03 seob50331F 0.03 fcrb6596 1F 0.03 fcrb35441F 0.03 fcrc5831 1F 0.03 fcrc28071F 0.03 seoc6004 1F 0.03 seoa00441F 0.03 fcrb2120 1F 0.03 mioc35931F 0.03 ncrc2439 1F 0.03 fcrb26971F 0.03 seoc1495 1F 0.03 hfcr50031F 0.03 seoa3863 1F 0.03 mioc70841F 0.03 ncrb3585 1F 0.03 seoc22211F 0.03 ncr9549 1F 0.03 seob50211F 0.03 fcrb7593 1F 0.03 fcrb33091F 0.03 fcrb9481 1F 0.03 fcrcl8281F 0.03 fcrc2222 1F 0.03 ncrc48511F 0.03 fcrc6084 1F 0.03 fcr0824 1F 0.03 fcrc7214 1F 0.03 mioa43261F 0.03 hfcr5706 1F 0.03 seoa56851F 0.03 mioa4037 1F 0.03 seoa93771F 0.03 mioa5468 1F 0.03 miob02021F 0.03 mioa6102 1F 0.03 fcrb59291F 0.03 mioa6734 1F 0.03 seob66251F 0.03 mioa8074 1F 0.03 ncr0019 1F 0.03 miob1139 1F 0.03 fcrc51371F 0.03 miob3044 1F 0.03 fcr1914 1F 0.03 miob5647 1F 0.03 fcrb17751F 0.03 mioc2204 1F 0.03 ncrc69251F 0.03 ncr3588 ~1F 0.03 ncrb33481F 0.03 ncr4040 1F 0.03 ncrc15371F 0.03 ncr8725 1F 0.03 seoa39081F 0.03 ncr9469 1F 0.03 ncr9003 1F 0.03 ncrb8203 1F 0.03 seoc60991F 0.03 ncrc0508 1F 0.03 fcr1060 1F 0.03 seoa24~43 1F 0.03 fcr3559 1F 0.03 seoa3429 1F 0.03 fcr4494 1F 0.03 seoa3701 1F 0.03 fcrb21901F 0.03 seoa7366 1F 0.03 fcrc25981F 0.03 seoa9711 1F 0.03 hfcr46401F 0.03 seob8914 1F 0.03 miob06361F 0.03 seoc2136 1F 0.03 miob32341F 0.03 mioc4472 1F 0.03 miod19081F 0.03 seoc0284 1F 0.03 ncrb41661F 0.03 ncrb1247 1F 0.03 ncrc18891F 0.03 hfcr2664 1F 0.03 ncrc31711F 0.03 miod4467 1F 0.03 ~fcrb86'9-0"""""'1F"""0": f3 seoa6364 1F 0.03 "

seoa69301F 0.03 mioc3137 1F 0.03 seob41451F 0.03 ncrc8903 1F 0.03 fcr3367 1F 0.03 fcrb3715 1F 0.03 fcr6630 1F 0.03 fcr5190 1F 0.04 miod6292lF 0.03 seoc5467 1F 0.04 hfcr62761F 0.03 mioa2783 lF 0.04 ncr3751 1F 0.03 ncr3219 1F 0.04 fcrb52721F 0.03 seoa9828 1F 0.04 ncrc57161F 0.03 miob9529 1F 0.04 hfcr28211F 0.03 miob8694 1F 0.04 fcrb3718lF 0.03 seoa8300 1F 0.04 fcrb53461F 0.03 seob0154 1F 0.04 ncrc88811F 0.03 ncrc1380 1F 0.04 miob06811F 0.03 hfcr0370 1F 0.04 ncrc88921F 0.03 fcrb7255 1F 0.04 mioa33351F 0.03 fcrb6208 1F 0.04 ncrb84961F 0.03 ncrc5959 1F 0.04 mioc77811F 0.03 fcr5316 1F 0.04 seob18341F 0.03 seob4579 lF 0.04 ncrc19951F 0.03 miod4367 lF 0.04 mioc19711F 0.03 fcr2442 1F 0.04 mioc23601F 0.03 fcr7561 1F 0.04 ncrcl7511F 0.03 fcrb1916 1F 0.04 seoa0860lF 0.03 fcrb3518 1F 0.04 seoa37371F 0.03 fcrb4231 1F 0.04 ncrc65871F 0.03 fcrc6997 1F 0.04 seoa46061F 0.03 hfcr0439 1F 0.04 seoa22441F 0.03 hfcr0489 1F 0.04 fcrb69491F 0.03 hfcr2584 1F 0.04 mioc20211F 0.03 mioa0601 1F 0.04 seob08721F 0.03 mioa0869 1F 0.04 ncr3118 1F 0.03 mioa2079 1F 0.04 miod60321F 0.03 mioa7069 1F 0.04 hfcr10371F 0.03 miob9714 1F 0.04 fcr4471 lF 0.03 mioc1125 1F 0.04 fcr3599 1F 0.03 miod4407 1F 0.04 fcrb18541F 0.03 ncr3718 1F 0.04 fcrc48411F 0.03 ncr7093 1F 0.04 hfcr36151F 0.03 ncrb0033 1F 0.04 mioa89981F 0.03 ncrc6825 1F 0.04 mioc02061F 0.03 ncrc7029 1F 0.04 mioc33001F 0.03 seoa0207 1F 0.04 mioc61141F 0.03 seoa9160 1F 0.04 miod51261F 0.03 seobl196 1F 0.04 ncr5709 1F 0.03 seobl586 1F 0.04 ncrb34451F 0.03 seob8710 1F 0.04 ncrc31001F 0.03 seoc5888 1F 0.04 seoa54551F 0.03 ncr2175 1F 0.04 seoa55201F 0.03 fcrc4188 1F 0.04 seob88071F 0.03 seoa6144 1F 0.04 seoc41871F 0.03 fcrb1741 1F 0.04 fcrb53391F 0.03 ncrb3638 1F 0.04 ncrc70381F 0.03 seoa2219 1F 0.04 fcrb37021F 0.03 ncrc2128 1F 0.04 fcrc48141F 0.03 miod1236 1F 0.04 fcr4984 1F 0.03 seoc4038 1F 0.04 mioa72581F 0.03 seoa8521 1F 0.04 seob68441F 0.03 fcrc3640 1F 0.04 fcrb87621F 0.03 fcrc5699 1F 0.04 seob06741F 0.03 fcrb4781 1F 0.04 mioc77641F 0.03 miob8249 1F 0.04 miob09311F 0.03 ncrc3953 1F 0.04 ncrb63781F 0.03 ncr2994 1F 0.04 "iriiob86'5'T"''"'T'~'' ""''0':'0'4' fcrc4180 1F 0.04 fcrb6281lF 0.04 seoc1073 1F 0.04 ncrb55951F 0.04 seob8693 1F 0.04 fcrbl5731F 0.04 fcrbl344 1F 0.04 seob16171F 0.04 mioa8973 1F 0.04 seob17571F 0.04 fcrb5181 1F 0.04 fcrb20341F 0.04 fcrc0839 1F 0.04 fcrb57631F 0.04 seoc0009 1F 0.04 fcrb78291F 0.04 fcrb6640 1F Ø04 mioal5131F 0.04 ncrc5907 1F 0.04 ncrb65811F 0.04 fcrb2624 1F 0.04 miod28371F 0.04 seoc0619 1F 0.04 miob60981F 0.04 seob5889 1F 0.04 seoa66211F 0.04 fcrb2196 1F 0.04 ncrc69201F 0.04 fcrb4985 1F 0.04 mioc18081F 0.04 fcrb8093 1F 0.04 ncrb63371F 0.04 fcrb8614 1F 0.04 ncrcl2471F 0.04 fcrc0112 1F 0.04 mioc09091F 0.04 hfcr3091 lF 0.04 miod08781F 0.04 hfcr6l10 1F 0.04 mioc02401F 0.04 ncr3686 1F 0.04 ncr0612 1F 0.04 ncrc1143 1F 0.04 seoa00401F 0.04 ncrc2949 1F 0.04 fcr0253 1F 0.04 ncrc6382 1F 0.04 fcr0186 1F 0.04 seoa8716 1F 0.04 fcr2218 1F 0.04 seoa9082 1F 0.04 fcrb48171F 0.04 seob4333 1F 0.04 fcrb67961F' 0.04 seob4612 1F 0.04 fcrb69391F 0.04 seoc4288 1F 0.04 fcrb77481F 0.04 mioa9127 1F 0.04 mioa47211F 0.04 seob0261 1F 0.04 mioa88361F 0.04 mioa3469 1F 0.04 mioc53361F 0.04 ncrc7005 1F 0.04 ncr0570 1F 0.04 seob4172 1F 0.04 ncrc09361F 0.04 seob1093 1F 0.04 ncrc39361F 0.04 fcrb3160 1F 0.04 ncrc47401F 0.04 ncrb8383 1F 0.04 ncrc52071F 0.04 fcrb7528 1F 0.04 seoa15841F 0.04 mioa8539 1F 0.04 seoa32511F 0.04 mioc3081 1F 0.04 seoa72501F 0.04 fcrc0271 1F 0.04 seoa76471F 0.04 ncrb8201 1F 0.04 seob16601F 0.04 seoc2504 1F 0.04 seob55581F 0.04 mioc7854 1F 0.04 seob88531F 0.04 seob3455 1F 0.04 ncrb39031F 0.04 ncr1699 1F 0.04 ncrc50161F 0.04 seoc4513 1F 0.04 mioa79551F 0.04 mioc6298 1F 0.04 hfcr31431F 0.04 miob6459 1F 0.04 seoa19241F 0.04 fcrb7803 1F 0.04 ncrb14381F 0.04 mioc4996 1F 0.04 seoa97121F 0.04 seoa2936 1F 0.04 fcrc41251F 0.04 fcrb1380 1F 0.04 ncrc32831F 0.04 fcrb5294 1F 0.04 seoc57801F 0.04 ncrc6459 1F 0.04 ncr3483 1F 0.04 miob8711 1F 0.04 seoc09991F 0.04 seoa3230 1F 0.04 ncrc92861F 0.04 miob0496 1F 0.04 mioc86711F 0.04 mioc6925 1F 0.04 mioa45421F 0.04 seoa9792 1F 0.04 mioal6871F 0.04 seob6008 1F 0.04 mioa41961F 0.04 seoa9046 1F 0.04 seob21611F 0.04 seob4515 1F 0.04 miob76271F 0.04 seoa4202 1F 0.04 riiiola47~:'2..~............ØØ4., seob0294 1G 4 .757821e-03 ~.~,..

seoa55281F 0.04 fcr4746 1G 5.04282e-03 ncrcl7611F 0.04 ncr1355 1G 5.04282e-03 miod33661F 0.04 miod5957 1G 5.342732e-03 fcrb84951F 0.04 ncrb0328 lG 5.342732e-03 fcrc04871F 0.04 seoa3639 1G 5.342732e-03 hfcr05941F 0.04 seoc2646 1G 5.342732e-03 mioa07761F 0.04 ncrb5117 1G 5.497749e-03 mioa70151F 0.04 fcr6635 1G 5.658209e-03 mioa71691F 0.04 fcrb3808 1G 5.658209e-03 mioa87731F 0.04 fcrc7087 1G 5.658209e-03 miob79381F 0.04 miob0189 1G 5.658209e-03 miod60251F 0.04 mioc7372 1G 5.658209e-03 ncr8628 1F 0.04 ncr8606 1G 5.989924e-03 ncrb39801F 0.04 fcrc4692 1G 6.33857e-03 ncrc27631F 0.04 miob4574 1G 6.33857e-03 ncrc40001F 0.04 ncrc4757 1G 6.33857e-03 ncrc51621F 0.04 hfcr2756 1G 6.704863e-03 seoa53661F 0.04 miob9030 1G 6.704863e-03 seob88171F 0.04 seoa2162 1G 6.704863e-03 ncr7941 1F 0.04 fcrb6211 1G 7.089543e-03 mioa36461F 0.04 hfcr5260 1G 7.089543e-03 seoc21911F 0.04 mioa5355 1G 7.089543e-03 fcrb20151F 0.04 ncr5975 1G 7.493369e-03 miod60581F 0.04 fcrb2624 1G 7.917127e-03 ncrc34681F 0.04 fcrc5372 1G 7.917127e-03 seob76581F 0.04 ncr2861 1G 7.917127e-03 ncrc69811F 0.04 seoa3121 1G 7.917127e-03 fcrb43911F 0.04 hfcr5003 1G 8.361625e-03 miod66091F 0.04 mioa9357 1G 8.361625e-03 mioc84791G 1.54e-05 ncrb8539 1G 8.361625e-03 fcrc05591G 3.01e-05 seob1808 1G 8.361625e-03 seoa6.5601G 7.42e-05 fcr4306 1G 8.590878e-03 miob79851G 3.4914e-04 mioa6659 1G 8.827692e-03 miob19461G 4.71848e-04 miob4413 1G 8.827692e-03 ncr3527 1G 5.88272e-04 ncrc2289 1G 8.827692e-03 fcrb68151G 9.02252e-04 mioc0999 1G 9.316185e-03 mioc62601G 1.110144e-03 seoa0488 1G 9.316185e-03 hfcr66111G 1.188468e-03 fcr2303 1G 9.827982e-03 mioa33991G 1.359944e-03 fcrb1769 1G 9.827982e-03 fcrb98511G 1.360187e-03 fcrb5100 1G 9.827982e-03 seoc20061G 1.360187e-03 fcrc0345 1G 9.827982e-03 ncr1699 1G 1.454132e-03 fcrc2573 1G 9.827982e-03 ncr1282 1G 1.89084e-03 fcrc5087 1G 9.827982e-03 fcrb50771G 2.016885e-03 miob9336 1G 9.827982e-03 mioc74441G 2.016885e-03 mioc3332 1G 9.827982e-03 miob01671G 2.291704e-03 mioc4747 1G 9.827982e-03 seoa82991G 2.291704e-03 fcr4160 1G 0.01 ncr0761 1G 2.441249e-03 miob9668 1G 0.01 seob30251G 2.766654e-03 fcrb5775 1G 0.01 seob80511G 3.13006e-03 miob9519 1G 0.01 mioa89701G 3.535205e-03 mioc5103 1G 0.01 ncrb01641G 3.535205e-03 ncrc3604 1G 0.01 seob56321G 3.535205e-03 ncrc5500 1G 0.01 seoa22331G 3.754676e-03 fcrb6509 1G 0.01 seob95521G 3.754676e-03 fcrc4408 1G 0.01 mioa06261G 3.986113e-03 miod4220 1G 0.01 fcrb92801G 4.230067e-03 ncr9105 1G 0.01 fcrc61191G 4.230067e-03 ncrc5150 1G 0.01 miob49561G 4.230067e-03 seoa1736 1G 0.01 ncr4860 1G 4.487107e-03 seoa8684 1G 0.01 fcrc45061G 4.757821e-03 hfcr3180 1G 0.01 miob22101G 4.757821e-03 ncr4189 1G 0.01 seoa58331G 4.757821e-03 seob3090 1G 0.01 seob-9574-"" """'c7": cT~: mioc2133 1G 0.
~ 1"G"" 02 seoc41351G 0.01 mioc3430 1G 0.02 miob77651G 0.01 ncrb6282 1G 0.02 seob03041G 0.01 seoa1734 1G 0.02 seoc40781G 0.01 miob5810 1G 0.02 seoc43801G 0.01 mioc5643 1G 0.02 fcrb22181G 0.01 ncr9337 1G 0.02 mioa10151G 0.01 ncrc1602 1G 0.02 mioa16601G 0.01 ncrc4402 1G 0.02 mioa40091G 0.01 seob0483 1G 0.02 mioa54611G 0.01 fcrc5041 1G 0.02 mioa97921G 0.01 mioc7471 1G 0.02 mioc35231G 0.01 miodl292 1G 0.02 ncrc38561G 0.01 ncrc0174 1G 0.02 seocl3481G 0.01 seoa4457 1G 0.02 seoc44841G 0.01 seob5748 1G 0.02 ~

miob93401G 0.01 fcrb6188 1G 0.02 mioc0302lG 0.01 mioa9630 1G 0.02 fcr2798 1G 0.01 miod0775 1G 0.02 fcrc55091G 0.01 ncrb6530 1G 0.02 mioa25221G 0.01 seoa5433 1G 0.02 fcrb28181G 0.01 seoa6695 1G 0.02 fcrc07271G 0.01 seob9734 1G 0.02 mioc02271G 0.01 fcr2598 1G 0.02 ncrb82031G 0.01 fcrb6281 1G 0.02 miod12001G 0.01 fcrb7965 1G 0.02 seob74091G 0.01 fcrc7243 1G 0.02 fcrb70631G 0.01 miob5969 1G 0.02 hfcr64861G 0.01 miob6245 1G 0.02 miob01801G 0.01 ncrc3408 1G 0.02 miob90681G 0.01 fcrb7861 1G 0.02 ncrb28701G 0.01 fcrc0967 1G 0.02 seoa37011G 0.01 mioa3856 1G 0.02 fcrb17411G 0.01 miob9284 1G 0.02 mioa56141G 0.01 mioc7542 1G 0.02 mioc37161G 0.01 miod4129 1G 0.02 mioc39581G 0.01 ncr8171 1G 0.02 ncr8041 1G 0.01 ncrb0513 1G 0.02 ncrb82071G 0.01 seoa5848 1G 0.02 ncrc93041G 0.01 seoa7383 1G 0.02 seoc33211G 0.01 seob4612 1G 0.02 fcrb47121G 0.01 seoc2681 1G 0.02 ncrc00971G 0.01 fcrc1043 1G 0.03 seoc72031G 0.01 fcrc2683 1G 0.03 mioa80321G 0.01 hfcr3902 1G 0.03 miob98821G 0.01 mioa0909 1G 0.03 seoc10771G 0.01 mioc8694 1G 0.03 miobl5591G 0.02 ncrc2618 1G 0.03 hfcr51571G 0.02 seoa4422 1G 0.03 mioc49831G 0.02 seoc3302 1G 0.03 mioc76651G 0.02 fcrb5503 1G 0.03 mioc84231G 0.02 fcrb6536 1G 0.03 seob80921G 0.02 fcrb7645 1G 0.03 seoc22241G 0.02 fcrc1563 1G 0.03 fcr4308 1G 0.02 fcrc2254 1G 0.03 mioal6741G 0.02 miob6354 1G 0.03 mioc82571G 0.02 mioc1971 1G 0.03 miod66711G 0.02 mioc5664 1G 0.03 ncrc06671G 0.02 ncrc2404 1G 0.03 fcr3525 1G 0.02 seoa3891 1G 0.03 fcrb45791G 0.02 seoa9729 1G 0.03 mioa90331G 0.02 seob1268 1G 0.03 miob80261G 0.02 seob2994 1G 0.03 miob85831G 0.02 seob5214 1G 0.03 seoc09571G "' 0": 0'3 seoa4366 1G 0 . 04 fcr3714 1G 0.03 seob6l56 1G 0.04 mioc36031G 0.03 fcrc2817 1G 0.04 ncr0612 1G 0.03 hfcr5220 1G 0.04 ncr0701 1G 0.03 ncrb5192 1G 0.04 ncrc04611G 0.03 seoa2087 1G 0.04 seob53421G 0.03 seoa3002 1G 0.04 miob58551G 0.03 seob0344 1G 0.04 mioc63741G 0.03 seocl561 1G 0.04 miod07081G 0.03 fcr2821 1G 0.04 miod28861G 0.03 fcrb4367 1G 0.04 ncrb30111G 0.03 fcrb5000 1G 0.04 seob21551G 0.03 fcrc0112 1G 0.04 seob45371G 0.03 fcrc0591 1G 0.04 seob92921G 0.03 mioa2374 1G 0.04 seoc12341G 0.03 miod7243 1G 0.04 seoc41371G 0.03 ncr2486 1G 0.04 fcrb81141G 0.03 ncrc9237 1G 0.04 fcrc00751G 0.03 seoa8443 1G 0.04 fcrc20131G 0.03 seob0918 1G 0.04 rnioal5321G 0.03 fcrb2633 1G 0.04 miob22931G 0.03 fcrb5903 1G 0.04 mioc23691G 0.03 hfcr2544 1G 0.04 miod74601G 0.03 miob4091 1G 0.04 ncr0673 1G 0.03 miob7319 1G 0.04 ncrc93431G 0.03 mioc4009 1G 0.04 seoa36281G 0.03 miod1908 1G 0.04 seobl1451G 0.03 ncrc1031 1G 0.04 seob51131G 0.03 seoa3633 1G 0.04 seob52191G 0.03 seob9674 1G 0.04 seob88171G 0.03 seoc3554 1G 0.04 seoc22951G 0.03 fcrb5l46 1G 0.04 fcr5509 1G 0.03 miob9010 1G 0.04 fCr6228 1G 0.03 mioc6946 1G 0.04 fcrb23181G 0.03 miod0455 1G 0.04 fcrb34741G 0.03 miod2211 1G 0.04 fcrb52591G 0.03 ncrc0262 1G 0.04 fcrc27871G 0.03 ncrc3093 1G 0.04 hfcr19681G 0.03 ncrc4384 1G 0.04 miod31321G 0.03 seoa1480 1G 0.04 ncr8538 1G 0.03 seoa2899 1G 0.04 ncr8594 1G 0.03 seoa6118 1G 0.04 ncrc01851G 0.03 seoa6144 1G 0.04 ncrc15311G 0.03 miob8143 1H 9.6e-05 ncrc33241G 0.03 fcrb6l91 1H 2.18748e-04 seoa40121G 0.03 fcrb4995 1H 5.08096e-04 fcrb24621G 0.03 hfcr4477 1H 5.46853e-04 fcrb43601G 0.03 miob7209 1H 7.83893e-04 fcrb59481G 0.03 fcrb9202 1H 9.02252e-04 miob47601G 0.03 ncr2472 1H 9.02252e-04 miob77941G 0.03 miob7554 1H 9.67277e-04 mioc43511G 0.03 fcrb3330 1H 1.188468e-03 ncr2926 1G 0.03 seoc2191 1H 1.188468e-03 ncrb13651G 0.03 miob4308 1H 1.659667e-03 seoa53921G 0.03 fcrb2162 1H 2.15038e-03 seob18891G 0.03 mioc2451 1H 2.15038e-03 seob64461G 0.03 miob8572 1H 2.291704e-03 seocl2781G 0.03 fcrc6345 1H 2.441249e-03 fcrc06611G 0.04 mioc1203 1H 2.441249e-03 hfcr12591G 0.04 seob3112 1H 2.441249e-03 hfcr44971G 0.04 seoc3965 1H 2.599424e-03 miod72461G 0.04 fcrb7944 1H 3.13006e-03 ncr3233 1G 0.04 fcrc2807 1H 3.13006e-03 ncrc53591G 0.04 hfcr1811 1H 3.32717e-03 'miod~70'3""'"1'Ti"''"'"3'":3'271"T'e-03fcrc5850 1H 0.02 ""' ncr8628 1H 3.986113e-03 fcrb4413 1H 0.03 ncrb3957 1H 3.986113e-03 ncr8413 1H 0.03 seob8741 1H 3.986113e-03 seob4545 1H 0.03 fcr7042 1H 4.487107e-03 fcr1312 1H 0.03 miod7421 1H 5.04282e-03 fcrc6010 1H 0.03 fcr1984 1H 5.342732e-03 mioa9581 1H 0.03 mioc3930 1H 5.342732e-03 mioa9649 1H 0.03 mioc0728 1H 5.989924e-03 ncr2812 1H 0.03 seob5478 1H 7.089543e-03 seoa8851 1H 0.03 fcrc5614 1H 7.917127e-03 seob3141 1H 0.03 ncrb8343 1H 7.917127e-03 miod5785 1H 0.03 fcr3323 1H 8.361625e-03 mioc1910 1H 0.03 fcrb1428 1H 8.827692e-03 fcrb3897 1H 0.03 mioc0669 1H 8.827692e-03 fcrb9430 1H 0.03 miod5122 1H 8.827692e-03 mioc1060 1H 0.03 ncrc0729 1H 9.316185e-03 miod4066 1H 0.03 fcrb4226 1H 0.01 ncrb5595 1H 0.03 seoa0429 1H 0.01 seob7929 1H 0.03 seob2966 1H 0.01 fcrb4981 1H 0.03 fcrb6031 1H 0.01 fcrc1781 1H 0.03 fcrc0166 1H 0.01 mioc2662 1H 0.03 fcrb2041 1H 0.01 ncrb8105 1H 0.03 fcrb9680 1H 0.01 seoal977 1H 0.03 hfcr3149 1H 0.01 seob0288 1H 0.03 ncr7292 1H 0.01 miob8773 1H 0.03 ncrc6382 1H 0.01 seoa5214 1H 0.03 mioc8153 1H 0.01 fcrb2051 1H 0.04 seob3485 1H 0.01 hfcr5905 1H 0.04 mioc7986 1H 0.01 mioa2073 1H 0.04 seob9406 1H 0.01 seoa3555 1H 0.04 miob2656 1H 0.01 seob0688 1H 0.04 ncrc0217 1H O.Ol fcr7295 1H 0.04 seob5069 1H 0.01 mioa5085 1H 0.04 mioa6442 1H 0.01 ncrc3598 1H 0.04 seoc0945 1H 0.01 ncrc5780 1H 0.04 fcr3664 1H 0.01 seoa5787 1H 0.04 fcrb5813 1H 0.01 fcr4471 1H 0.04 mioc3139 1H 0.01 miob7373 1H 0.04 mioc6937 1H 0.01 ncr5651 1H 0.04 ncrc8892 1H 0.01 ncrc5844 1H 0.04 seoa2641 1H 0.01 fcrc4916 1H 0.04 seob8321 1H 0.01 mioc2074 1H 0.04 fcrc4380 1H 0.02 seob2959 1H 0.04 ncr6142 1H 0.02 seoc1025 1H 0.04 seoa2448 1H 0.02 seoc6182 1H 0.04 seoc0499 1H 0.02 mioa8851 1H 0.04 fcrb9686 1H 0.02 miob9788 1H 0.04 ncr3037 1H 0.02 mioc5695 1H 0.04 seob0497 1H 0.02 seoa9709 1H 0.04 fcrb9324 1H 0.02 seob6535 1I 1.74042e-04 hfcr4007 1H 0.02 ncrc0262 1I 4.7724e-04 mioc2997 1H 0.02 seoc3965 1I 5.85651e-04 ncrb4331 1H 0.02 ncr9105 1I 8.16756e-04 ncrc1885 1H 0.02 seob5962 1I 9.30251e-04 seob4925 1H 0.02 fcrb6202 1I 1.127264e-03 mioc0162 1H 0.02 mioa3629 1I 1.127264e-03 seoa2381 1H 0.02 fcrc3907 1I 1.200832e-03 seoa5554 1H 0.02 ncr2484 1I 1.200832e-03 seoc4720 1H 0.02 ~seob6395 1I 1.200832e-03 hfcr2250 1H 0.02 fcrb8994 1I 1.361048e-03 mioa2377 1H 0.02 seoa1747 1I 1.44814e-03 seob4197 1H 0.02 mioa4014 1I 1.540198e-03 fcrc2670 1H 0.02 seoa5986 1I 1.848665e-03 "~cr0187 1I ~ T:963134e-03 miob7136 1T 0.01 - ' fcrb2759 1I 1.963134e-03 miod0126 12 0.01 mioc5113 1I 1.963134e-03 hfcr5691 1I 0.01 fcrb6005 1I 2.083894e-03 mioa0247 1I 0.01 miob0178 1I 2.083894e-03 mioa8818 1I 0.01 seoa3322 1I 2.083894e-03 mioa8899 1I 0.01 ncrc2831 1I 2.486944e-03 ncrc3072 1T 0.01 ncr5713 1I 2.63595e-03 ncrc4033 1T O.C1 ncrb8538 1I 2.792845e-03 fcrb7183 lI 0.01 ncrc6712 1I 2.792845e-03 mioa4057 1I 0.01 miod0340 1I 2.957987e-03 mioc0226 1I 0.01 fcrb7453 1I 3.131743e-03 mioc7782 1I 0.01 ncrc1374 12 3.131743e-03 ncr9337 1I 0.01 miod1809 1I 3.314495e-03 seoa1552 1I 0.01 ncr2382 1I 3.314495e-03 seoa8655 1I 0.01 ncrb0513 1I 3.314495e-03 fcrb5269 1I 0.01 ncrc0304 1I 3.314495e-03 ncr3368 1I 0.01 ncrc1203 1I 3.314495e-03 ncrc4259 1I 0.01 mioa8318 1I 3.506638e-03 seoa5766 1I 0.01 seoc1934 1I 3.70858e-03 mioa9323 1I 0.01 fcrb1917 1I 3.920744e-03 ncrc1379 1I 0.01 seoa3108 1I 4.623001e-03 ncrc5500 1I 0.01 seob2303 1I 4.623001e-03 seoa6151 1I 0.01 fcrb4995 1I 5.150669e-03 fcrb9289 1I 0.01 mioa0647 1I 5.150669e-03 fcrb9871 1I 0.01 ncrb8063 1I 5.150669e-03 fcrc7358 1I 0.01 seoa2854 1I 5.150669e-03 mioa8774 1I 0.01 hfcr4176 1I 5.433839e-03 ncr4384 1I 0.01 mioa3913 1I 5.433839e-03 seoa2801 1I 0.01 nc:rc10481I 5.433839e-03 seob0928 1I 0.01 fcrb8208 1I 5.730596e-03 seob4333 1I O.Ol miob0795 1I 5.730596e-03 seob5021 1I 0.01 fcrb6829 1I 6.041484e-03 fcrb0131 1I 0.01 miob2601 1I 6.041484e-03 fcrc2231 1I 0.01 seoa5396 1I 6.041484e-03 hfcr2894 1I 0.01 fcrb4367 1I 6.367062e-03 seoa4317 1I 0.01 fcrb6728 1I 6.367062e-03 fcrb2318 1I 0.01 fcrc2314 1I 6.367062e-03 fcrb4918 1I 0.01 ncrb0045 1I 6.707903e-03 mioa6585 1I 0.01 fcrb3664 1I 7.064602e-03 ncrc0728 1I 0.01 seoc2144 1I 7.064602e-03 fcrb2715 1I 0.01 mioa1353 1I 7.437766e-03 ncr7093 1I -0.01 miob1789 1I 7.437766e-03 ncr8112 1I 0.01 ncr1526 1I 7.437766e-03 seob1008 1I 0.01 ncrc9187 1I 7.437766e-03 fcrb5000 1I 0.01 fcrb4717 1I 7.828021e-03 fcrb5622 1I 0.01 miod5682 1I 7.828021e-03 mioa4782 1I 0.01 seob6139 1I 7.828021e-03 ncrc1775 1I 0.01 seob9193 1I 7.828021e-03 fcrb1855 1I 0.01 seob1052 1I 8.236011e-03 fcrb2113 1I 0.01 seoc7566 1I 8.236011e-03 hfcr2584 1I 0.01 miod0978 1I 8.662397e-03 mioa2478 1I 0.01 seoa5253 1I 8.662397e-03 mioc2204 1I 0.01 fcrb1441 1I 9.107857e-03 miod5894 1I 0.01 fcrb8202 1I 9.107857e-03 seoc0514 1I 0.01 fcrc2472 1I 9.107857e-03 fcr3856 1I 0.01 hfcr1646 1I 9.107857e-03 fcrb7616 1I 0.01 miod3160 1I 9.107857e-03 fcrb7703 1I 0.01 fcr7667 1I 9.573086e-03 fcrc0396 1I 0.01 fcrb4929 1I 9.573086e-03 fcrc6460 1I 0.01 fcrb9069 1I 9.573086e-03 miob3199 1I 0.01 mioa3693 1I 0.01 miob6688 1I 0.01 mioa3963 1I 0.01 seoc2336 1I 0.01 mioa9154 1I 0.01 fcrb1644 1I 0.01 ""fcr~6l~~~.......,....1.-1. ,.."...Ø_.x.1seoa0101 1I 0.02 mioa3440 1I 0.01 fcr2088 1I 0.02 ncrc2161 1I 0.01 miob8373 1I 0.02 seoa2822 1I 0.01 mioc6956 1I 0.02 hfcr4055 1I 0.01 ncrb5537 1I 0.02 mioc1187 1I 0.01 ncrc5724 1I 0.02 ncr7341 1I 0.01 seoa2837 1I 0.02 seoa1611 1I 0.01 seoa'7923 1I 0.02 seoa3748 1I 0.01 seoc2317 1T 0.02 seoa4717 1I 0.01 fcrb6937 1I 0.02 seoa5848 1I 0.01 mioa0494 1I 0.02 seob8807 1I 0.01 mioa9492 1I 0.02 fcrb2796 1I 0.01 seoa8351 1I 0.02 fcrb5087 1I 0.01 seobl219 1I 0.02 fcrc5799 1I 0.01 fcrb9751 1I 0.02 mioa5836 1I 0.01 fcrc3958 1I 0.02 ncr7973 1I 0.01 mioc7781 1I 0.02 ncrc4815 1I 0.01 ncr3587 1I 0.02 ncrc7016 1I 0.01 ncr7666 1I 0.02 seob8291 1I 0.01 ncrc6796 1I 0.02 seoc0951 1I 0.01 fcrb6917 1I 0.03 fcr1984 1I 0.02 mioa9719 1I 0.03 fcrb2996 1I 0.02 miodl925 1I 0.03 fcrb8936 1I 0.02 ncrc1884 1I 0.03 fcrc4688 1I 0.02 ncrc4869 1I 0.03 hfcrl811 1I 0.02 fcrb3017 1I 0.03 mioa6704 1T 0.02 fcrc2280 1I 0.03 ncrb8392 1I 0.02 fcrc6940 1I 0.03 seob7082 1I 0.02 mioa2290 1I 0.03 fcrc6972 1I 0.02 mioa6471 1I 0.03 seca7129 1I 0.02 miob1778 1I 0.03 seob4612 1I 0.02 mioc8635 1I 0.03 fcrb0146 1I 0.02 ncrb0916 1I 0.03 fcrb6759 1I 0.02 fcrb1329 1I 0.03 miob8487 1I 0.02 hfcr1053 1I 0.03 miocl248 1I 0.02 mioa4484 1I 0.03 ncr2260 1I 0.02 mioc2097 1I 0.03 seoa2528 1I 0.02 mioc9775 1I 0.03 fcr0680 1I 0.02 miod7246 1I 0.03 fcr5120 1I 0.02 seoa9656 1I 0.03 fcrb5631 1I 0.02 seoc5842 1I 0.03 mioa2620 1I 0.02 fcr2079 1I 0.03 miob3911 1I 0.02 fcrb9068 1I 0.03 miob4057 1I 0.02 fcrc6470 1I 0.03 mioc2726 1I 0.02 mioc7216 1I 0.03 ncrb4869 1I 0.02 mioc7998 1I 0.03 ncrc0558 1I 0.02 ncr3412 1I 0.03 ncrc8863 1I 0.02 ncrc6584 1I 0.03 seoa5461 1I 0.02 seoa4017 1I 0.03 seob6417 1I 0.02 seoa5156 1I 0.03 seoc4960 1I 0.02 fcrb4994 1I 0.03 fcrb6968 1I 0.02 fcrb7939 1I 0.03 fcrc0892 1I 0.02 fcrb9543 1I 0.03 fcrc5068 1I 0.02 fcrc2619 1I 0.03 hfcr3921 1I 0.02 hfcr2963 1I 0.03 mioc0019 1I 0.02 miob9201 1I 0.03 ncr5168 1I 0.02 mioc5692 1I 0.03 ncrc2701 1I 0.02 miod4938 1I 0.03 ncrc5245 1I 0.02 ncrb1518 1I 0.03 seoa8401 1I 0.02 ncrb8177 1I 0.03 mioc0911 1I 0.02 ncrc7171 1I 0.03 miod3785 1I 0.02 seoc2191 1I 0.03 ncrb0163 1I 0.02 fcrb4229 1I 0.03 ncrb6833 1I 0.02 fcrb6549 1I 0.03 lifcrd"(~2~~1T "'"" ""~"0: ncrb0571 1I 0. 04 mioa03321I 0.03 ncrc0539 1I 0.04 mioa97881T 0.03 seob2807 1I 0.04 miob64421I 0.03 seoc4135 1I 0.04 ncr2836 1I 0.03 fcr0955 1I 0.04 seoc22051I 0.03 fcrb2510 1I 0.04 hfcr51571I 0.03 fcrc0677 1I 0.04 miod70521I 0.03 fcrc5831 1I 0.04 ncrb00461I 0.03 miob1734 1I 0.04 ncrc13491I 0.03 miod1792 1I 0.04 seoc45131I 0.03 ncr0007 1I 0.04 fcr2542 1I 0.04 ncr5055 1I 0.04 fcrb46961I 0.04 ncrb3056 1I 0.04 fcrc49681I 0.04 ncrc2319 1I 0.04 fcrc54711I 0.04 seob3419 1I 0.04 hfcr31341I 0.04 seob5379 1I 0.04 mioc46411I 0.04 seoc0861 1I 0.04 mioc62691I 0.04 mioa9581 1J 5.11163e-04 miod18251I 0.04 ncrb8752 1J 5.47259e-04 miod68451I 0.04 seoa3516 1J 8.71839e-04 ncr7934 1I 0.04 mioa9555 1J 1.057776e-03 ncrb43191I 0.04 seoc0778 1J 1.540198e-03 ncrc44481I 0.04 mioc5751 1J 1.963134e-03 seoa47271I 0.04 fcrb4890 1J 2.083894e-03 seoa83741I 0.04 fcrc6228 1J 2.211242e-03 seob41221I 0.04 seoa5911 1J 2.345486e-03 seob68441I 0.04 seoa0740 1J 2.486944e-03 fcrb19951I 0.04 seoa5554 1J 2.792845e-03 fcrb66351I 0.04 fcrb8187 1J 2.957987e-03 mioa24751I 0.04 ncrc9712 1J 2.957987e-03 miob01541I 0.04 fcrb7510 1J 3.131743e-03 mioc24511I 0.04 fcrb4727 1J 3.314495e-03 ncrc45311I 0.04 mioa2185 1J 3.314495e-03 seoa97111I 0.04 fcrb6808 1J 3.70858e-03 seob02481I 0.04 seob7346 1J 3.70858e-03 seob21491I 0.04 fcrb4378 1J 4.143566e-03 seoc50021I 0.04 miod0592 1J 4.143566e-03 fcrb58101I 0.04 miod3306 1J 4.377496e-03 fcrb68171I 0.04 seoc5039 1J 4.377496e-03 fcrc08391I 0.04 mioa3945 1J 4.88056e-03 fcrc13811I 0.04 ncrc3624 1J 5.150669e-03 miob74431I 0.04 seoc0843 1J 5.433839e-03 miob7985lI 0.04 hfcr0734 1J 5.730596e-03 mioc34841I 0.04 seoa4436 1J 5.730596e-03 seob21851I 0.04 ncrc4135 1J 6.041484e-03 seob74631I 0.04 seob4752 1J 6.041484e-03 fcrb25961I 0.04 mioc2074 1J 7.064602e-03 fcrc50241I 0.04 ncrc4089 1J 7.064602e-03 hfcr52601I 0.04 seoc1175 1J 7.064602e-03 ncr2472 1I 0.04 ncrb3980 1J 7.828021e-03 ncr8538 1I 0.04 ncrc9944 1J 8.236011e-03 seob35591I 0.04 seoa7383 1J 8.662397e-03 seob60001I 0.04 fcrb2452 1J 9.107857e-03 seob79031I 0.04 ncr2160 1J 9.573086e-03 fcr2167 1I 0.04 seoa5234 1J 9.573086e-03 fcrb30561I 0.04 ncrb8689 1J 0.01 fcrb45421I 0.04 seob0992 1J 0.01 fcrb67401I 0.04 miod1532 1J 0.01 mioa10151I 0.04 miod5703 1J 0.01 mioa65521I 0.04 ncrc9704 1J 0.01 mioa99891I 0.04 mioa6093 1J 0.01 mioc35651I 0.04 miob2466 1J 0.01 mioc69461I 0.04 miob8259 1J 0.01 ncr6144 1I 0.04 seob1322 1J 0.01 ""~ f'crb2~~~"~"""~1J ncrb6903 1K 6. 737022e-03 "'"~ "'""
~~~~~~0':
0~1 ~

ncrc4780 1J 0.01 seoa3694 1K 6.737022e-03 seoa5157 1J 0.01 seoc1484 1K 6.737022e-03 miob3308 1J 0.01 hfcr6700 1K 7.311813e-03 seob1061 1J 0.01 fcrb7760 1K 7.928973e-03 fcrc0654 1J 0.01 ncrb7516 1K 7.928973e-03 mioc2039 1J 0.01 miod4775 1K 8.591056e-03 mioc1126 1J 0.01 miob4684 1K 9.300726e-03 fcrb1503 1J 0.01 mioc0852 1K 0.01 fcrb0193 1J 0.02 mioc7910 1K 0.01 mioa8851 lJ 0.02 fcrb4789 1K 0.01 ncr7904 1J 0.02 fcrb9286 1K 0.01 ncrc0640 1J 0.02 ncrc5438 1K 0.01 miob9614 1J 0.02 seob6758 1K 0.01 ncrc3936 1J 0.02 fcrb2849 1K 0.01 fcrb2198 1J 0.02 fcrb4656 1K 0.01 mioa4064 1J 0.02 fcrb8680 1K 0.01 mioa0528 1J 0.02 miod5612 1K 0.01 ncr9140 1J 0.02 seob5342 1K 0.01 seoa9389 1J 0.02 fcrbl441 1K 0.01 fcrb8542 lJ 0.02 ncrc8851 1K 0.01 fcrb9655 1J 0.02 seob0937 1K 0.01 mioa6738 1J 0.02 seob1039 1K 0.01 mioc4978 1J 0.02 seob7392 1K 0.01 seob4270 1J 0.02 seob8999 1K 0.01 fcrb6279 1J 0.03 seoc1305 1K 0.01 fcrc0351 1J 0.03 mioc1085 1K 0.01 fcrc4669 1J 0.03 ncrb3314 1K 0.01 mioc1205 1J 0.03 fcrb2041 1K 0.01 seobl399 1J 0.03 fcrc6138 1K 0.01 seob6379 1J 0.03 ncr2484 1K 0.01 mioa5586 1J 0.03 ncrb2200 1K 0.01 seob2169 1J 0.03 fcrb7861 1K 0.01 ncrc0936 1J 0.03 hfcr2295 1K 0.01 seob1411 1J 0.03 mioc0909 1K 0.01 fcrc4161 1J 0.03 ncrc1049 1K 0.01 mioa6102 1J 0.03 ncrc5327 1K 0.01 fcrb3518 1J 0.03 seoc1664 1K 0.01 ncrc4994 1J 0.03 fcrbl691 1K 0.01 miob7373 1J 0.04 fcrc4456 1K 0.01 mioc0206 1J 0.04 miob3426 1K O.Ol ncr2288 1J 0.04 miod6162 1K 0.01 ncrcl567 1J 0.04 fcrb5702 1K 0.01 seoa4485 1J 0.04 miob9533 1K 0.01 hfcr4485 lJ 0.04 mioc2720 1K 0.01 ncrc3856 1J 0.04 seob3367 1K 0.01 fcrb2350 1J 0.04 fcr2821 1K 0.02 fcrb6734 1J 0.04 fcrb5867 1K 0.02 mioa9821 1J 0.04 fcrb7944 1K 0.02 miob8657 1J 0.04 ncrc1608 1K 0.02 ncr1780 1J 0.04 seob5764 1K 0.02 mioa6832 1J 0.04 fcrb6185 1K 0.02 seob1362 1J 0.04 mioc0164 1K 0.02 mioc0181 1J 0.04 mioc0530 1K 0.02 ncr4946 1J 0.04 seoc5228 1K 0.02 ncrc4287 1J 0.04 mioa5902 1K 0.02 fcrc2745 1J 0.04 seoa0085 1K 0.02 seob2195 1J 0.04 seob0787 1K 0.02 seoc5140 1K 3.392191e-03 hfcr5719 1K 0.02 seob5193 1K 4.048349e-03 mioc3593 1K 0.02 fcrb5259 1K 4.416653e-03 ncrb8332 1K 0.02 mioc6204 1K 4.814199e-03 ncrc4588 1K 0.02 fcr0604 1K 5.242925e-03 seoa0137 1K 0.02 seob7928 1K 6.202152e-03 seob0971 1K 0.02 "'seob5556~ '~" 'D':i72 miob3845 1K 0. 04 ~ "IK

seob3499 1K 0.02 ncrb5197 1K 0.04 fcrb6779 1K 0.02 seoa5833 1K 0.04 fcrc6551 1K 0.02 seoa9060 1K 0.04 mioc0384 1K 0.02 seob1158 1K 0.04 miod0977 1K 0.02 seoc1402 1K 0.04 ncrb2131 1K 0.02 fcr0703 1K 0.04 ncrc2857 1K 0.02 fcrb7098 1K 0.04 seoa4461 1K 0.02 fcrb7831 1K 0.04 fcrb1876 1K 0.02 mioa5812 1K 0.04 fcrb6031 1K 0.02 miob9403 1K 0.04 miob2210 1K 0.02 mioc3669 1K 0.04 ncrc9491 1K 0.02 miod6044 1K 0.04 seoa1269 1K 0.02 ncr5522 1K 0.04 seoa4163 1K 0.02 ncrc0423 1K 0.04 seob1419 1K 0.02 seob2717 1K 0.04 seob1848 1K 0.02 seob4967 1K 0.04 seoc0651 1K 0.02 seob7569 1K 0.04 fcrb2979 1K 0.03 seoc1203 1K 0.04 fcrb4786 1K 0.03 seoc1236 1K 0.04 miob5829 1K 0.03 fcrb3848 1K 0.04 miob9163 1K 0.03 fcrc0305 1K 0.04 mioc6937 1K 0.03 fcrc2439 1K 0.04 fcrb3627 1K 0.03 fcrc6381 1K 0.04 fcrb7244 1K 0.03 miob8989 1K 0.04 mioa4628 1K 0.03 seoa0256 1K 0.04 ncr7631 1K 0.03 seoa2854 1K 0.04 ncrc6127 1K 0.03 seoa4640 1K 0.04 seoa8401 1K 0.03 seob6096 1K 0.04 seob3892 1K 0.03 seob6386 1K 0.04 fcr3282 1K 0.03 seob8300 1K 0.04 fcrb1337 1K 0.03 fcrb7324 1L 3.7556e-04 fcrb2763 1K 0.03 ncrc5780 1L 1.774441e-03 mioa2652 1K 0.03 seoc1508 1L 3.707446e-03 mioc7372 1K 0.03 ncrc9642 1L 4.048349e-03 ncrc0534 1K 0.03 seoa4802 1L 4.416653e-03 ncrc1.4211K 0.03 seob2938 1L 4.416653e-03 ncrc2080 1K 0.03 miob3072 1L 5.242925e-03 fcr7561 1K 0.03 seob2937 1L 5.704865e-03 fcrb7247 1K 0.03 fcr4328 1L 6.202152e-03 fcrc2457 1K 0.03 seoa4327 1L 6.737022e-03 fcrc3993 1K 0.03 seob5658 1L 7.311813e-03 ncr3435 1K 0.03 fcrb3476 1L 7.928973e-03 ncrc8841 1K 0.03 miod5682 1L 7.928973e-03 seoa6078 1K 0.03 ncrc5959 1L 7.928973e-03 seoa8867 1K 0.03 fcr2293 1L 8.591056e-03 seob0386 1K 0.03 mioc0238 1L 9.300726e-03 seoc0809 1K 0.03 ncrc0393 1L 0.01 fcrb1877 1K 0.04 seob8212 1L 0.01 fcrb4994 1K 0.04 mioc3523 1L 0.01 fcrc0160 1K 0.04 mioc7471 1L 0.01 fcrc0720 1K 0.04 ncrb3329 1L 0.01 fcrc2131 1K 0.04 mioa9792 1L 0.01 mioc1963 1K 0.04 fcr2088 1L 0.01 miod0355 1K 0.04 hfcr3486 1L 0.01 seoa2652 1K 0.04 fcrb8202 1L 0.01 seob3370 1K 0.04 fcrb3920 1L 0.01 seob4689 1K 0.04 fcrb5016 1L 0.01 fcr5712 1K 0.04 fcrb8236 1L 0.01 fcrb1397 1K 0.04 fcrc0775 1L 0.01 fcrb1523 1K 0.04 miob6087 1L 0.01 fcrb2051 1K 0.04 ncrb5704 1L 0.01 fcrb7803 1K 0.04 seoa9997 1L 0.01 mioa6738 1K 0.04 seobl399 1L 0.01 fcr5369 1L 0'~ 01 ncrc6407 1L 0. 04 ""

miod53011L 0.01 seob6836 1L 0.04 ncr5613 1L 0.01 fcrb3217 1L 0.04 seob77471L 0.01 hfcr3514 lL 0.04 mioc56331L 0.01 miob2448 1L 0.04 seob26611L 0.01 seoa8501 1L 0.04 fcr4444 1L 0.02 seob9001 1L 0.04 ncrc67561L 0.02 fcrc0180 1L 0.04 seob50321L 0.02 miob8515 1L 0.04 fcrb59261L 0.02 ncrc2377 1L 0.04 miob99011L 0.02 seoc6169 1L 0.04 fcrb23801L 0.02 fcrc6651 1M 1.12314e-04 hfcr02851L 0.02 seoa8299 1M 1.913367e-03 seoa16151L 0.02 ncr3825 1M 2.829671e-03 seoa20421L 0.02 fcrb3219 1M 2.830108e-03 fcrc51601L 0.02 seocl535 1M 3.05452e-03 mioc25921L 0.02 mioc5367 2M 3.825425e-03 ncr5027 1L 0.02 fcrc2807 1M 4.118173e-03 ncrc04211L 0.02 ncrc0821 1M 4.118173e-03 ncrc66971L 0.02 seoa3578 1M 4.118173e-03 seoc42881L 0.02 mioc1279 1M 4.430531e-03 fcrb94501L 0.02 ncrb8224 1M 4.763601e-03 fcrc48411L 0.02 fcr7424 1M 5.11853e-03 mioc34131L 0.02 ncrb0696 1M 5.11853e-03 mioc34921L 0.02 seob7250 1M 6.32669e-03 ncr3869 1L 0.02 miob7435 1M 6.781521e-03 seob57111L 0.02 ferc5873 1M 7.264698e-03 fcrb30161L 0.02 mioa8857 1M 7.264698e-03 fcrb62111L 0,02 mioc5179 1M 8.321953e-03 micc14401L 0.02 ncr4790 1M 8.321953e-03 mioc74441L 0.02 seob5582 1M 8.321953e-03 miod53491L 0.02 miob6485 1M 8.899102e-03 ncr3718 1L 0.02 seoc2549 1M 8.899102e-03 ncr9378 IL 0.02 miod4564 1M 9.510735e-03 ncrc51501L 0.02 seob1906 1M 9.510735e-03 seoc16281L 0.02 ncr4551 1M 0.01 fcrb30741L 0.03 ncrc9528 1M 0.01 hfcr23671L 0.03 mioa0577 1M 0.01 mioa96301L 0.03 mioc7974 1M 0.01 mioc02761L 0.03 ncr8780 1M 0.01 ncr2408 1L 0.03 ncrc2495 1M 0.01 fcrb30241L 0.03 fcr2498 1M 0.01 rniod34171L 0.03 fcr6329 1M 0.01 ncrb24001L 0.03 fcrbi446 1M 0.01 seoa76471L 0.03 fcrb8215 1M 0.01 fcrb57231L 0.03 fcrc2024 1M 0.01 miod06251L 0.03 miob3149 1M 0.01 seob11441L 0.03 miob9901 1M 0.01 seob95521L 0.03 miob9274 1M 0.01 fcrc04561L 0.03 mioc2188 1M 0.01 fcrc49681L 0.03 fcrb1876 1M 0.01 mioa46741L 0.03 fcr7561 IM 0.01 miod34211L 0.03 ncrc5039 1M 0.01 ncrc71271L 0.04 seob8029 1M 0.01 fcrb42871L 0.04 fcrb5305 1M 0.01 fcrb70361L 0.04 fcrc7243 1M 0.01 fcrc54181L 0.04 miob6437 1M 0.01 hfcr26861L 0.04 ncr2145 1M 0.01 miob92841L 0.04 ncrc1949 1M 0.01 miod22321L 0.04 ncrc9530 1M 0.01 seob05641L 0.04 seob9218 1M 0.01 seoc25041L 0.04 ncrb8239 1M 0.01 fcrb81211L 0.04 seob8483 1M 0.01 ncrc11931L 0.04 mioa5054 1M 0.01 . ......
'ncrb~2~~2~~"' "'0:01- mioa5836 1M 0.02 llvI

seob08791M 0.01 miod2211 1M 0.02 fcr38801M 0.01 ncr8111 1M 0.02 fcrc52701M 0.01 ncrb4435 1M 0.02 miob83201M 0.01 ncrc8863 1M 0.02 ncrc19991M 0.01 seoc0861 1M 0.02 seoa97401M 0.01 fcr0793 1M 0.03 seob72621M 0.01 fcrb6406 1M 0.03 mioa53101M 0.01 fcrc4067 1M 0.03 miob564'71M 0.01 hfcr3834 1M 0.03 mioc84121M 0.01 mioc0181 1M 0.03 miod53101M 0.01 mioc1749 1M 0.03 ncrb66751M 0.01 miod2058 1M 0.03 seoa20871M 0.01 ncr7088 1M 0.03 seob18601M 0.01 ncr8538 1M 0.03 mioc2539IM 0.02 ncrc4985 1M 0.03 mioc42421M 0,02 seoa3344 IM 0.03 mioc52401M 0.02 seob3307 1M 0.03 mioc84811M 0.02 seob6414 1M 0.03 ncr09761M 0.02 seoc1483 IM 0.03 ncr23041M 0.02 fcrb8422 1M 0.03 ncrb67201M 0.02 fcrc2683 1M 0.03 ncrb88171M 0.02 ncr2996 1M 0.03 seoa75831M 0.02 ncrb6762 1M 0.03 seob79411M 0.02 miob6721 1M 0.03 fcrb21971M 0.02 ncr3834 1M 0.03 fcrc28001M 0.02 ncr7962 1M 0.03 ncrc4153IM 0.02 ncrcl421 1M 0.03 fcr13881M 0.02 seoa7542 1M 0.03 fcrc45601M 0.02 seob6096 1M 0.03 mioa33351M 0.02 fcr1992 1M 0.03 miod27071M 0.02 fcrb6484 1M 0.03 seoa1977IM 0.02 mioa67'72 1M 0.03 seob40391M 0.02 mioc5636 1M 0.03 seob831I1M 0.02 miod0167 1M 0.03 mioa06841M 0.02 ncrc1168 1M 0.03 miob07851M 0.02 ncrc1311 1M 0.03 miodl7921M 0.02 ncrc7171 2M 0.03 miod45401M 0.02 seoc0619 2M 0.03 ncrc35441M 0.02 seoc4078 1M 0.03 seoa45241M 0.02 miod5338 1M 0.03 seoa57841M 0.02 fcr1347 1M 0.03 seoa98141M 0.02 fcr2220 1M 0.03 seoc22201M 0.02 fcrb1690 1M 0.03 mioa47381M 0.02 mioc6269 1M 0.03 mioc86401M 0.02 miod2853 1M 0.03 miod42691M 0.02 seoa5528 1M 0.03 ncr41131M 0.02 seoa7926 1M 0.03 ncr41221M 0.02 seob4165 1M 0.03 ncr49931M 0.02 fcr0680 1M 0.03 ncr62321M 0.02 fcr3856 1M 0.03 ncrb02651M 0.02 mioa0220 1M 0.03 seob50211M 0.02 miobl789 1M 0.03 seob54901M 0.02 ncrb4437 1M 0.03 fcr39871M 0.02 seoa5'731 1M 0.03 fcrb21001M 0.02 seob3226 1M 0.03 fcrc40331M 0,02 seoc5125 1M 0.03 fcrc41381M 0.02 fcrc0076 1M 0.04 mioa02471M 0.02 fcrc2090 1M 0.04 mioa65431M 0.02 fcrc4157 1M 0.04 mioc02791M 0.02 miob0942 1M 0.04 ncrb68461M 0.02 miod0773 1M 0.04 seoa98891M 0,02 miod0832 1M 0.04 fcrb84851M 0.02 ncr5975 1M 0.04 ncrb5535""~"-'-~ "- ~ 0 . 04 fcrc2014 1N 2 . 620469e-03 ~~I~

ncrb73501M 0.04 ncrb1420 1N 2.620469e-03 ncrc70921M 0.04 mioa4845 1N 2.830108e-03 seob61981M 0.04 fcrc2096 1N 3.551234e-03 mioa36021M 0.04 seob4363 1N 3.551234e-03 fcr49001M 0.04 fcrb6005 1N 3.825425e-03 fcrb30161M 0.04 seoc2723 1N 3.825425e-03 fcrb37601M 0.04 ncrc0807 1N 5.496516e-03 fcrb42311M 0.04 ncrc8909 1N 5.898804e-03 hfcr11151M 0.04 fcrb6890 1N 6.32669e-03 mioa54611M 0.04 fcrb4271 IN 6.781521e-03 mioa60911M 0.04 ncrb4319 1N 6.781521e-03 mioc32061M 0.04 seoa3852 1N 6.781521e-03 miod47351M 0.04 fcrb5550 1N 7.264698e-03 ncrc02791M 0.04 mioc2662 1N 7.264698e-03 ncrc20871M 0.04 ncrc6623 1N 7.777673e-03 ncrc30491M 0.04 miod4464 1N $.321953e-03 seob09491M 0.04 ncr0451 1N 8.321953e-03 seoc05231M 0.04 seob1285 1N 8.321953e-03 seoa02311M 0.04 miod6324 1N 8.899102e-03 fcr41061M 0.04 mioc0302 1N 9.510735e-03 fcrb64421M 0.04 mioc5203 1N 9.510735e-03 fcrc03321M 0.04 mioc6320 1N 9.510735e-03 fcrc26701M 0.04 ncr3316 1N 9.510735e-03 mioa17751M 0.04 ncrc0544 1N 9.510735e-03 mioa20731M 0.04 ncrc2110 1N 9.510735e-03 mioa47921M 0.04 fcrb4321 1N 0.01 mioa54521M 0.04 seoa9817 1N 0.01 mioa87961M 0.04 ncrc1049 1N 0.01 mioa95051M 0.04 seoa7897 1N 0.01 miob17741M 0.04 seob8386 1N 0.01 miob85321M 0.04 fcrb3169 1N 0.01 mioc09111M 0.04 mioa0820 1N 0.01 mioc3430lM 0.04 mioa3620 1N 0.01 ncrc51491M 0.04 miod7429 1N 0.01 seoa7.6611M 0.04 ncrc6846 1N 0.01 seob03701M 0.04 seoa3102 1N 0.02 seob79031M 0.04 seoa4056 1N 0.01 fcrb13971M 0.04 seoa8993 1N 0.01 fcrb28511M 0.04 fcr5350 1N 0.01 fcrb31271M 0.04 fcr5836 1N 0.01 fcrc56141M 0.04 fcrb5219 1N 0.01 mioa36931M 0.04 miob9905 1N 0.01 miod51221M 0.04 hfcr0521 1N 0.01 ncr4416lM 0.04 mioc0902 1N 0.01 ncr61971M 0.04 mioc1590 1N 0.01 ncrb13731M 0.04 ncr3189 1N 0.01 ncrc35511M 0.04 ncrc6811 1N 0.01 ncrc65871M 0.04 seob5478 IN 0.01 ncrc71051M 0.04 seoc5209 1N 0.01 seoa21351M 0.04 fcrb4275 1N 0.01 seoa2l621M 0.04 mioc3663 1N 0.01 seoa873$1M 0.04 ncrb6394 1N 0.01 seoa94451M 0.04 hfcr3375 1N 0.01 ncrc05391N 2.76601e-04 mioa9821 1N 0.01 seoc50391N 4.04054e-04 mioc3746 1N 0.01 fcrb15801N 9.05702e-04 seoa5691 1N 0.01 mioc30421N 1.500619e-03 fcr2218 1N 0.01 seoa38471N 1.500619e-03 fcrb5181 1N 0.01 seob04421N 1.628342e-03 ncrc0646 1N 0.01 ncrb86651N 1.913367e-03 ncrc6697 1N 0.01 seob42701N 2.071956e-03 seoa6152 1N 0.01 fcrb24521N 2.42476e-03 seoc3487 IN O.OI

fcrb94011N 2.42476e-03 ncrc1003 1N 0.01 ~7 """ ~1N miod2888 10 4.231915e-03 '~"- -~~~~"~0~.02 ~
~f 99 ncrc4815 10 4.728956e-03 cr0 fcrc2346 1N 0.02 fcrc2724 10 5.141808e-03 mioa0252 1N 0.02 ncr6343 l0 5.398569e-03 ncrc9232 1N 0.02 seoc5039 10 5.492474e-03 seoa4040 1N 0.02 mioa0577 10 5.937728e-03 hfcr2890 1N 0.02 fcrb5123 10 6.493198e-03 ncr1523 1N 0.02 ncrc4219 10 6.586678e-03 fcr1562 1N 0.02 fcrb9639 10 6.645906e-03 ncrb8721 1N 0.02 ncrc2859 10 6.734392e-03 mioa2185 1N 0.02 fcrb7099 10 6.749149e-03 ncr6141 1N 0.02 mioa3725 10 7.103544e-03 ncrc3045 1N 0.02 seoa3827 10 7.157172e-03 ncrc5207 1N 0.02 seoc0861 10 7.27845e-03 seoa1439 1N 0.02 ncr9587 10 7.280627e-03 mioc5270 1N 0.02 mioc6360 10 7.319551e-03 miod0708 1N 0.02 seola7765 10 7.57129e-03 ncrc0863 1N 0.02 mioa3392 10 7.67629e-03 ncr4648 1N 0.03 mioc3605 10 8.235796e-03 seob6853 1N 0.03 mioc3042 10 8.28488e-03 fcrb7780 1N 0.03 fcrb5181 10 8.370927e-03 mioc2997 1N 0.03 mioa4845 10 8.41047e-03 mioc329~ 1N 0.03 seoa9409 10 8.522999e-03 ncrc0262 1N 0.03 seoa8501 10 9.074357e-03 seoc0780 1N 0.03 seob4363 10 9.215346e-03 mioc1783 1N 0.03 ncr9175 10 9.251372e-03 ncr3713 1N 0.03 ncrb2125 10 9.300568e-03 ncrb3314 1N 0.03 seob1906 10 9.52088e-03 seob2953 1N 0.03 seob5003 10 9.992094e-03 fcrb1399 1N 0.03 h fcr6501 1N 0.03 fcrc4157 10 0.01 mioc4994 1N 0.03rt seoa8370 10 0.01 miod4629 1N 0.03 fcrb9914 10 0.01 ncr1526 1N 0.03 miod4539 10 0.01 ncrc2394 1N 0.03 fcrb4378 10 0.01 seoa6118 1N 0.03 miod6068 10 0.01 fcr3861 1N 0.04 mioc5367 10 0.01 mioa3945 1N 0.04 fcrb9401 10 0.01 mioc6374 1N 0.09 seob7764 10 O.Ol seoc2518 1N 0.09 ncrb8665 10 0.01 fcrb3592 1N 0.04 seoa6151 10 0.01 hfcr6370 1N 0.04 ncrc5672 10 0.01 mioa3471 1N 0.04 ncrc0852 10 0.01 mioa6731 1N 0.04 fcrb2306 10 0.01 miod3327 1N 0.04 seoa3102 10 0.01 ncrb6581 1N 0.04 fcrb8915 10 0.01 ncrc9637 1N 0.04 fcrc2050 10 0.01 seob6703 1N 0.04 mioa8857 10 0.01 fcrc6470 1N 0.04 fcrb2449 10 0.01 mioa9604 1N 0.04 ncrc2103 10 0.01 miod4895 1N 0.04 mioa5054 10 0.01 seoa2744 1N 0.04 mioa4532 10 0.01 seob7765 1N 0.04 seob1285 10 0.01 fcr2195 1N 0.04 mioc0424 10 0.01 ncrc2080 1N 0.04 ncrc5571 10 0.01 seoa7669 1N 0.04 ncr1235 10 0.01 seob0089 1N 0.04 ncrb1420 10 0.01 seob6584 1N 0.04 ncr3827 10 0.01 fcrc6651 10 1.70693e-03 seoa3847 10 0.01 seoa0536 10 2.030178e-03 mioc3296 10 0.01 ncr2507 10 3.138435e-03 ncrc9052 10 0.01 fcrc2619 10 3.288685e-03 seoa0792 10 0.01 seoa8993 10 3.289707e-03 fcrb6091 10 0.01 mioa0474 10 4.026625e-03 fcrb3181 10 0.01 mioa3888 10 4.070197e-03 ncr7382 10 0.01 seob5080~~~w'~'~ "m~~~: ~~ hfcr3531 10 0.
~0~~ 02 seob403910 0.01 hfcr1115 10 0.02 fcrc587310 0.01 fcrc5577 10 0.02 fcrb600010 0.01 fcrb5550 10 0.02 seoa829910 0.01 ncrc0445 10 0.02 fcr2821 10 0.01 mioa5540 10 0.02 miob093110 0.01 miod5058 10 0.02 fcrb280010 0.01 miob6459 10 0.02 seoa612910 0.01 seoa0085 10 0.02 fcrc399810 0.01 mioa5310 10 0.02 miob256910 0.01 mioa3963 10 0.02 fcrb906910 0.01 ncrc1049 10 0.02 ncrc662310 0.01 mioa0307 10 0.02 fcrb947610 0.01 hfcr3503 10 0.02 seob427010 0.01 seoc1661 10 0.02 fcrb795110 0.01 ncrc2825 10 0.02 ncr2363 10 0.01 mioa0220 10 0.02 seoc153510 0.01 fcrc0076 l0 0.02 mioa1513l0 0.01 miod3306 l0 0.02 seob015410 0.01 fcrb6406 10 0.02 seob670410 0.01 fcrc6305 10 0.02 ncrc209810 0.01 ncrc9020 10 0.02 miob560810 0.01 mioa3471 10 0.02 ncrc027910 0.01 fcrb4470 10 0.02 seoc522210 0.01 ncrc1999 l0 0.02 seoc490910 0.01 miob2287 10 0.02 ncr3052 10 0.01 mioa6583 10 0.02 seob178710 0.01 fcrc2807 10 0.02 ncr7945 10 0.01 fcrc6969 10 0.02 mioc103010 0.01 miod3327 10 0.02 fcrb344510 0.01 fcrc2014 10 0.02 ncrc211010 O.U1 fcrc6028 10 0.02 ncrc6264i0 0.02 ncr3825 10 0.02 ncrc355110 0.02 fcrb3017 10 0.02 fcrb251610 0.02 seob4001 10 0.03 fcrb600510 0.02 fcrb6202 10 0.03 fcr0419 10 0.02 seob0089 10 0.03 ncrc080710 0.02 mioc5240 10 0.03 seob627210 0.02 seoc0009 10 0.03 hfcr289010 0.02 miob6485 10 0.03 fcrc267010 0.02 fcr5350 10 0.03 ncrb055010 0.02 seoa9165 10 0.03 ncr4648 10 0.02 seob8403 10 0.03 seob986910 0.02 ncrc3049 10 0.03 hfcr286310 0.02 mioc1279 10 0.03 fcrc027110 0.02 fcrb4533 10 0.03 seoc460910 0.02 fcrb3056 10 0.03 miod473210 0.02 ncr0451 10 0.03 ncr9789 10 0.02 seob1081 10 0.03 ncr8588 10 0.02 mioa0820 10 0.03 fcrb206010 0.02 fcrb3205 10 0.03 miod285310 0.02 seoc1175 10 0.03 seob622910 0.02 fcr7424 10 0.03 ncr3568 10 0.02 ncrb4319 10 0.03 fcrb245210 0.02 fcrb5929 10 0.03 seob596210 0.02 ncr4790 10 0.03 seob976410 0.02 fcr0997 10 0.03 mioa367210 0.02 ncr0046 10 0.03 ncr6072 10 0.02 mioc6341 10 0.03 mioa706910 0.02 ncr6170 10 0.03 seob373110 0.02 fcrc5311 10 0.03 seoc344310 0.02 hfcr2820 10 0.03 fcrc424410 0.02 fcrc6997 10 0.03 hfcr254710 0.02 mioc6320 10 0.03 'iri'i 2'0 ""'" b": f?'~mi o a 6 0 3 10 0 . 0 4 o'c'7 "~ " 4 8 T$'"' '"'"

fcrb9099 10 0.03 seoa0396 10 0.04 miob7092 10 0.03 seob0523 10 0.04 ncr2472 10 0.03 fcr2293 10 0.04 hfcr0734 10 0.03 seoc5209 10 0.04 fcrb8732 10 0.03 ncrc4994 10 0.04 ncrb5537 10 0.03 ncrb4435 10 0.04 ncr8096 10 0.03 miob3320 10 0.04 seoa3852 10 0.03 ncrc3544 10 0.04 mioa4738 10 0.03 ncrc6587 10 0.04 seoc0018 10 0.03 ncrc5663 10 0.04 mioa6739 10 0.03 fcrb5896 10 0.04 ncrc9877 10 0.03 fcr6427 10 0.04 mioc3573 10 0.03 seob2148 10 0.04 seocl694 10 0.03 mioc2541 10 0.04 ncrc5025 10 0.03 seob6131 10 0.04 mioc2381 10 0.03 fcrb6574 10 0.04 seoa1737 10 0.03 ncrb8714 10 0.04 fcr7561 10 0.03 ncrb7350 10 0.04 fcrb4275 10 0.03 fcrc6313 10 0.04 fcrb6436 10 0.03 fcrb5070 10 0.04 miob0974 10 0.03 ncr4551 10 0.04 seob3734 10 0.03 fcrb9520 10 0.04 fcr1060 10 0.03 fcrc2376 1P 2.27e-04 ncrb6394 10 0.03 ncrc2381 1P 7.13e-04 seob9872 10 0.03 ncrb3957 1P 9.67e-04 seoa2087 10 0.03 ncrc7171 1P 1.034296e-03 seob6279 10 0.04 seoa0925 1P 1.319416e-03 fcrb2305 10 0.04 fcrc1906 1P 1.365415e-03 hfcr4176 10 0.04 miob9441 1P 1.426128e-03 fcrc0072 10 0.04 fcrb7753 1P 1.481468e-03 ncr4122 10 0.04 mioc0621 1P 1.503579e-03 mior_090210 0.04 miod5184 1P 1.661105e-03 fcrc0180 10 0.04 miod3160 1P 1.763432e-03 ncrb1899 10 0.04 seob8854 1P 1.841199e-03 miod4496 10 0.04 fcrb9963 1P 1.859959e-03 ncrb1179 10 0.04 mioa3940 1P 1.859959e-03 seoa9817 10 0.04 ncrc4219 1P 2.079782e-03 mioc0019 10 0.04 miob9403 1P 2.104585e-03 fcrb3169 10 0.04 ncr4030 1P 2.18945e-03 seob8300 10 0.04 fcrc2775 1P 2.201026e-03 seob0294 10 0.04 miod4998 1P 2.322457e-03 miob6437 10 0.04 ncrc7065 1P 2.322457e-03 fcrc2254 10 0.04 rnioc3127 1P 2.476955e-03 seoa3578 10 0.04 mioc4103 1P 2.476955e-03 fcrb8422 10 0.04 fcrc4948 1P 2.588614e-03 seoa0029 10 0.04 fcrc5134 1P 2.589996e-03 fcrb6460 10 0.04 mioa2783 1P 2.589996e-03 ncrc0508 10 0.04 ncr7813 1P 2.589996e-03 fcrb9633 10 0.04 fcrc1689 1P 2.804609e-03 fcrb5527 10 0.04 ncr8314 1P 3.121843e-03 seob0386 10 0.04 miob2492 1P 3.20844e-03 fcrb4599 10 0.04 miob8347 1P 3.20844e-03 fcrc4669 10 0.04 seoa5234 1P 3.20844e-03 fcrb2350 10 0.04 fcrb2765 1P 3.308966e-03 mioc0302 10 0.04 mioc3413 1P 3.490738e-03 seob9232 10 0.04 fcrb7380 1P 3.564101e-03 fcrb8505 10 0.04 miob5873 1P 3.585792e-03 ncrb2131 10 0.04 fcrc2280 1P 3.590799e-03 ncrc4259 10 0.04 seoc7885 1P 3.612999e-03 fcrb3896 10 0.04 seob6872 1P 3.677261e-03 fcrc5379 10 0.04 miob6688 1P 3.723139e-03 ncrb3468 10 0.04 fcrb3518 1P 3.874043e-03 fcrc6990 10 0.04 miocl438 1P 3.954153e-03 inioc215'2""' """'3":'9'5415'3'e-03fcrb2993 1P 8.683395e-03 """1'P' ncr3700 1P 3.954153e-03 fcrb8936 1P 8.683395e-03 ncrb75611P 4.138654e-03 miob9817 1P 8.683395e-03 ncrc09721P 4.19215e-03 mioc2360 1P 8.683395e-03 miob75541P 4.228055e-03 miod2707 1P 8.683395e-03 fcrb12021P 4.381373e-03 ncr3834 1P 8.683395e-03 fcrb27961P 4.381373e-03 ncrc0393 1P 8.683395e-03 fcrc06371P 4.381373e-03 ncrc5508 1P 8.683395e-03 ncr0615 1P 4.381373e-03 seobl839 1P 8.683395e-03 seoa51561P 4.381373e-03 seoc0780 1P 8.684081e-03 miob71881P 4.531729e-03 fcrcl165 1P 8.876286e-03 ncrc08291P 4.789246e-03 miob7209 1P 8.90642e-03 seoa98741P 4.817268e-03 seob1574 1P 9.211766e-03 mioa68111P 4.848713e-03 fcrb8432 1P 9.228453e-03 miodl4541P 4.848713e-03 mioa9033 1P 9.277008e-03 ncrb39031P 4.848713e-03 ncrc9739 1P 9.281176e-03 seoc20291P 4.848713e-03 miod6947 1P 9.37261e-03 fcr3664 1P 4.853433e-03 seoc3277 1P 9.391913e-03 fcrc61271P 4.994669e-03 mioa0826 1P 9.529688e-03 mioc14161P 5.177558e-03 ncr0791 1P 9.529688e-03 mioc27281P 5.300295e-03 ncrc3049 1P 9.529688e-03 fcr3559 1P 5.359298e-03 ncrc3171 1P 9.529688e-03 miob40371P 5.359298e-03 seob3699 1P 9.529688e-03 ncr1055 1P 5.359298e-03 fcrb6776 1P 9.549328e-03 seoa26311P 5.359298e-03 fcrb8730 1P 9.812826e-03 ncrb69491P 5.470501e-03 seoc2785 1P 9.857578e-03 seoa89601P 5.592575e-03 seoc1345 1P 0.01 seob01281P 5.718877e-03 seoa7403 1P 0.01 ncrc54051P 5.752711e-03 seobl667 1P 0.01 fcrc46581P 5.916437e-03 fcrc5850 1P 0.01 mioc75091P 5.916437e-03 miod5651 1P 0.01 miob46131P 5.926508e-03 seoa0044 1P 0.01 miob281.41P 6.166182e-03 seoa7115 1P 0.01 mioa34671P 6.426036e-03 seoa8351 1P 0.01 ncrc24631P 6.523627e-03 seob4117 1P 0.01 seoa53921P 6.523627e-03 seob5843 1P 0.01 fcrc51381P 6.64384e-03 fcrc4045 1P 0.01 miod74861P 6.651906e-03 fcrb7510 1P 0.01 fcrbl4921P 7.184558e-03 fcrb9843 1P 0.01 fcrb66671P 7.184558e-03 seoc2031 1P 0.01 fcrb92221P 7.184558e-03 mioc0899 lP 0.01 fcrc68881P 7.184558e-03 fcr7419 1P 0.01 mioal8851P 7.184558e-03 fcrb7700 1P 0.01 mioa55861P 7.184558e-03 fcrc4971 1P 0.01 miobl8331P 7.184558e-03 fcrc5007 1P 0.01 miod51261P 7.184558e-03 mioa8679 1P 0.01 miod53491P 7.184558e-03 miob3042 1P 0.01 ncr0420 1P 7.184558e-03 mioc6956 1P 0.01 ncr3960 1P 7.184558e-03 miod4467 1P 0.01 ncr9919 1P 7.184558e-03 ncr3404 1P 0.01 ncrc41131P 7.184558e-03 ncr3412 1P 0.01 seob68821P 7.184558e-03 ncr3527 1P 0.01 mioc12451P 7.224439e-03 ncrb2091 1P 0.01 seob47491P 7.682062e-03 ncrc4079 1P 0.01 miob34111P 7.697079e-03 fcrb2344 1P 0.01 fcrb96111P 7.903117e-03 seoa0464 1P 0.01 mioc51131P 7.903117e-03 seob1318 1P 0.01 ncrc07441P 7.903117e-03 ncrc0445 1P 0.01 ncrc45311P 7.903117e-03 fcrb1552 1P 0.01 fcrb61601P 7.966664e-03 fcrb3001 1P 0.01 seoa36341P 8.179896e-03 fcrc0672 1P 0.01 hfcr40281P 8.260407e-03 fcrc5164 1P 0.01 fcrb57531P 8.563533e-03 fcrc5402 1P 0.01 seob68351P 8.57233e-03 fcrc6282 1P 0.01 ..fcrc7 ,...,.. ~ mioc1978 1P 0 0.x.5.... : ~.i , .
"",. 01 '1'p mioal5841P 0.01 mioc7668 1P 0.01 ncr2866 1P 0.01 ncrc5372 1P 0.01 ncr3357 1P 0.01 ncrc1153 lP 0.01 ncrc08581P 0.01 miob8096 lP 0.01 ncrc3842lP O.Ol fcrb4077 1P 0.01 ncrc58441P 0.01 fcrb6190 1P 0,01 fcrc44081P 0.01 hfcr2963 1P 0.01 mioc00191P 0.01 hfcr6640 lP 0.01 ncr0212 1P 0.01 miob2466 1P 0.01 seob27241P 0.01 mioc2019 1P O.Ol ncr3434 lP 0.01 mioc5695 1P 0.01 ncrc65971P 0.01 ncr3944 1P 0.01 fcr6630 1P 0.01 seoa5253 1P 0.01 miob23861P 0.01 seoa7126 1P O.Ol miob76101P 0.01 seob3303 1P 0.01 mioc03101P 0.01 mioa9891 lP 0.01 mioc07601P 0.01 miob2478 1P 0.01 mioc1l071P 0.01 mioa2691 1P 0.01 mioc52101P 0.01 mioc3716 1P O.Ol ncr4700 1P 0.01 miod1316 lP O.Ol ncrb15181P 0.01 seoc1476 1P 0.01 ncrc97291P 0.01 seoc1484 1P 0.01 seob93531P 0.01 ncr2575 1P 0.01 seoc5134lP 0.01 fcrb4391 1P 0.01 miob92481P 0.01 fcrb7525 lP 0.01 ncrc47571P 0.01 mioa0294 1P 0.01 ncrc57581P 0.01 mioa3080 1P 0.01 hfcr66111P 0.01 miob2686 1P 0.01 fcrb28511P 0.01 miob6290 1P 0.01 mioa34691P 0.01 miob7638 1P 0.01 mioc63911P 0.01 mioc0486 1P 0.01 ncr3163 1P 0.01 mioc4925 1P 0.01 ncr4545 1P 0.01 mioc6260 1P 0.01 seoa55771P 0.01 ncr4118 1P 0.01 seoa58431P 0.01 seob0219 1P 0.01 seob48961P 0.01 seob9282 1P 0.01 mioc48881P 0.01 seoc5612 1P 0.01 seob92411P 0.01 ncrb7465 1P 0.01 mioc39061P 0.01 fcrb9909 1P 0.01 fcrb94491P 0.01 seoc0945 1P 0.02 mioc41611P 0.01 seoc4078 1P 0.02 fcrc69721P 0.01 ncr0612 1P 0.02 seoc4l871P 0.01 miob3595 1P 0.02 fcrc63351P 0.01 miob5810 1P 0.02 seoc20001P 0.01 seoc0551 1P 0.02 mioa96301P 0.01 mioc1995 1P 0.02 hfcr30191P 0.01 fcrb3870 1P 0.02 mioa24931P 0.01 fcrb7951 1P 0.02 ncr3432 1P 0.01 fcrc4876 1P 0.02 ncr9933 1P 0.01 hfcr1709 1P 0.02 ncrb11631P 0.01 miob4144 1P 0.02 ncrb60871P 0.01 miob5675 1P 0.02 ncrc1595lP 0.01 mioc2133 1P 0.02 ncrc34361P 0.01 miod0592 1P 0.02 ncrc38801P 0.01 ncr3339 1P 0.02 seoa27751P 0.01 ncrc3520 1P 0.02 seob79291P 0.01 ncrc4496 1P 0.02 fcrb95431P 0.01 ncrc5571 lP 0.02 fcrc22221P 0.01 seoa0135 1P 0.02 miob63551P 0.01 seob0817 1P 0.02 seob60081P 0.01 seob1808 1P 0.02 ncrc34431P 0.01 seob1908 lP 0.02 fcrb66391P 0.01 seoc2249 1P 0.02 ricrc332~4~' 0~:~0~~ ' miod0773 1P 0.

miob83911P 0.02 fcrb5841 1P 0.02 seob82611P 0.02 fcrb9629 1P 0.02 mioc21231P 0.02 hfcr3067 1P 0.02 ncr3465 1P 0.02 mioa3379 lP 0.02 ncrc88651P 0.02 miob7922 1P 0.02 mioc74711P 0.02 miob9065 1P 0.02 ncrc08491P 0.02 mioc3574 1P 0.02 seob86071P 0.02 ncr0496 1P 0.02 mioc27991P 0.02 ncr7537 1P 0.02 seoa75421P 0.02 ncrcl349 1P 0.02 fcr5415 1P 0.02 ncrc4302 1P 0.02 fcrb47601P 0.02 ncrc6242 1P 0.02 fcrb56031P 0.02 seoa6238 1P 0.02 fcrb68081P 0.02 seob2169 1P 0.02 fcrc03761P 0.02 seob3189 1P 0.02 fcrc08391P 0.02 seob5064 1P 0.02 fcrc68681P 0.02 mioc7433 1P 0.02 hfcr00111P 0.02 ncr9469 1P 0.02 miob97141P 0.02 ncrc8851 lP 0.02 miod74401P 0.02 seoa7286 1P 0.02 miod74611P 0.02 miob9325 1P 0.02 ncr2994 1P 0.02 fcrb3946 1P 0.02 ncr4140 1P 0.02 seoa9409 1P 0.02 ncr6212 1P 0.02 mioc3962 1P 0.02 ncr8156 1P 0.02 fcrb1466 1P 0.02 ncrc42311P 0.02 fcrb1633 1P 0.02 ncrc57441P 0.02 fcrb3330 1P 0.02 ncrc68991P 0.02 fcrb4656 1P 0.02 seoa20511P 0.02 fcrb8080 1P 0.02 seoa32871P 0.02 fcrc2014 1P 0.02 seoc21311P 0.02 fcrc5233 1P 0.02 mioc71701P 0.02 mioa3987 1P 0.02 fcrb90691P 0.02 miob2341 1P 0.02 ncrc03411P 0.02 miob2944 1P 0.02 hfcr26931P 0.02 mioc7620 1P 0.02 fcrb61121P 0.02 miod2641 1P 0.02 fcrb57231P 0.02 miod7227 1P 0.02 fcrb64321P 0.02 ncrb5244 1P 0.02 mioa89451P 0.02 ncrc1811 1P 0.02 mioc14401P 0.02 ncrc3856 1P 0.02 miod42201P 0.02 ncrc6423 1P 0.02 miod47841P 0.02 seoa4675 1P 0.02 miod48951P 0.02 seoa5387 1P 0.02 ncr1101 1P 0.02 seob1844 1P 0.02 ncr5557 1P 0.02 seob9756 1P 0.02 ncr5668 1P 0.02 seoc1661 1P 0.02 ncr9378 1P 0.02 seoc1856 1P 0.02 ncrb44411P 0.02 mioc5740 1P 0.02 ncrb57371P 0.02 ncr2175 1P 0.02 ncrc31211P 0.02 seoa8229 1P 0.02 seoa36391P 0.02 ncrc1310 1P 0.02 seob88391P 0.02 fcrc0430 1P 0.02 mioc19631P 0.02 seoa6930 1P 0.02 miob73731P 0.02 ncrc7131 1P 0.02 fcrc61961P 0.02 ncr0159 1P 0.02 ncrc94281P 0.02 ncrc9517 1P 0.02 seob74441P 0.02 seoc4900 1P 0.03 mioc11261P 0.02 mioc2577 1P 0.03 seob29381P 0.02 ncr3233 1P 0.03 mioc76861P 0.02 seoc4941 1P 0.03 miob45741P 0.02 miob5098 1P 0.03 ncrc06631P 0.02 seoa8640 1P 0.03 miob24321P 0.02 mioc0317 1P 0.03 fcrb154'7-~-~--"'~~~~0~:'0'3fcr0637 1P 0.
--~ 1h' 03 fcrb1990lP 0.03 fcr4803 lP 0.03 fcrb27131P 0.03 fcrb1380 1P 0.03 fcrb37261P 0.03 fcrb1689 1P 0.03 fcrb43111P 0.03 miob6562 1P 0.03 fcrb69491P 0.03 mioc2872 1P 0.03 fcrb85361P 0.03 mioc4022 1P 0.03 hfcr28951P 0.03 ncr8177 lP 0.03 mioa41771P 0.03 ncr9324 1P 0.03 miob18731P 0.03 ncrc2079 1P 0.03 mioc45571P 0.03 ncrc2730 1P 0.03 ncrc07041P 0.03 ncrc3598 1P 0.03 ncrc95191P 0.03 seoa3422 1P 0.03 seob13851P 0.03 seob0815 lP 0.03 seob40571P 0.03 seob1766 1P 0.03 seob50541P 0.03 seob4095 lP 0.03 miod52181P 0.03 seob4333 1P 0.03 miod44491P 0.03 seoc2348 1P 0.03 fcrb76931P 0.03 seobl960 1P 0.03 fcrc0569lP 0.03 fcr2700 1P 0.03 miob91851P 0.03 mioa9007 1P 0.03 fcrb2754lP 0.03 seoa7178 1P 0.03 mioa95231P 0.03 seob7039 1P 0.03 ncrc71131P 0.03 fcrc6916 1P 0.03 seobl8891P 0.03 ncrb2085 1P 0.03 ncr2812 1P 0.03 seob8065 lP 0.03 seob30641P 0.03 . miob8707 lP 0.03 fcrc64521P 0.03 seob5032 1P 0.03 fcr5474 lP 0.03 ncrc0863 1P 0.03 fcrb21891P 0.03 miob8609 lP 0.03 fcrb24601P 0.03 hfcr2770 1P 0.03 fcrc73731P 0.03 mioa6585 1P 0:03 mioa05281P 0.03 mioa0582 1P 0.03 mioa08621P 0.03 ncrc9039 1P 0.03 mioa34401P 0.03 fcrb3237 1P 0.03 miob34271P 0.03 fcrb6666 1P 0.03 mioc24431P 0.03 hfcr1762 lP 0.03 mioc25961P 0.03 hfcr2287 1P 0.03 ncrb85851P 0.03 hfcr2629 1P 0.03 ncrc03831P 0.03 hfcr6613 1P 0.03 ncrc67781P 0.03 mioa2998 1P 0.03 ncrc70491P 0.03 mioa5326 1P 0.03 seoa09131P 0.03 mioa8796 1P 0.03 seoa73661P 0.03 miob2375 1P 0.03 seoa82681P 0.03 miob5770 1P 0.03 seoa94451P 0.03 ncr5168 1P 0.03 seob21391P 0.03 ncr7178 1P 0.03 seob38871P 0.03 ncrb3498 1P 0.03 seob40391P 0.03 ncrc0427 1P 0.03 seob54781P 0.03 ncrc0544 1P 0.03 seob55561P 0.03 ncrc2888 1P 0.03 ncr7284 1P 0.03 seoa2381 1P 0.03 mioa65801P 0.03 seoa2854 1P 0.03 seob47661P 0.03 seoa3207 1P 0.03 seob00651P 0.03 seob0063 1P 0.03 fcr3043 1P 0.03 seob6139 1P 0.03 seob61771P 0.03 seob6585 1P 0.03 seob83881P 0.03 seoc0651 1P 0.03 seoc06571P 0.03 seoc1804 1P 0.03 fcrb82221P 0.03 mioc4117 1P 0.03 ncr9125 1P 0.03 hfcr5514 1P 0.03 seoa44221P 0.03 ncrc0375 1P 0.03 fcrc64131P 0.03 mioa4135 1P 0.04 seoa40121P 0.03 fcrc2126 1P 0.04 """~fcrb5"38~4~ """""'~0'.~04miob3320 1P 0.
~~~" 04 "'~1'P

mioc7561 1P 0.04 mioc1560 1P 0.04 ncrb7340 1P 0.04 ncr3141 1P 0.04 seocl631 1P 0.04 ncr3177 1P 0.04 ncr4215 1P 0.04 ncr5568 1P 0.04 seoc0596 1P 0.04 ncr7924 1P 0.04 fcrb5902 1P 0.04 ncrb8721 1P 0.04 fcr4699 1P 0.04 ncrc0564 1P 0.04 fcrbl916 1P 0.04 ncrc3100 1P 0.04 fcrb4799 1P 0.04 ncrc3464 1P 0.04 fcrb8236 1P 0.04 ncrc6587 1P 0.04 fcrc0487 1P 0.04 ncrc9159 1P 0.04 fcrc1565 1P 0.04 ncrc9700 1P 0.04 hfcr2955 1P 0.04 seoa1910 1P 0.04 mioa1097 1P 0.04 seoa4167 1P 0.04 mioa2038 1P 0.04 seob1891 1P 0.04 mioa8484 1P 0.04 seob3684 1P 0.04 miob5646 1P 0.04 seoc1872 1P 0.04 mioc1085 1P 0.04 fcrc2231 1P 0.04 mioc2735 1P 0.04 fcrb4533 1P 0.04 ncr3963 1P 0.04 fcrc5506 1P 0.04 ncr7539 1P 0.04 mioa5729 1P 0.04 ncrc1259 1P 0.04 mioc0662 1P 0.04 ncrc3596 1P 0.04 mioc6385 1P 0.04 ncrc5633 lP 0.04 ncrc9712 1P 0.04 seoa4202 1P 0.04 seob0564 1P 0.04 seoa4717 1P 0.04 fcrb9407 1P 0.04 seob17.531P 0.04 fcrb5763 1P 0.04 seob1196 1P 0.04 hfcr3183 1P 0.04 seob2735 1P 0.04 fcrb8014 1P 0.04 fcrb6028 1P 0.04 fcrc2849 1P 0.04 fcr2018 1P 0.04 mioc6987 1P 0.04 fcrc0835 1P 0.04 ncrc4600 1P 0.04 hfcr5383 1P 0.04 mioa6545 1P 0.04 fcrb5108 1P 0.04 seoc3487 1P 0.04 mioc4318 1P 0.04 seoc2681' 1P 0.04 ncrc1032 1P 0.04 mioc0424 1P 0.04 miob9748 1P 0.04 mioa0857 1P 0.04 mioc9205 1P 0.04 mioc2911 1P 0.04 seoa1900 1P 0.04 mioc3573 1P 0.04 fcr1098 1P 0.04 mioc7662 1P 0.04 seob2685 1P 0.04 mioa0249 1P 0.04 seob9645 1P 0.04 miob2166 lP 0.04 fcr4503 1P 0.04 mioc0911 1P 0.04 hfcr3224 1P 0.04 mioc2726 1P 0.04 ncrc6825 1P 0.04 miod4752 1P 0.04 ncrc4267 1P 0.04 ncrc9338 1P 0.04 fcrb3654 1P 0.04 seob0261 1P 0.04 fcrc5547 1P 0.04 fcrb2715 1P 0.04 mioc4366 1P 0.04 fcrb3933 1P 0.04 mioc8423 1P 0.04 fcrb5181 1P 0.04 fcr4900 1P 0.04 fcrb6785 1P 0.04 fcrb3083 1P 0.04 fcrb7608 1P 0.04 fcrb4248 1P 0.04 fcrc6345 1P 0.04 fcrb5928 1P 0.04 miob1115 1P 0.04 fcrb6502 1P 0.04 miob5412 1P 0.04 fcrb7723 1P 0.04 mioc0384 1P 0.04 fcrb9371 1P 0.04 mioc1028 1P 0.04 fcrb9856 1P 0.04 mioc2110 1P 0.04 fcrc1740 1P 0.04 mioc6983 1P 0.04 fcrc5137 1P 0.04 miod2845 1P 0.04 hfcr5864 1P 0.04 ncr0547 1P 0.04 mioa1603 1P 0.04 ncr5713 1P 0.04 mioa2261 1P 0.04 ncrbl956 1P 0.04 ~~~ncrb772~6~~ ~~~~~~~ ~~~~0.~04miod3776 1Q 2 .13004e-03 ~~~~~~
~~~~1P

ncrc1103 1P 0.04 ncrb0487 1Q 2.13004e-03 ncrc2128 1P 0.04 fcrb9017 1Q 2.158692e-03 ncrc3045 1P 0.04 ncrc4920 1Q 2.172062e-03 ncrc3735 1P 0.04 seoa1720 1Q 2.200519e-03 ncrc3953 1P 0.04 miob8830 1Q 2.222443e-03 ncrc4575 1P 0.04 fcr0770 1Q 2.292688e-03 ncrc8932 1P 0.04 fcr2006 1Q 2.342381e-03 ncrc9469 1P 0.04 fcr2079 1Q 2.342381e-03 seoa2528 1P 0.04 fcr5779 1Q 2.342381e-03 seoa7295 1P 0.04 mioa1015 1Q 2.342381e-03 seob3517 1P 0.04 ncrc9704 1Q 2.342381e-03 seoc2589 1P 0.04 seob1197 1Q 2.342381e-03 seoc3690 1P 0.04 seoa5235 1Q 2.36091e-03 seoc4824 1P 0.04 mioa5614 1Q 2.467117e-03 ncr0664 1P 0.04 fcr0707 1Q 2.573342e-03 seob0046 1Q 4.27e-05 ncrc3258 1Q 2.573342e-03 mioc7662 1Q 8.24e-05 seoa5433 1Q 2.733174e-03 miod5590 1Q 1.44e-04 miob5010 1Q 2.824307e-03 seob2950 1Q 1.52e-04 seob7082 1Q 2.824307e-03 hfcr3413 1Q 2.00513e-04 seob8099 1Q 2.828664e-03 fcrb2330 1Q 2.07e-04 seoc8306 1Q 2.882251e-03 miob7135 1Q 3.26e-04 fcrb6431 1Q 3.096741e-03 fcrb3454 1Q 3.56e-04 ncr0847 1Q 3.096741e-03 hfcr2616 1Q 3.60223e-04 ncrc9562 1Q 3.096741e-03 seob5324 1Q 4.3e-04 seoa1419 1Q 3.096741e-03 ncrb3768 1Q 4.92e-04 seob1399 1Q 3.096741e-03 fcrb8127 1Q 5.63262e-04 seob0031 1Q 3.219186e-03 fcrb9655 1Q 5.63262e-04 seoa4795 1Q 3.26458e-03 ncrc5653 1Q 6.2804e-04 fcr0997 1Q 3.392193e-03 fcrb3763 1Q 7.78162e-04 . mioa3944 1Q 3.392193e-03 miod6068 1Q 7.78162e-04 mioa5461 1Q 3.392193e-03 seob6432 1.Q 7.78162e-04 mioc0501 1Q 3.392193e-03 fcrb2094 1Q 8.2e-04 mioc1600 1Q 3.392193e-03 fcrc6127 1Q 8.95e-04 ncrc0427 1Q 3.392193e-03 fcrc0959 1Q 9.59902e-04 ncrc9793 1Q 3.392193e-03 seob6015 1Q 1.030399e-03 ncrc9855 1Q 3.392193e-03 ncr1526 1Q 1.063703e-03 seoa0792 1Q 3.392193e-03 fcrc4848 1Q 1.106641e-03 seob5886 1Q 3.592626e-03 fcrc5391 1Q 1.178958e-03 fcrb6359 1Q 3.619754e-03 miob4860 1Q 1.178958e-03 fcr5190 1Q 3.712298e-03 miod5080 1Q 1.254238e-03 fcrb3461 1Q 3.712298e-03 fcrc6228 1Q 1.304487e-03 fcrb5202 1Q 3.712298e-03 mioa0891 1Q 1.304487e-03 mioa6738 1Q 3.712298e-03 ncrc1665 1Q 1.441864e-03 mioc2577 1Q 3.727477e-03 seoa3422 1Q 1.441864e-03 mioa8767 1Q 3.739684e-03 fcrc1019 1Q 1.484277e-03 miod5218 1Q 3.808421e-03 mioc3419 1Q 1.484277e-03 ncrc0174 1Q 3.808421e-03 ncrb3957 1Q 1.484277e-03 mioc0567 1Q 3.983145e-03 ncrc5592 1Q 1.584296e-03 ncrc1751 1Q 4.026284e-03 mioa3620 1Q 1.59205e-03 fcrb1344 1Q 4.058782e-03 hfcr2536 1Q 1.605711e-03 fcrb5994 1Q 4.058782e-03 seoa5547 1Q 1.674509e-03 fcrc0654 1Q 4.058782e-03 fcr0027 1Q 1.75607e-03 mioa3548 1Q 4.058782e-03 miob2960 1Q 1.75607e-03 miob8146 1Q 4.058782e-03 mioc4270 1Q 1.75607e-03 ncrc9284 1Q 4.058782e-03 seoa0085 1Q 1.756632e-03 ncrc9438 1Q 4.058782e-03 seob3112 1Q 1.781563e-03 ncrc9712 1Q 4.058782e-03 fcrb6464 1Q 1.801576e-03 seoa1410 1Q 4.058782e-03 mioc3573 1Q 1.88355e-03 seobl748 1Q 4.058782e-03 seoa4601 1Q 1.912246e-03 ncrc0798 1Q 4.123577e-03 fcr2139 1Q 1.935013e-03 ncr0016 1Q 4.130421e-03 ncrc4135 1Q 1.935013e-03 mioa2072 1Q 4.159968e-03 fcrc5850 1Q 1.963872e-03 seoa6106 1Q 4.168552e-03 seob0763~m~lQ~ ~4.178-879e-03fcrb6738 lQ 6.800293e-03 ~~.~ ~
-~

fcrb52871Q 4.217128e-03 miob8825 1Q 6.800293e-03 fcrb62361Q 4.251227e-03 ncrb8203 1Q 6.800293e-03 mioc56611Q 4.281336e-03 seoa9931 lQ 6.800293e-03 fcrb00721Q 4.290663e-03 seobQ514 1Q ~.800293e-03 ncrc54921Q 4.291267e-03 ncrc9795 1Q 7.052912e-03 fcrb21621Q 4.433463e-03 miob0644 1Q 7.214905e-03 mioa61721Q 4.433463e-03 hfcr0750 1Q 7.323261e-03 ncr3811 1Q 4.433463e-03 seoc3554 1Q 7.323261e-03 ncr9599 1Q 4.433463e-03 fcrb2080 1Q 7.387873e-03 ncrb84681Q 4.433463e-03 fcrb5092 1Q 7.387873e-03 seoa00401Q 4.433463e-03 fcrc0412 1Q 7.387873e-03 seob29661Q 4.433463e-03 hfcr3500 1Q 7.387873e-03 mioa81921Q 4.459835e-03 mioa6102 1Q 7.387873e-03 fcrb36441Q 4.811926e-03 miod0993 1Q 7.387873e-03 fcrb4868lQ 4.821954e-03 miod7418 1Q 7.387873e-03 fcr0894 1Q 4.838255e-03 seob7015 1Q 7.387873e-03 fcrb19901Q 4.838255e-03 mioc2123 1Q 7.393331e-03 fcrc11151Q 4.838255e-03 seoa7178 1Q 7.471525e-03 mioa64761Q 4.838255e-03 fcrb1496 1Q 7.708383e-03 ncr5871 1Q 4.838255e-03 seob2685 1Q 7.994213e-03 ncrc40151Q 4.838255e-03 fcr2132 1Q 8.019199e-03 fcrb56621Q 5.0613442-03 fcrb3483 1Q 8.019199e-03 mioa84841Q 5.124865e-03 mioa3997 1Q 8.019199e-03 seob02501Q 5.168742e-03 mioa6807 1Q 8.019199e-03 seoc32691Q 5.172067e-03 mioc4731 1Q 8.019199e-03 fcrb56451Q 5.275168e-03 ncre5569 1Q 8.019199e-03 fcrc57991Q 5.275168e-03 seoa9930 1Q 8.019199e-03 miod60581Q 5.275168e-03 seob2169 1Q 8.019199e-03 ncrc10371Q 5.275168e-03 ncre1029 1Q 8.067263e-03 ncrc97001Q 5.275168e-03 ncr9934 1Q 8.203479e-03 seoa42841Q 5.340812e-03 seob3503 1Q 8.224653e-03 ncrc03241Q 5.366126e-03 fcrb6442 1Q 8.263283e-03 mioa53261Q 5.506529e-03 ncrb0749 1Q 8.275966e-03 seob61891Q 5.603216e-03 fcrb6262 1Q 8.280361e-03 fcrc06511Q 5.627127e-03 miod2311 1Q 8.338691e-03 ncrc59471Q 5.731692e-03 miod4493 1Q 8.523402e-03 fcrc28081Q 5.746317e-03 fcr3269 1Q 8.696926e-03 mioa33211Q 5.746317e-03 fcr4328 1Q 8.696926e-03 mioa68321Q 5.746317e-03 fcr4782 1Q 8.696926e-03 miob86571Q 5.746317e-03 mioc0528 1Q 8.696926e-03 miocl4161Q 5.746317e-03 miod3509 1Q 8.696926e-03 ncr6142 1Q 5.746317e-03 seoa0429 1Q 8.696926e-03 ncrc99471Q 5.746317e-03 seoa9086 1Q 8.696926e-03 ncrc34081Q 5.837544e-03 seob6856 1Q 8.696926e-03 hfcr59911Q 5.941877e-03 seoc1934 1Q 8.696926e-03 fcrb24241Q 5.995387e-03 miob0213 1Q 8.722193e-03 fcrc55471Q 6.157035e-03 fcr1760 1Q 9.055237e-03 fcrb51141Q 6.253918e-03 miod7081 1Q 9.130039e-03 seoa39081Q 6.253918e-03 miob9462 1Q 9.296152e-03 seob63861Q 6.253918e-03 seob5592 1Q 9.365596e-03 seob80821Q 6.253918e-03 ncrb0027 1Q 9.4225e-03 seoc41871Q 6.255406e-03 fcrb1995 1Q 9.423822e-03 seoc00091Q 6.349542e-03 fcrb5336 1Q 9.423822e-03 fcrb29261Q 6.355659e-03 fcrb5~5& 1Q 9.423822e-03 fcrc0357IQ 6.483101e-03 miob9185 1Q 9.423822e-03 fcrb59181Q 6,544455e-03 ncr3034 IQ 9.423822e-03 mioa22131Q 6.604566e-03 ncrc0151 1Q 9.423822e-03 fcrb433I1Q 6.673407e-03 ncrc7131 1Q 9.423822e-03 seob29361Q 6.715058e-03 seoa0302 1Q 9.423822e-03 seob58991Q 6.749391e-03 seob3360 1Q 9.423822e-03 fcr0419 1Q 6.800293e-03 seob5099 1Q 9.423822e-03 fcr2861 1Q 6,800293e-03 fcrb5455 1Q 9.594248e-03 fcrb45061Q 6.800293e-03 seob6584 1Q 9.61842e-03 ncrcl941 1Q ~9~.~7b9113e-03fcrb4860 1Q 0.01 fcrb5926 1Q 9.821331e-03 fcrb5305 lQ 0.01 hfcr5045 1Q 0.01 fcrb8910 1Q 0.01 ncrb8201 1Q 0.01 fcrc1763 1Q 0.01 fcr5075 1Q 0.01 fcrc4068 1Q 0.01 fcrb9569 1Q 0.01 fcrc6419 1Q 0.01 miob0178 1Q 0.01 fcrc6970 1Q 0.01 seoa1802 1Q 0.01 hfcr2934 1Q 0.01 seoa3121 1Q 0.01 miob8639 1Q 0.01 seob1158 1Q 0.01 mioc4526 1Q 0.01 seob3189 1Q 0.01 miod4084 1Q 0.01 fcrb3704 1Q 0.01 ncr0496 1Q 0.01 ncrc1168 1Q 0.01 ncr3614 1Q 0.01 miob7268 1Q 0.01 ncrc5072 1Q 0.01 mioa2343 1Q 0.01 seoa9777 1Q 0.01 ncr4648 1Q 0.01 seob2807 1Q 0.01 fcr1337 1Q 0.01 seob5418 1Q 0.01 seob6851 1Q 0.01 seob5726 1Q 0.01 ncrc1153 1Q 0.01 seob7891 1Q 0.01 seob6781 1Q 0.01 seob2958 1Q 0.01 fcrb1807 1Q 0.01 fcrb5351 1Q 0.01 fcrb7699 1Q 0.01 fcrb5896 1Q 0.01 fcrc0607 1Q 0.01 ncrc6439 1Q 0.01 fcrc5402 1Q 0.01 seob9898 1Q 0.01 mioa0707 1Q 0.01 fcrc5379 1Q 0.01 mioa6093 1Q 0.01 fcrb4719 1Q 0.01 ncr7813 1Q 0.01 ncrc2715 1Q 0.01 seoc2382 1Q 0.01 fcrb7068 1Q 0.01 seoa1427 1Q 0.01 seob1737 1Q 0.01 fcrb8222 1Q 0.01 ncr8686 1Q 0.01 ncrcl349 1Q 0.01 miob8711 1Q 0.01 seoa4783 1Q 0.01 ncrcl615 1Q 0.01 seoa4739 1Q 0.01 ncrb6949 1Q 0.01 fcr1997 1Q 0.01 mioc5226 1Q 0.01 fcrb1387 1Q 0.01 ncrc5844 1Q 0.01 fcrb5562 1Q 0.01 ncrc7023 1Q 0.01 fcrb5702 1Q 0.01 fcrb2979 1Q 0.01 fcrc0839 1Q 0.01 fcrc0959 1Q 0.01 fcrc6381 1Q 0.01 mioa1655 1Q 0.01 mioa3392 1Q 0.01 mioa9555 1Q 0.01 mioa3945 1Q 0.01 miob6351 1Q 0.01 miob2569 1Q 0.01 miod7270 1Q 0.01 miob7922 1Q 0.01 ncr1437 1Q 0.01 mioc6345 1Q 0.01 ncrb8319 1Q 0.01 ncrb0513 1Q 0.01 ncrc3141 1Q 0.01 ncrb4477 1Q 0.01 ncrc4864 1Q 0.01 ncrc2675 1Q 0.01 seoa1644 1Q 0.01 ncrc7171 1Q 0.01 seoa2765 1Q 0.01 seoa1552 1Q 0.01 seoa4587 1Q 0.01 seob0058 1Q 0.01 seoa9705 1Q 0.01 seob0263 1Q 0.01 seob1093 1Q 0.01 seob1009 1Q 0.01 seob1757 1Q 0.01 seob1322 1Q 0.01 seoc1664 1Q 0.01 seob5891 1Q 0.01 seoc2248 1Q 0.01 seoc1906 1Q 0.01 seoc4824 1Q 0.01 seoc3588 1Q 0.01 ncrc2857 1Q 0.01 fcrc1654 1Q 0.01 ncrc4815 1Q 0.01 ncr2575 1Q 0.01 miob9228 1Q 0.01 mioc5736 1Q 0.01 fcrb5389 1Q 0.01 fcrb6779 1Q 0.01 seoa7157 1Q 0.01 fcr4846 1Q 0.01 miob0865 1Q 0.01 fcrb9007 1Q 0.01 miob7985 1Q 0.01 fcr1555 1Q 0.01 mioc2728 1Q 0.01 fcrb2759 1Q 0.01 fcrb5087 1Q 0.01 seoa5894~-~~~ ~~~TO.O~l seob2149 1Q 0.01 ~m~
1.Q~

fcrb51771Q 0.01 seob4192 1Q 0.01 seoc22211Q 0.01 seob4293 1Q 0.01 seob18531Q 0.01 seob7465 1Q 0.01 fcrc47341Q 0.01 mioc4103 1Q 0.01 mioc05301Q 0.01 seoa5746 1Q 0.01 miod53101Q 0.01 fcr0999 1Q 0.01 ncr13871Q 0.01 ncr0075 1Q 0.01 ncr48601Q 0.01 ncr8067 1Q 0.01 ncrb65571Q 0.01 ncrc3011 1Q 0.01 ncrc28311Q 0.01 ncrc9010 1Q 0.01 seoa34081Q 0.01 hfcr3902 1Q 0.01 seoa95661Q 0.01 ncrb8343 1Q 0.01 seob05341Q 0.01 ncrc1765 1Q 0.01 seoa28011Q 0.01 ncrc3413 1Q 0.01 fcrb87141Q 0.01 fcrc4408 1Q 0.01 fcrb24831Q 0.01 ncrc2290 1Q 0.01 fcrb60621Q 0.01 fcrb3073 1Q 0.01 seoa08601Q 0.01 fcrb9118 1Q 0.01 seoa00231Q 0.01 fcrc2216 1Q 0.01 ncr80961Q 0.01 mioa7299 1Q 0.01 seob04831Q 0.01 mioa8147 1Q 0.01 mioa47821Q 0.01 mioa8796 1Q 0.01 mioa24131Q 0.01 miod1316 1Q 0.01 fcr39361Q 0.01 ncr3815 1Q 0.01 fcrb45151Q 0.01 ncr6943 1Q 0.01 fcrc01301Q 0.01 ncrb8398 1Q 0.01 fcrc06371Q 0.01 ncrc3104 1Q 0.01 fcrc64701Q 0.01 ncrc3526 1Q 0.01 mioa21851Q O.Ol seoa1910 1Q 0.01 ncr70501Q 0.01 seob3367 1Q 0.01 ncr9060lQ 0.01 seocl345 1Q 0.01 ncrc20801Q 0.01 seoc2218 1Q 0.01 seob09761Q 0.01 seoa9712 1Q 0.01 fcrb20771Q 0.01 miod6018 1Q 0.01 fcrb52531Q 0.01 seob0831 1Q 0.01 fcrc51691Q 0.01 mioc2385 1Q 0.01 ncrc12591Q 0.01 ncrc6953 1Q 0.01 ncrb36281Q 0.01 seob2l95 1Q 0.01 fcrb67181Q 0.01 seoa9997 1Q 0.01 ncrc24131Q 0.01 fcrc0210 1Q 0.01 fcrb44091Q 0.01 miob9559 lQ 0.01 fcrb2546lQ 0.01 mioc2726 1Q 0.01 seob19561Q 0.01 mioa1933 1Q 0.02 miob85651Q 0.01 seoa5741 1Q 0.02 fcrb42781Q 0.01 fcrb2049 1Q 0.02 mioa08201Q 0.01 miob0942 1Q 0.02 fcrb42321Q 0.01 mioc2152 1Q 0.02 ncrc46541Q 0.01 miod7227 1Q 0.02 fcr06651Q 0.01 ncr0836 1Q 0.02 fcr08371Q 0.01 ncrc0856 1Q 0.02 fcrb18761Q 0.01 seoa1653 1Q 0.02 fcrb31271Q 0.01 miob9248 1Q 0.02 fcrb36541Q 0.01 fcrb6171 1Q 0.02 fcrb59291Q 0.01 fcrb8119 1Q 0.02 fcrb86281Q 0.01 fcrb5204 1Q 0.02 fcrc58981Q 0.01 fcrb3314 1Q 0.02 miob71561Q 0.01 miob9781 1Q 0.02 miod06251Q 0.01 seoa0501 1Q 0.02 miod19081Q 0.01 seob0182 1Q 0.02 ncr85941Q 0.01 mioc1245 1Q 0.02 ncrc69811Q 0.01 ncrc9819 1Q 0.02 seoa17651Q 0.01 seoa1770 1Q 0.02 seoa36621Q 0.01 miob4613 1Q 0.02 seoc17851Q U.UZ fcrc2439 1Q 0.02 fcr3323 1Q 0.02 fcrc5372 1Q 0.02 fcrb20131Q 0.02 hfcr5228 1Q 0.02 fcrb25161Q 0.02 hfcr5237 1Q 0.02 fcrb60311Q 0.02 mioa7957 1Q 0,02 fcrc18341Q 0.02 mioa9505 1Q 0.02 fcrc51421Q 0.02 miob4090 1Q 0.02 fcrc56141Q 0.02 miob9805 1Q 0.02 hfcr05601Q 0.02 miod5369 1Q 0.02 mioa42851Q 0.02 miod6263 1Q 0.02 miob17891Q 0.02 ncrc5724 IQ 0.02 miob71361Q 0.02 ncrc9867 1Q 0.02 mioc11251Q 0.02 seoa2381 1Q 0.02 miod72251Q 0.02 seoa2817 1Q 0.02 ncr0547 1Q 0.02 seoa2837 1Q 0.02 ncrb81891Q 0.02 seoa5253 1Q 0.02 ncrc44481Q 0.02 seoa6151 1Q 0.02 seoa01451Q 0.02 seoal646 1Q 0.02 seoa83511Q 0.02 ncrc6503 1Q 0.02 seob77291Q 0.02 seoa5387 1Q 0.02 seob88391Q 0.02 seob3378 1Q 0.02 seob901'71Q 0.02 fcrc5270 1Q 0.02 fcrb94011Q 0.02 ncrc1606 1Q 0.02 hfcr53831Q 0.02 fcrb5725 1Q 0.02 fcrb19161Q 0.02 fcrb5507 1Q 0.02 mioa76271Q 0.02 mioc0546 1Q 0.02 hfcr26931Q 0.02 fcrb2317 1Q 0.02 seob02981Q 0.02 ncrb8171 1Q 0.02 fcrb15031Q 0.02 fcr0529 1Q 0.02 fcrb64321Q 0.02 fcr2218 1Q 0.02 fcrc60411Q 0.02 fcrb4925 1Q 0.02 mioa03321Q 0.02 fcrb6917 1Q 0.02 mioa14171Q 0.02 fcrc2050 1Q 0.02 mioa65521Q 0.02 hfcr2041 1Q 0,02 mioc8016IQ 0.02 hfcr2323 1Q 0.02 miod08071Q 0.02 hfcr5620 1Q 0.02 miod09781Q 0.02 ncrc2859 1Q 0.02 ncr3961 1Q 0.02 ncrc6888 1Q 0.02 ncr5488 1Q 0.02 seoa1599 1Q 0.02 ncrb83831Q 0.02 seoa4017 1Q 0.02 ncrc19951Q 0.02 seoa9792 1Q 0.02 ncrc94281Q 0.02 seoa9981 1Q 0.02 seoa40121Q 0.02 seobl766 1Q 0.02 seob14111Q 0.02 seob7419 1Q 0.02 seob64151Q 0.02 ncrc0803 1Q 0.02 seoa89091Q 0.02 miod1863 1Q 0.02 fcrb24601Q 0.02 fcrb9355 1Q 0.02 fcrb94991Q 0.02 seob1898 1Q 0.02 miob94951Q 0.02 miod4367 1Q 0.02 fcrb47811Q 0.02 miob8627 1Q 0.02 fcrc04301Q 0.02 fcrb3017 1Q 0.02 ncrc67961Q 0.02 ncr0133 1Q 0.02 mioc35491Q 0.02 ncrc0583 1Q 0.02 ncrc04421Q 0.02 fcrc5831 1Q 0.02 fcrb43421Q 0.02 seoa3639 1Q 0.02 fcrb89591Q 0.02 seoa9209 1Q 0.02 fcrb23341Q 0.02 ncrc4296 1Q 0.02 mioc06211Q 0.02 seoa4246 1Q 0.02 ncrb74651Q 0.02 fcrb8485 1Q 0.02 miob39421Q 0.02 seoc2670 1Q 0.02 fcr0768 1Q 0.02 ncrc9304 1Q 0.02 fcr5712 1Q 0.02 miob2825 1Q 0.02 fcrb42311Q 0.02 seoc4585 1Q 0.02 fcrb74431Q 0.02 fcr0893 1Q 0.02 22~

fcr4212 ~1Q 0.0~ fcrb1940 1Q 0.03 fcrb80141Q 0.02 seob3139 1Q 0.03 mioal4731Q 0.02 seob7575 1Q 0.03 mioa45421Q 0.02 seob5673 1Q 0.03 mioa87741Q 0.02 fcrb1909 1Q 0.03 miob31201Q 0.02 fcrb2166 1Q 0.03 miob71051Q 0.02 fcrb2933 1Q 0.03 miob96711Q 0.02 fcrb3080 1Q 0.03 mioc25921Q 0.02 fcrb3539 1Q 0.03 miod32541Q 0.02 fcrb4799 1Q 0.03 ncrc44441Q 0.02 fcrb8080 1Q 0.03 seoa04691Q 0.02 mioa4245 1Q 0.03 seoa15751Q 0.02 miob7290 1Q 0.03 seoa17891Q 0.02 mioc3962 1Q 0.03 seoa22441Q 0.02 miod3826 1Q 0.03 seoa98731Q 0.02 ncrc5631 1Q 0.03 seob21391Q 0.02 seoa1584 1Q 0.03 seob55791Q 0.02 seoa1924 1Q 0.03 seob72001Q 0.02 seoa3429 1Q 0.03 seob81941Q 0.02 seoa3910 1Q 0.03 seob83291Q 0.02 seoa9709 1Q 0.03 seob91451Q 0.02 seob2717 1Q 0.03 seoc38701Q 0.02 seob4584 1Q 0.03 fcrb78521Q 0.02 mioc1357 1Q 0.03 seoc49001Q 0.02 seob7929 1Q 0.03 miob86871Q 0.02 ncrc2007 1Q 0.03 ncr0612 1Q 0.02 miob1814 1Q 0.03 seob59541Q 0.02 ncrc5039 1Q 0.03 mioa153210 0.02 seob2161 1Q 0.03 fcrc48761Q 0.02 miob8418 1Q 0.03 seoa93631Q 0.02 fcr1557 1Q 0.03 fcrb51731Q 0.02 ncrc1502 1Q 0.03 fcrc28291Q 0.03 fcrb2190 1Q 0.03 miod45121Q 0.03 seoa1318 1Q 0.03 seoa35441Q 0.03 miod0777 1Q 0.03 seoa40531Q 0.03 fcrb5726 1Q 0.03 ncrc22891Q 0.03 seob1586 1Q 0.03 ncrb00461Q 0.03 miob9124 1Q 0.03 miob66101Q 0.03 ncrc1193 1Q 0.03 seoa44641Q 0.03 seob3226 1Q 0.03 fcrb21241Q 0.03 miob8454 1Q 0.03 fcrb65081Q 0.03 ncrc9338 1Q 0.03 fcrb65091Q 0.03 miob0154 1Q 0.03 fcrb82151Q 0.03 mioal496 1Q 0.03 fcrb96111Q 0.03 ncrc3358 1Q 0.03 fcrc46881Q 0.03 seob0810 1Q 0.03 fcrc72431Q 0.03 fcrb4926 1Q 0.03 hfcr30891Q 0.03 fcrb8515 1Q 0.03 mioa79551Q 0.03 fcrb8719 1Q 0.03 mioc74211Q 0.03 fcrc6976 1Q 0.03 miod48671Q 0.03 fcrc7228 1Q 0.03 ncrb71771Q 0.03 hfcr3486 1Q 0.03 ncrc12031Q 0.03 hfcr5009 1Q 0.03 ncrc24951Q 0.03 hfcr6677 1Q 0.03 ncrc37771Q 0.03 mioa1657 1Q 0.03 ncrc69961Q 0.03 mioa6999 1Q 0.03 seob75051Q 0.03 miob0361 1Q 0.03 ncrb01641Q 0.03 miob6029 1Q 0.03 ncrc03421Q 0.03 mioc3716 1Q 0.03 hfcr19151Q 0.03 mioc8945 1Q 0.03 seob67511Q 0.03 ncrb6261 1Q 0.03 fcrb57961Q 0.03 ncrb7386 1Q 0.03 miob67211Q 0.03 ncrc3802 1Q 0.03 mioc39061Q 0.03 seoa5544 1Q 0.03 seob1538~~~~ ~~~~~~~0~.-03mioal524 1Q 0.
~~~~ 1Q~~ ~ 03 seob29591Q 0.03 miob2375 1Q 0.03 seoc09571Q 0.03 miob3693 1Q 0.03 seoc51251Q 0.03 miob9614 1Q 0.03 ncr3700 1Q 0.03 mioc0662 lQ 0.03 fcrb48371Q 0.03 mioc4119 1Q 0.03 seobl3231Q 0.03 mioc8471 1Q 0.03 fcrb92161Q 0.03 miod6961 1Q 0.03 seob37311Q 0.03 ncr3163 lQ 0.03 ncrc43131Q 0.03 ncr5713 1Q 0.03 fcrb53601Q 0.03 ncr6878 1Q 0.03 fcrbl6941Q 0.03 ncrb8569 1Q 0.03 miob85321Q 0.03 ncrc9469 1Q 0.03 fcr1633 1Q 0.03 seoa5662 1Q 0.03 fcrb75881Q 0.03 seob2658 1Q 0.03 fcrc50071Q 0.03 seob5044 1Q 0.03 ncrc04241Q 0.03 seob6087 1Q 0.03 seob20111Q 0.03 seob8425 1Q 0.03 seob87861Q 0.03 seoc3469 1Q 0.03 mioc01211Q 0.03 ncrc0259 1Q 0.03 hfcr65091Q 0.03 seoa5520 1Q 0.03 seob58121Q 0.03 seoa0014 1Q 0.03 fcr3880 1Q 0.03 ncrb8670 1Q 0.04 fcrb24521Q 0.03 fcrb8133 lQ 0.04 fcrb47881Q 0.03 fcrb2804 1Q 0.04 fcrb75101Q 0.03 mioc2961 1Q 0.04 fcrb82081Q 0.03 fcrb9253 1Q 0.04 fcrc18491Q 0.03 miod1333 1Q 0.04 fcrc62821Q 0.03 miob5495 1Q 0.04 hfcr66801Q 0.03 seoc2622 1Q 0.04 mioa13701Q 0.03 fcrb2472 1Q 0.04 mioa89981Q 0.03 ncrc0413 1Q 0.04 mioa98211Q 0.03 seoa6172 lQ 0.04 mioc11221Q 0.03 fcrc5092 1Q 0.04 miod44671Q 0.03 mioa4753 1Q 0.04 ncr0144 1Q 0.03 fcrb4542 1Q 0.04 ncr5374 1Q 0.03 ncr3419 1Q 0.04 ncrc57751Q 0.03 seob6670 1Q 0.04 seoa10801Q 0.03 fcrb9449 1Q 0.04 seoa17491Q 0.03 fcrb5108 1Q 0.04 seoa72961Q 0.03 fcrb2321 1Q 0.04 seob01571Q 0.03 fcrb4372 1Q 0.04 seob85001Q 0.03 fcr0253 1Q 0.04 seob88071Q 0.03 fcrb1380 1Q 0.04 mioc10491Q 0.03 fcrb2208 1Q 0.04 ncr3827 1Q 0.03 fcrb7617 1Q 0.04 seoa43171Q 0.03 hfcr5691 1Q 0.04 fcr5123 1Q 0.03 mioa9294 1Q 0.04 ncrb03231Q 0.03 miob8583 1Q 0.04 fcrc05221Q 0.03 mioc2872 1Q 0.04 seoc02761Q 0.03 mioc7818 1Q 0.04 seob18181Q 0.03 mioc8782 1Q 0.04 seob49281Q 0.03 miod5651 1Q 0.04 mioc75421Q 0.03 ncr4551 1Q 0.04 seoc00341Q 0.03 ncr6144 1Q 0.04 fcrb27151Q 0.03 ncrb0262 1Q 0.04 ncrc08641Q 0.03 ncrb1179 1Q 0.04 seob11531Q 0.03 ncrc1949 1Q 0.04 mioa90671Q 0.03 seoa3088 1Q 0.04 ncr8966 1Q 0.03 seoa3108 1Q 0.04 fcr2182 1Q 0.03 seoa3578 1Q 0.04 fcr4214 1Q 0.03 seoa5784 1Q 0.04 hfcr17091Q 0.03 seoa8979 1Q 0.04 mioa06841Q 0.03 seob0937 1Q 0.04 seob38871Q 004 seoa3694 1Q 0.04 seob82611Q 0.04 seoa5461 1Q 0.04 seoc52181Q 0.04 hfcr5860 1Q 0.04 ncr3396 1Q 0.04 fcrb8808 1Q 0.04 fcrc20081Q 0.04 fcrb2034 1Q 0.04 seoa17371Q 0.04 fcrb2993 1Q 0.04 mioc41611Q 0.04 fcrb3083 1Q 0.04 mioa41831Q 0.04 fcrb7608 1Q 0.04 hfcr34531Q 0.04 fcrb8334 1Q 0.04 seob16601Q 0.04 fcrc1745 1Q 0.04 ncrc07471Q 0.04 mioa0294 1Q 0.04 ncrc30761Q 0.04 mioc0999 1Q 0.04 mioa13031Q 0.04 ncr0644 1Q 0.04 fcrb00971Q 0.04 ncrb7167 1Q 0.04 miob85861Q 0.04 ncrc3544 1Q 0.04 miob91211Q 0.04 seoa1559 1Q 0.04 hfcr34451Q 0.04 seoa5977 1Q 0.04 fcrb03551Q 0.04 seoa9389 1Q 0.04 fcr3757 1Q 0.04 seob2987 1Q 0.04 fcrb25451Q 0.04 seob6133 1Q 0.04 fcrc01981Q 0.04 seob7530 1Q 0.04 fcrc27241Q 0.04 seoc0866 1Q 0.04 hfcr53811Q 0.04 seoc4561 1Q 0.04 hfcr63701Q 0.04 fcrb4929 1Q 0.04 miob34741Q 0.04 seob0751 1Q 0.04 mioc39041Q 0.04 fcrb2308 1Q 0.04 miod25561Q 0.04 fcrb8765 1Q 0.04 miod46861Q 0.04 miod7408 1Q 0.04 miod74141Q 0.04 seoa5552 1Q 0.04 ncrb25241Q 0.04 ncrcllll 1Q 0.04 ncrc18851Q 0.04 fcrc0529 1Q 0.04 ncrc97721Q 0.04 miob'7435 1Q 0.04 seoa07831Q 0.04 seoa7366 1Q 0.04 seoal7391Q 0.04 miod5672 1Q 0.04 seoa21811Q 0.04 ncrb8559 1Q 0.04 seoa38211Q 0.04 seob4570 1Q 0.04 seoa84011Q 0.04 seoc7566 1R 4.46e-05 seob22831Q 0.04 miod1450 1R 1.6e-04 seob41971Q 0.04 mioc8437 1R 1.73e-04 seob53791Q 0.04 fcrb8940 1R 2.2e-04 seob64461Q 0.04 seoc1311 1R 3.9e-04 seoc56121Q 0.04 mioc8016 1R 4.68e-04 seob01851Q 0.04 seob4333 1R 4.76e-04 fcr1004 1Q 0.04 fcrb5007 1R 5.07e-04 fcrb45701Q 0.04 ncr0429 1R 5.52e-04 mioa20131Q 0.04 miod5672 1R 7.23e-04 seob65721Q 0.04 ncrc9279 1R 7.82e-04 seoc10231Q 0.04 seob3982 1R 9.0e-04 miob04961Q 0.04 fcrb8243 1R 9.59e-04 fcrb56391Q 0.04 ncrb8721 1R 1.162766e-03 mioa39631Q 0.04 seoa8979 1R 1.319654e-03 ncrc06461Q 0.04 fcrc5071 1R 1.418474e-03 miob58291Q 0.04 seob4036 1R 1.454856e-03 ncr3404 1Q 0.04 mioa9821 1R 1.580436e-03 seob65411Q 0.04 fcrb7505 1R 1.701834e-03 seob11551Q 0.04 ncr7876 1R 1.750917e-03 seob74631Q 0.04 ncrc5758 1R 2.140366e-03 ncrc43841Q 0.04 fcr0788 1R 2.965188e-03 fcrb27491Q 0.04 mioa5681 1R 2.965188e-03 fcrb69901Q 0.04 ncr3971 1R 2.965188e-03 ncr3830 1Q 0.04 fcrb1856 1R 3.128651e-03 fcrb21991Q 0.04 mioc2082 1R 3.266717e-03 fcrb18901Q 0.04 fcrb7487 1R 3.605772e-03 seob62171Q 0.04 seob5612 1R 3.815575e-03 .~u : ~.r,:
hfcr~6085~~", ~'. ~~~~~~15~'e-03mioa9429 1R 0.
~'~t'' 01 ""

mioa09361R 3.869158e-03 ncr4984 1R 0.01 ncr5473 1R 4.172868e-03 ncrc4996 1R 0.01 ncr2995 1R 4.406314e-03 ncrc6846 1R 0.01 seob11331R 4.406314e-03 seoa5303 1R 0.01 fcrb64811R 4.591403e-03 seob3503 1R 0.01 miob48641R 5.008048e-03 seoc0203 1R 0.01 fcrb45511R 5.070437e-03 ncr7037 1R 0.01 seob91241R 5.140962e-03 ncr6197 1R 0.01 fcrc03021R 5.314391e-03 fcr0469 1R 0.01 ncr7284 1R 5.341076e-03 seoa8432 1R 0.01 fcr3973 1R 5.680774e-03 seob0999 1R 0.01 hfcr06761R 5.680774e-03 fcrc0700 1R 0.01 ncr3419 1R 5.680774e-03 fcrc0287 1R 0.01 seob85621R 5.799323e-03 miob7610 1R 0.01 ncr6415 1R 5.811451e-03 seoa1065 1R 0.01 seob45391R 5.985149e-03 ncrb5227 1R 0.01 mioc09781R 6.234951e-03 seoa9373 1R 0.01 ncr2160 1R 6.38343e-03 fcrb4311 1R 0.01 miod43481R 6.431377e-03 fcrb5840 1R 0.01 ncr9123 1R 6.431377e-03 mfob1059 1R 0.01 mioc23811R 6.542403e-03 miob3426 1R 0.01 fcrc62281R 6.563288e-03 mioc6345 1R 0.01 fcrc71801R 6.981505e-03 miod2834 1R 0.01 fcrb61061R 7.267233e-03 ncrb8330 1R 0.01 mioa69691R 7.267233e-03 ncrc2463 1R 0.01 miob43461R 7.267233e-03 ncrc9010 1R 0.01 miob90731R 7.267233e-03 seoa7094 1R 0.01 ncr4993 1R 7.267233e-03 fcrc6397 1R 0.01 seoa81951R 7.267233e-03 hfcr0560 1R 0.01 seob47261R 7.267233e-03 ncr6893 1R 0.01 seoc04381R 7.267233e-03 fcrb2800 1R 0.01 seoc21311R 7.267233e-03 fcr5470 1R 0.01 mioc28871R 7.625229e-03 mioc1978 1R 0.01 hfcr39221R 7.634307e-03 ncrb3314 1R 0.01 fcrb98591R 7.640161e-03 hfcr6043 1R 0.01 fcrb35501R 8.107057e-03 fcrb7430 1R 0.01 hfcr53831R 8.196224e-03 fcrb8196 1R 0.01 miob65361R 8.196224e-03 mioa7088 1R 0.01 ncr7768 1R 8.23771e-03 miod6552 1R 0.01 miob94951R 8.352358e-03 ncr5890 1R 0.01 mioa34281R 8.522496e-03 ncr7097 1R 0.01 fcr3033 1R 8.949232e-03 ncrb7635 1R 0.01 miod40831R 8.954907e-03 ncrc0090 1R 0.01 mioa61721R 9.121639e-03 seoa7403 1R 0.01 ncr7852 1R 9.121639e-03 seob0872 1R 0.01 seoa53961R 9.201027e-03 seoc5627 1R 0.01 fcr3987 1R 9.226752e-03 fcrb6269 1R 0.01 fcrb66661R 9.226752e-03 mioa8796 1R 0.01 mioa86471R 9.226752e-03 ncrb8134 1R 0.01 ncr5709 1R 9.226752e-03 ncr7631 1R 0.01 ncrb87511R 9.226752e-03 fcr5092 1R 0.01 miod62341R 9.25374e-03 hfcr1189 1R 0.01 fcr3269 1R 9.538296e-03 fcrb1977 1R 0.01 fcr2986 1R 9.702695e-03 mioc0371 1R 0.01 ncr3402 1R 9.933523e-03 mioc3603 1R 0.01 ncrc22901R 9.933523e-03 ncr2581 1R 0.01 ncrc40761R 9.933523e-03 ncrc1003 1R 0.01 ncr0159 1R 9.971493e-03 seoa4174 1R 0.01 hfcr59701R 0.01 fcrb9225 1R 0.01 fcr0018 1R 0.01 miod1825 1R 0.01 fcr0288 1R 0.01 ncrc9739 1R 0.01 fcr3005 1R 0.01 fcrc5418 1R 0.01 mioa60931R 0.01 miod2635 1R 0.01 seoc0920"~""'.'~F ~,.. ~.1 ncrb11G3 1R 0 .
~.k.~ 02 a,.~.

fcrb87281R 0.01 ncrbG385 1R 0.02 seoa55781R 0.01 ncrc9549 1R 0.02 fcrb93241R 0.01 seoa7899 1R 0.02 fcrOG37 1R 0.01 seoa8636 1R 0.02 fcrc28291R 0.01 ncr7753 1R 0.02 hfcr05211R 0.01 seob6541 1R 0.02 mioa20791R 0.01 mioc1002 1R 0.02 mioa22041R 0.01 seoc4260 1R 0.02 miob18731R 0.01 ncr1692 1R 0.02 miob23551R 0.01 seob6781 1R 0.02 miob28581R 0.01 hfcr3G74 1R 0.02 miob34741R 0.01 ncr3965 1R 0.02 miod24831R 0.01 fcrc6460 1R 0.02 ncr1780 1R 0.01 seoc2G81 1R 0.02 ncr7570 1R 0.01 seoa1747 1R 0.02 seoa18831R 0.01 mioa8380 1R 0.02 seoa28011R 0.01 seob5203 1R 0.02 seoa729G1R 0.01 fcrb5720 1R 0.02 seob40751R 0.01 ncrc6817 1R 0.02 seob898G1R 0.01 hfcr0400 1R 0.02 seoc225G1R 0.01 fcr1060 1R 0.02 miob80771R 0.01 fcr3599 1R 0.02 miob90201R 0.01 fcrc5721 1R 0.02 mioc15981R 0.01 fcrc6582 lR 0.02 miob97881R 0.01 hfcr2708 1R 0.02 miod15781R 0.01 miob2093 1R 0.02 hfcr2G42lR 0.01 miob8572 1R 0.02 fcrc20401R 0.01 miod4184 1R 0.02 fcrc26191R 0.01 miod4998 1R 0.02 ncrb59091R 0.01 miod6243 1R 0.02 fcr2598 1R 0.01 ncr632G 1R 0.02 seobG25G1R 0.01 ncr7178 1R 0.02 miod19081R 0.01 ncrc1102 1R 0.02 fcr0470 1R 0.01 seoa73G9 1R 0.02 fcrOG80 1R 0.01 seob1061 1R 0.02 fcrb21901R 0.01 seob1808 1R 0.02 fcrb45701R 0.01 seoc1535 1R 0.02 mioa16031R 0.01 seoc140G 1R 0.02 mioa55941R 0.01 fcrc5547 1R 0.02 miod13581R 0.01 ncrb0328 1R 0.02 seob57731R 0.01 seoc5267 1R 0.02 mioc49781R 0.01 fcrb9499 1R 0.02 fcrc07201R 0.01 mioc4290 1R 0.02 ncrcG8751R 0.01 seoa1460 1R 0.02 seoa56911R 0.01 seoa9302 1R 0.02 fcrb21971R 0.01 seob7047 1R 0.02 fcrb58671R 0.01 fcrb2159 1R 0.02 fcrb77231R 0.01 seoa4284 1R 0.02 mioc390G1R 0.01 seob4424 1R 0.02 fcrc51541R 0.01 ncrc9483 1R 0.02 hfcr32071R 0.01 seoa5766 1R 0.02 ncrc95251R 0.01 fcr1068 1R 0.02 fcr0553 1R 0.02 ncrc5553 1R 0.02 fcrb86231R 0.02 seoa7897 1R 0.02 fcrb9G021R 0.02 seob5730 1R 0.02 fcrc51261R 0.02 seob9122 1R 0.02 hfcr57191R 0.02 seob6415 1R 0.02 mioa44841R 0.02 fcr0793 1R 0.02 mioc20211R 0.02 fcrc5136 1R 0.02 mioc84121R 0.02 mioa0890 1R 0.02 miod19371R 0.02 mioa2528 1R 0.02 ncrb04871R 0.02 miob2668 1R 0.02 ncrb0G021R 0.02 miob3942 1R 0.02 ~...~~ .,.
...~., 1'1~'~~"~~:'~~ fcrc5850 1R 0.03 "
ncr9779 "'~

ncrb0367 1R 0.02 mioa6585 1R 0.03 seob0949 1R 0.02 mioc2961 1R 0.03 seob9871 1R 0.02 miod1885 1R 0.03 fcr1608 1R 0.02 ncrc5232 lR 0.03 fcr3181 1R 0.02 seoa7366 1R 0.03 seocl957 1R 0.02 seoa8609 1R 0.03 ncrc3526 1R 0.02 seob3545 1R 0.03 ncr6811 1R 0.02 seob6020 1R 0.03 fcrb6410 lR 0.02 seob6680 1R 0.03 fcrb1924 1R 0.02 seoa5525 1R 0.03 seoa6387 1R 0.02 mioc7296 1R 0.03 seoa0221 1R 0.02 seob6132 1R 0.03 ncrc1374 1R 0.02 seoc3269 1R 0.03 mioa0707 1R 0.02 miob8698 1R 0.03 hfcr0543 1R 0.02 mioc4769 1R 0.03 fcrc4390 1R 0.02 ncrb3638 1R 0.03 fcr3539 1R 0.02 seobl238 1R 0.03 fcr5536 1R 0.02 fcr2729 1R 0.03 fcrb2460 1R 0.02 mioc3523 1R 0.03 fcrb3169 1R 0.02 ncrc9168 1R 0.03 fcrb5440 1R 0.02 fcrb2484 1R 0.03 fcrb7885 1R 0.02 fcrb8808 1R 0.03 fcrc0042 1R 0.02 seob8639 1R 0.03 fcrc3998 1R 0.02 miob9678 1R 0.03 fcrc6560 1R 0.02 hfcr0383 1R 0.03 hfcr3446 1R 0.02 fcr3620 1R 0.03 hfcr6322 1R 0.02 fcrb3201 1R 0.03 ~

mioa1055 1R 0.02 ncrc0744 1R 0.03 mioa6102 1R 0.02 seoc2218 1R 0.03 mioa6567 1R 0.02 fcrb2094 1R 0.03 mioc4552 1R 0.02 fcrb6785 1R 0.03 ncr3588 1R 0.02 mioc7904 1R 0.03 ncrb3768 1R 0.02 ncr3434 1R 0.03 ncrb5837 1R 0.02 ncrb8385 1R 0.03 ncrb8102 1R 0.02 ncrc2868 1R 0.03 ncrc2807 1R 0.02 ncrb8790 1R 0.03 seoa0388 1R 0.02 fcr7415 1R 0.03 seoa6510 1R 0.02 fcrb4932 1R 0.03 seob9353 1R 0.02 fcrb4995 1R 0.03 seoc2510 1R 0.02 fcrb5688 1R 0.03 seob5064 1R 0.02 fcrc5713 1R 0.03 fcr4568 1R 0.02 hfcr4007 1R 0.03 seoa2385 1R 0.02 mioa0874 1R 0.03 seoa7077 1R 0.02 miobl789 1R 0.03 fcrb8891 1R 0.02 mioc0699 1R 0.03 miob8487 1R 0.02 mioc2587 1R 0.03 fcrb4918 1R 0.02 mioc2728 1R 0.03 miob2466 1R 0.02 mioc7895 1R 0.03 seob1100 1R 0.02 ncr6232 1R 0.03 fcrc4846 1R 0.02 ncr7923 1R 0.03 fcr2935 1R 0.02 ncrb2013 1R 0.03 seob8839 1R 0.02 ncrb3444 1R 0.03 fcrc0839 1R 0.02 ncrb4402 1R 0.03 ncrc2531 1R 0.02 ncrc1192 1R 0.03 ferb4504 1R 0.02 ncrc4411 1R 0.03 fcrc5092 1R 0.02 seoa4167 1R 0.03 seob8311 1R 0.02 seoa5366 1R 0.03 seoc2518 1R 0.02 seob3523 1R 0.03 seoa0010 1R 0.02 seob8242 1R 0.03 fcr2293 1R 0.03 seoc4960 1R 0.03 fcrb6141 1R 0.03 seoc5134 1R 0.03 fcrb6817 1R 0.03 mioc1049 1R 0.03 fcrc5113 1R 0.03 mioc3726 1R 0.03 mioa30841R 0.03 mioa8622 1R 0.04 mioc39621R 0.03 fcr3379 1R 0.04 ncr0138 1R 0.03 fcrb9151 1R 0.04 ncr8199 1R 0.03 fcrb9620 1R 0.04 seob11581R 0.03 hfcr3467 1R 0.04 seob18911R 0.03 hfcr3615 1R 0.04 fcrb43421R 0.03 hfcr5237 1R 0.04 hfcr31101R 0.03 mioa0891 1R 0.04 seoc50061R 0.03 mioa5531 1R 0.04 seoc25891R 0.03 mioa8820 1R 0.04 fcrb51991R 0.03 mioc2828 1R 0.04 mioa88511R 0.03 ncr1712 1R 0.04 ncrb71021R 0.03 ncr1768 1R 0.04 ncrc31001R 0.03 ncr7595 1R 0.04 fcr2018 1R 0.03 ncrb3373 1R 0.04 fcr2684 1R 0.03 ncrb7211 1R 0.04 miod51981R 0.03 seoa0207 1R 0.04 mioc11261R 0.03 seoa5575 1R 0.04 ncr4126 1R 0.03 seob6812 1R 0.04 seob87861R 0.03 fcrc2745 1R 0.04 mioa08201R 0.03 mioa0245 1R 0.04 fcrb51141R 0.03 mioa0763 1R 0.04 ncr9781 1R 0.03 hfcr3444 1R 0.04 seoa50901R 0.03 mioc6412 1R 0.04 fcr0990 1R 0.03 ncr3339 1R 0.04 fcr3053 1R 0.03 mioc3716 1R 0.04 fcrb55641R 0.03 mioc7668 1R 0.04 fcrb88701R 0.03 fcr4272 1R 0.04 fcrc22801R 0.03 fcr2306 1R 0.04 fcrc54581R 0.03 fcrc5482 1R 0.04 hfcr22751R 0.03 seoc5228 1R 0.04 hfcr39621R 0.03 miod3205 1R 0.04 miob82491R 0.03 seob9368 1R 0.04 mioc46031R 0.03 fcrc2050 1R 0.04 mioc63911R 0.03 fcrc4157 1R 0.04 ncrc09641R 0.03 mioc0301 1R 0.04 ncrc58061R 0.03 ncrcl326 1R 0.04 ncrc59591R 0.03 seoa9363 1R 0.04 ncrc59721R 0.03 seob4804 1R 0.04 seoa40551R 0.03 fcrb2124 1R 0.04 seoa97351R 0.03 fcr3957 1R 0.04 fcrb22921R 0.03 fcr0253 1R 0.04 fcrb63631R 0.03 fcr1562 1R 0.04 ncr7151 1R 0.03 fcr5474 1R 0.04 ncrc66871R 0.03 fcrc5309 1R 0.04 miod04551R 0.03 hfcr6700 1R 0.04 ncrc66971R 0.03 mioa0601 1R 0.04 fcrc55161R 0.03 mioa8774 1R 0.04 mioa30181R 0.03 miob4368 1R 0.04 ncrc33431R 0.03 mioc3127 1R 0.04 fcr2167 1R 0.03 mioc6385 1R 0.04 fcrc21311R 0.03 mioc8635 1R 0.04 seob02191R 0.03 ncr8481 1R 0.04 seob60081R 0.03 ncr9487 1R 0.04 seob65581R 0.03 ncrb7167 1R 0.04 seob54581R 0.03 ncrc0856 1R 0.04 seoc15041R 0.03 ncrc5738 1R 0.04 mioa19711R 0.03 seoa4670 1R 0.04 mioc92061R 0.03 seob1574 1R 0.04 ncrc02621R 0.03 seob4105 1R 0.04 seob61391R 0.03 seob4498 1R 0.04 ncr9933 1R 0.03 seoc0513 1R 0.04 miod73261R 0.03 fcr4214 1R 0.04 miod23671R 0.03 fcrb6522 1R 0.04 ncrc2443 hit 0.04 seoc2030 1R 0.04 ncr6637 1R 0.04 seoa1644 1V 2.4e-07 fcrb5679 1R 0.04 fcrb4985 1V 7.42e-07 seob6028 1R 0.04 seob3307 1V 8.48e-06 seoa0238 1R 0.04 seob2775 1V 1.28e-05 fcrc0559 1R 0.04 fcrb5214 1V 2.5e-05 ncrc0803 1R 0.04 fcrb8489 1V 2.68e-05 mioc6925 1R 0.04 mioc6211 1V 3.64e-05 hfcr3224 1R 0.04 miod6234 1V 3.66e-05 ncrc0863 1R 0.04 hfcr3180 1V 4.19e-05 miob4760 1R 0.04 fcrb2256 1V 4.65e-05 seob5954 1R 0.04 ncrc9436 1V 5.3e-05 seoa5302 1R 0.04 fcrb4981 1V 5.58e-05 fcrb6301 1R 0.04 seoa9997 1V 6.35e-05 ncrc0922 1R 0.04 seoc0009 1V 6.35e-05 fcrc5007 1R 0.04 miob9805 1V 7.39e-05 fcrc1313 1R 0.04 seoa0099 1V 8.58e-05 mioc0240 1R 0.04 seob4197 1V 8.72e-05 miod7099 1R 0.04 fcr5836 1V 9.95e-05 ncrc0341 1R 0.04 fcrc5604 1V 1.06e-04 fcr0343 1R 0.04 mioa4229 1V 1.08e-04 182 1R 0.04 fcrb4781 1V 1.15081e-04 fcr?

_ 1R 0.04 fcrb2137 1V 1.21e-04 fcrb1604 fcrbl854 1R 0.04 seob2937 1V 1.23e-04 fcrc4005 1R 0.04 fcrb3966 1V 1.27e-04 hfcr3067 1R 0.04 seob2938 1V 1.52e-04 hfcr3070 1R 0.04 hfcr2287 1V 1.53205e-04 hfcr4170 1R 0.04 miod4564 1V 1.53205e-04 hfcr4275 1R 0.04 seoa1460 1V 1.53205e-04 oa2165 1R 0.04 ncrc5019 1V 1.72e-04 mi . 1R 0.04 hfcr5969 1V 1.76303e-04 mioa5902 mioa8452 1R 0.04 mioa2993 1V 2.02535e-04 mioa8594 1R 0.04 miod5505 1V 2.02535e-04 miob5119 1R 0.04 fcrb5123 1V 2.14e-04 miod4407 1R 0.04 hfcr2844 1V 2.22e-04 iod7367 1R 0.04 miod7099 1V 2.26e-04 m 1R 04 fcr5339 1V 2.32274e-04 ncr4612 . miob9652 1V 2.32274e-04 ncr5975 1R 0.04 seoa1883 1V 2.32274e-04 ncr8725 1R 0.04 seoa5652 1V 2.32274e-04 ncrb0090 1R 0.04 seob5100 1V 2.32274e-04 ncrb3011 1R 0.04 ncrc5150 1V 2.45e-04 ncrb7376 1R 0.04 seob6217 1V 2.45e-04 ncrc1653 1R 0.04 seob7500 1V 2.45e-04 ncrc6129 1R 0.04 mioc0121 1V 2.45493e-04 ncrc7171 1R 0.04 seob4306 1V 2.49e-04 seoa0064 1R 0.04 seob2950 1V 2.65e-04 seoa6238 1R 0.04 ncr2892 1V 2.65933e-04 seoa8266 1R 0.04 ncrb3980 1V 2.65933e-04 seob8854 1R 0.04 seoa6178 1V 2.65933e-04 seoc1539 1R 0.04 seoc3519 1V 2.65933e-04 seoc1764 1R 0.04 fcrc5160 1V 3.03966e-04 seoc7498 1R 0.04 mioa3940 1V 3.03966e-04 seob3485 1R 0.04 seob7765 1V 3.03966e-04 ncrb8468 1R 0.04 fcr3575 1V 3.19001e-04 fcrb6005 1R 0.04 miod5030 1V 3.46873e-04 ncrb8539 1R 0.04 ncrc5016 1R 0.04 fcrb3544 1V 3.47e-04 ncr2288 1R 0.04 seoc1235 1V 3.49e-04 fcrb1920 1R 0.04 seoa5156 1V 3.5e-04 fcrb8020 1R 0.04 ncrb0164 1V 3.54e-04 seoa4132 1R 0.04 ncrc5025 1V 3.55e-04 seoa9870 1R 0.04 seoa6661 1V 3.66e-04 seob0876 1R 0.04 ncrc4985 1V 3.68e-04 seob2990 1R 0.04 seob7622 1V 3.68e-04 fcr2860 1V 3.7e-04 fcrc7219 1V 5.78978e-04 fcr4128 1V 3.7e-04 mioc2872 1V 5.78978e-04 fcrbl5391V 3.7e-04 nerc2857 1V 5.78978e-04 fcrb19901V 3.7e-04 seob9851 1V 5.78978e-04 fcrb23301V 3.7e-04 mioa8380 1V 5.8e-04 fcrb31191V 3.7e-04 hfcr0229 1V 6.04e-04 fcrb52691V 3.7e-04 fcrb5351 1V 6.14e-04 fcrb55361V 3.7e-04 seob9614 1V 6.2227e-04 fcrb55641V 3.7e-04 fcrb5896 1V 6.33e-04 fcrb62511V 3.7e-04 fcre2724 1V 6.33e-04 fcrb65961V 3.7e-04 mioc7370 1V 6.33e-04 fcrb66391V 3.7e-04 fcr2299 1V 6.35519e-04 fcrb70841V 3.7e-04 fcr3053 1V 6.55553e-04 ferb76931V 3.7e-04 fcrc5137 1V 6.55553e-04 fcrc27451V 3.7e-04 fcrc5614 1V 6.55553e-04 fcrc28491V 3.7e-04 mioa3668 1V 6.55553e-04 fcrc54801V 3.7e-04 miod4493 1V 6.55553e-04 miob15611V 3.7e-04 seoa5898 1V 6.55553e-04 miob34561V 3.7e-04 seoa8993 1V 6.55553e-04 miob88161V 3.7e-04 seoa9160 1V 6.55553e-04 miob92481V 3.7e-04 seob2987 1V 6.55553e-04 miob93931V 3.7e-04 ncrc3460 1V 7.23e-04 mioc40221V 3.7e-04 seob3088 1V 7.31e-04 miod56221V 3.7e-04 fcrb9611 1V 7.41141e-04 ncr1545 1V 3.7e-04 fcrc5107 1V 7.41141e-04 ncrb22881V 3.7e-04 mioc3682 1V 7.41141e-04 ncrb74821V 3.7e-04 ncrb0749 1V 7.41141e-04 ncrc24431V 3.7e-04 seoc3690 1V 7.41141e-04 ncrc53631V 3.7e-04 seoc1642 1V 7.46e-04 seoa22091V 3.7e-04 fcrb8340 1V 7.85e-04 seoa32301V 3.7e-04 miob8825 1V 7.85e-04 seoa734U1V 3.7e-04 seob9869 1V 8.13e-04 seoa89971V 3.7e-04 mioc4161 lV 8.18e-04 seob03031V 3.7e-04 fcrb8668 1V 8.3666e-04 seob07631V 3.7e-04 fcrc5041 1V 8.3666e-04 seob08721V 3.7e-04 fcrc5086 1V 8.3666e-04 seob14261V 3.7e-04 fcrc6932 1V 8.3666e-04 seoc16611V 3.7e-04 hfcr3058 1V 8.3666e-04 seoc22211V 3.7e-04 seob6437 1V 8.3666e-04 mioa47531V 3.89e-04 seob9772 1V 8.3666e-04 fcrb81141V 3.91e-04 seoa1318 1V 8.39e-04 seoal7491V 3.95201e-04 fcrb9420 1V 8.44e-04 miob91241V 4.2e-04 fcrb6715 1V 8.66e-04 fcrc04301V 4.28e-04 ncr2575 1V 9.04e-04 seoa59771V 4.43e-04 ncrc4875 1V 9.26e-04 fcr1772 1V 4.49e-04 miob9285 1V 9.36e-04 fcrc55771V 4.49548e-04 fcrb1684 1V 9.43108e-04 miob77941V 4.49548e-04 fcrb2113 1V 9.43108e-04 ncr4030 1V 4.49548e-04 fcrb3686 1V 9.43108e-04 ncrc05761V 4.49548e-04 mioa8864 1V 9.43108e-04 seoa04291V 4.68e-04 miob3695 1V 9.43108e-04 fcrb36801V 4.71e-04 miob4574 1V 9.43108e-04 ncr0133 1V 4.83e-04 miob8609 1V 9.43108e-04 fcrb67851V 5.10569e-04 seoa0799 1V 9.43108e-04 miob07641V 5.10569e-04 seob6379 1V 9.43108e-04 miod38271V 5.10569e-04 seoc4380 1V 9.43108e-04 ncrb00461V 5.10569e-04 seob6380 1V 9.5e-04 seob13621V 5.10569e-04 seob2994 1V 9.76e-04 seoc01491V 5.10569e-04 ncrc5039 1V 9.97e-04 fcrc41801V 5.27e-04 mioc1354 1V 1.061561e-03 fcrb37631V 5.67e-04 ncrb1167 1V 1.073748e-03 ncrc32831V 5.67e-04 seoa3895 1V 1.119756e-03 ncrc33581V 5.67e-04 seoc1307 1V 1.151704e-03 fcr4328 1V 5.78978e-04 seoa1737 1V 1.155164e-03 ~F1V~- ncrc3598 1V 1.476515e-03 ~~~~~~1.170347e-03 ncrc9899 170828e-03 ncrc4076 1V 1.476515e-03 fcrb3734 1V 1. ncrc5688 1V 1.476515e-03 seob0133 1V 1.181737e-03 rc8988 1V 1.476515e-03 miob3618 1V 1.193188e-03 nc 1V 1.476515e-03 oa1856 miob7106 1V 1.193188e-03 se lV 1.476515e-03 oa2768 miob7638 1V 1.193188e-03 se 1V 476515e-03 ncrb0328 1V 1.193188e-03 seoa3891 .
476515e-03 ncrc0461 1V 1.193188e-03 seoa4681 1V .
V 476515e-03 seob4555 1V 1.193188e-03 seoa5235 1 .
476515e-03 seoc0034 1V 1.193188e-03 seoa6557 1V .
476515e-03 mioa0192 1V 1.218739e-03 seoa7517 1V .
476515e-03 miod4735 1V 1.252641e-03 seoa7530 1V .
476515e-03 mioa1906 1V 1.255827e-03 seoa8424 1V .

ioa8973 1V 1.256375e-03 seoa9627 1V 1.476515e m 1V 27253e-03 seoa9792 1V 1.476515e-03 seoa9724 . seob0370 1V 1.476515e-03 miod1316 1V 1.29623e- eob1947 1V 1.476515e-03 ncrc9159 1V 1.326356e-03 s 1V 476515e-03 fcr3001 1V 1.339245e-03 seob5556 .
476515e-03 fcrb8908 1V 1.339245e-03 seoc2722 1V .
493237e-03 fcrc0591 1V 1.339245e-03 mioa8987 1V 1.
494352e-03 fcrc2775 1V 1.339245e-03 miob7319 1V .
501088e-03 fcrc7056 1V 1.339245e-03 mioa0311 1V .
501088e-03 hfcr2629 1V 1.339245e-03 mioa2173 1V .
V 501088e-03 miob2836 1V 1.339245e-03 ncr4135 1 .

ioc0669 1V 1.339245e-03 ncrb2544 1V 1.501088e m 1V 1.339245e-03 ncrb4166 1V 1.501088e-03 iod3946 m 1V 339245e-03 seoa0536 1V 1.501088e-03 ncrc0964 . seob8562 1V 1.501088e-03 339245e-03 ncrc9052 1V . hfcr2670 1V 1.548822e-03 339245e-03 seoa6172 1V . seoa5554 1V 1.560016e-03 339245e-03 seob4734 1V 1. miob9748 1V 1.637446e-03 339245e-03 seob7424 1V 1. hfcr0263 1V 1.680177e-03 411242e-03 fcr4380 1V 1. hfcr5991 1V 1.680177e-03 3142e-03 miod3914 1V . hfcr6613 1V 1.680177e-03 .
426601e-03 miod7081 1V . mioa3913 1V 1.680177e-03 miod1323 1V e ncrb3768 1V 1.680177e-03 1.44079 fcrbl922 1V 1.458522e-03 fcr0748 1V 1.690349e-03 ncr4189 1V 1.458522e-03 rb2292 1V 1.706646e-03 f fcr1756 1V 1.476515e-03 c 1V 1.722062e-03 rc6990 f fcr4494 1V 1.476515e-03 c 1V 1.79599e-03 eob3226 fcr4902 1V 1.476515e-03 s 1V 1.80089e-03 eob2728 fcrbl575 1V 1.476515e-03 s 1V 1.835381e-03 hfcr3444 fcrb2483 1V 1.476515e-03 r2102 1V 1.844189e-03 f fcrb4470 1V 1.476515e-03 c 1V 1.878078e-03 r3856 f fcrb4717 1V 1.476515e-03 c 1V 1.878078e-03 ioa8952 fcrb7036 1V 1.476515e-03 m 1V 1.878078e-03 iob8487 fcrb9454 1V 1.476515e-03 m 1V 1.878078e-03 iod5060 fcrc0529 1V 1.476515e-03 m 1V 1.878078e-03 miod5123 fcrc5471 1V 1.476515e-03 r6344 1V 1.878078e-03 n fcrc7222 1V 1.476515e-03 c 1V 1.878078e-03 oa0501 hfcr3647 1V 1.476515e-03 se 1V 1.878078e-03 eoa8239 mioa1626 1V 1.476515e-03 s 1V 1.878078e-03 oc5228 miob3307 1V 1.476515e-03 se 1V 1.898609e-03 fcrb2800 miob9734 1V 1.476515e-03 r0634 1V 1.92683e-03 miob9788 1V 1.476515e-03 nc 1V 1.961962e-03 oa4795 mioc0911 1V 1.476515e-03 se 1V 1.97218e-03 ob7906 mioc7364 1V 1.476515e-03 se 1V 2.019211e-03 rc5482 f miod1448 1V 1.476515e-03 c 1V 2.026769e-03 rc5230 miod5092 1V 1.476515e-03 nc 1V 2.092754e-03 eoa2805 miod7212 1V 1.476515e-03 s 1V 2.096472e-03 r2276 f ncr0025 1V 1.476515e-03 c 1V 2.096472e-03 fcrb4345 ncr2967 1V 1.476515e-03 rc1563 1V 2.096472e-03 f ncr8538 1V 1.476515e-03 c 1V 096472e-03 ncrb3638 1V 1.476515e-03 fcrc5290 .
096472e-03 ncrc1780 1V 1.476515e-03 fcrc6002 1V .

... ~~.a .a, ~97~~~~~ 1V~~ .~... ncr1712 1V 2.893252e-03 0 ~~~~~"0964'72e-03 o 2~

. ncrc2161 1V 2.893252e-03 mi 096472e-03 a 2 miob7662 1V . seoa2391 1V 2.893252e-03 096472e-03 miob8320 1V . seob4145 1V 2.893252e-03 096472e-03 miod5010 1V . seob9543 1V 2.893252e-03 096472e-03 miod5256 1V . ncr0491 1V 2.89732e-03 096472e-03 ncr3825 1V . seoc1628 1V 3.049387e-03 096472e-03 ncrb6109 1V . seob6156 1V 3.053926e-03 096472e-03 seoa0207 1V . ncrc4033 1V 3.104345e-03 134438e-03 ncrb2092 1V . seob8287 1V 3.182459e-03 209983e-03 fcrb1992 1V . mioc0302 1V 3.183214e-03 225021e-03 seoa3761 1V . mioc0621 1V 3.183214e-03 228163e-03 miod4407 1V . ncr2930 1V 3.183214e-03 23127e-03 seoa5433 1V . mioc3369 1V 3.203555e-03 299089e-03 seoa6620 1V . ncrc9877 1V 3.209695e-03 337161e-03 fcrb3615 1V . fcr1555 1V 3.212904e-03 337161e-03 fcrb3870 1V . fcr4129 1V 3.212904e-03 337161e-03 fcrb5296 1V . fcrb5467 1V 3.212904e-03 337161e-03 fcrb6508 1V . fcrb8236 1V 3.212904e-03 337161e-03 fcrb8119 1V . mioa7617 1V 3.212904e-03 337161e-03 fcrb8215 1V . miod4269 1V 3.212904e-03 337161e-03 fcrc5139 1V . ncr4656 1V 3.212904e-03 337161e-03 miod5198 1V . ncr7923 1V 3.212904e-03 337161e-03 miod6560 1V . ncr8153 1V 3.212904e-03 337161e-03 ncr3782 1V . seoal599 1V 3.212904e-03 337161e-03 ncrb0027 1V . seoa8912 1V 3.212904e-03 337161e-03 ncrb4477 1V . mioc1060 1V 3.280556e-03 337161e-03 ncrb7166 1V . ncrc2675 1V 3.327646e-03 337161e-03 seoa1173 1V . ~-~cr8725 1V 3.386385e-03 337161e-03 seob18U8 1V . fcrb2713 1V 3.387122e-03 342651e-03 seob6206 1V . fcrb6031 1V 3.409264e-03 3779e-03 miob9087 1V . hfcr2536 1V 3.539273e-03 379735e-03 seob0321 1V . mioa9491 1V 3.555572e-03 406345e-03 miob3594 1.V . fcr2940 1V 3.558947e-03 48884e-03 fcrb4067 1V . fcr5536 1V 3.563359e-03 540573e-03 seob0937 1V . fcrb1731 1V 3.563359e-03 541667e-03 mioc3220 1V . mioc2166 1V 3.563359e-03 564635e-03 fcrb9684 1V . ncrb0045 1V 3.563359e-03 564635e-03 seob7575 1V . ncrb0074 1V 3.563359e-03 602068e-03 fcr0061 1V . ncrb2517 1V 3.563359e-03 602068e-03 fcr_3282 1V . ncrc6846 1V 3.563359e-03 1V 602068e-03 fcrc0379 1V . seoa1992 1V 3.563359e-03 602068e-03 fcrc4390 1V . seoa6152 1V 3.563359e-03 602068e-03 miob5940 1V . seoa8177 1V 3.563359e-03 602068e-03 mioc3603 . seob0058 1V 3.563359e-03 602068e-03 miod1377 1V . seob6133 1V 3.563359e-03 602068e-03 miod5771 1V . seoc0951 1V 3.563359e-03 602068e-03 miod6324 1V . seoc3659 1V 3.563359e-03 602068e-03 ncrc8863 1V . seoc4928 1V 3.563359e-03 602068e-03 seoa1598 1V . ncr4454 1V 3.575847e-03 602068e-03 seob4303 1V . seob3882 1V 3.664302e-03 604001e-03 fcrb6929 1V . seob0085 1V 3.701885e-03 604001e-03 ncrc2959 1V . mioa6726 1V 3.711483e-03 697142e-03 fcrbl337 1V . ncrb8646 1V 3.711654e-03 697142e-03 fcrc5690 1V . seoa3717 1V 3.749508e-03 697142e-03 ncr0165 1V . seoa8399 1V 3.841578e-03 721704e-03 seoa6131 1V . fcrb5375 1V 3.881761e-03 721914e-03 fcr4503 1V . miob9533 1V 3.895277e-03 870616e-03 miod4857 1V . ncrb6742 1V 3.895528e-03 893252e-03 fcr1328 1V . seoc5006 1V 3.929979e-03 893252e-03 fcrc2573 1V . fcrb1876 1V 3.9471e-03 893252e-03 fcre5671 1V . fcrc2954 1V 3.9471e-03 893252e-03 hfcr5611 1V . mioa2204 1V 3.9471e-03 893252e-03 mioc3669 1V . mioa8970 1V 3.9471e-03 miod0080 1V e 2.8 mioa960.4-~~.~ ~3.9.471e~-03 ncrc2831 1V 4.204851e-03 miob86911V 3.9471e-03 ncrc2919 1V 4.204851e-03 mioc72011V 3.9471e-03 ncrc4907 1V 4.204851e-03 miod53101V 3.9471e-03 ncrc5312 1V 4.204851e-03 miod68451V 3.9471e-03 ncrc8841 1V 4.204851e-03 ncr8481 1V 3.9471e-03 ncrc8949 1V 4.204851e-03 ncrb82391V 3.9471e-03 ncrc9328 1V 4.204851e-03 ncrc38551V 3.9471e-03 ncrc9910 1V 4.204851e-03 ncrc53761V 3.9471e-03 seoa0388 1V 4.204851e-03 ncrc65871V 3.9471e-03 seoa3147 1V 4.204851e-03 seoa02561V 3.9471e-03 seoa6137 1V 4.204851e-03 seob72781V 3.9471e-03 seoa6393 1V 4.204851e-03 miob87541V 3.989036e-03 seoa6497 lV 4.204851e-03 seob75841V 4.012618e-03 seoa6598 1V 4.204851e-03 ncrb81711V 4.135195e-03 seoa6718 1V 4.204851e-03 mioc14161V 4.135238e-03 seoa7157 1V 4.204851e-03 seob00461V 4.135238e-03 seoa8543 1V 4.204851e-03 miob92091V 4.157852e-03 seoa9302 1V 4.204851e-03 fcrb23081V 4.204851e-03 seob0034 1V 4.204851e-03 fcrb31811V 4.204851e-03 seob0154 1V 4.204851e-03 fcrb32011V 4.204851e-03 seob0426 1V 4.204851e-03 fcrb42411V 4.204851e-03 seob1081 1V 4.204851e-03 fcrb42701V 4.204851e-03 seob4645 1V 4.204851e-03 fcrb42941V 4.204851e-03 seob5319 1V 4.204851e-03 fcrb43601V 4.204851e-03 seob5743 1V 4.204851e-03 fcrb44171V 4.204851e-03 seob5748 1V 4.204851e-03 fcrb49631V 4.204851e-03 seob5899 1V 4.204851e-03 fcrb56881V 4.204851e-03 seob7474 1V 4.204851e-03 fcrb84491V 4.204851e-03 seob8483 1V 4.204851e-03 fcrb93711V 4.204851e-03 seoc0843 1V 4.204851e-03 fcrcl7581V 4.204851e-03 fcr0903 1V 4.316218e-03 fcrc23061V 4.204851e-03 fcr1499 1V 4.339479e-03 fcrc55831V 4.204851e-03 hfcr3143 1V 4.349057e-03 hfcr30221V 4.20485ie-03 mioc3139 1V 4.358139e-03 mioa15201V 4.204851e-03 fcrb3627 1V 4.366761e-03 mioa33211V 4.204851e-03 fcrb5503 1V 4.366761e-03 mioa36731V 4.204851e-03 fcrc1402 1V 4.366761e-03 mioa40761V 4.204851e-03 fcrc5771 1V 4.366761e-03 mioa45641V 4.204851e-03 hfcr2895 1V 4.366761e-03 mioa50851V 4.204851e-03 miob2668 1V 4.366761e-03 mioa56921V 4.204851e-03 miod5672 1V 4.366761e-03 mioa59551V 4.204851e-03 ncr4416 1V 4.366761e-03 miob39531V 4.204851e-03 ncrb0696 1V 4.366761e-03 mioc12291V 4.204851e-03 ncrb8585 1V 4.366761e-03 mioc22091V 4.204851e-03 seoa1616 1V 4.366761e-03 mioc23481V 4.204851e-03 seob4293 1V 4.366761e-03 mioc24511V 4.204851e-03 seoc0284 1V 4.366761e-03 mioc25771V 4.204851e-03 seoc0742 1V 4.366761e-03 mioc41121V 4.204851e-03 seoc4381 1V 4.366761e-03 mioc63741V 4.204851e-03 fcrc5007 1V 4.388153e-03 miod21281V 4.204851e-03 seob0185 1V 4.388153e-03 miod31601V 4.204851e-03 miod7414 1V 4.474241e-03 miod60481V 4.204851e-03 seoc2549 1V 4.489551e-03 miod65211V 4.204851e-03 mioc4351 1V 4.538604e-03 miod73241V 4.204851e-03 mioc8481 1V 4.678627e-03 ncr0638 1V 4.204851e-03 miod4686 1V 4.740394e-03 ncr1305 1V 4.204851e-03 fcr2972 1V 4.825136e-03 ncr1550 1V 4.204851e-03 fcrc4551 1V 4.825136e-03 ncrb02071V 4.204851e-03 fcrc6174 1V 4.825136e-03 ncrb33481V 4.204851e-03 hfcr4497 1V 4.825136e-03 ncrb85381V 4.204851e-03 miob2285 1V 4.825136e-03 ncrc05391V 4.204851e-03 mioc1600 1V 4.825136e-03 ncrc12471V 4.204851e-03 miod4932 1V 4.825136e-03 ncrc15951V 4.204851e-03 ncr8096 1V 4.825136e-03 n~crc97841V 4 . 025136e-03 fcr2935 1V 5. 990508e-03 ~~~~~
rv~~~

seoa4107 1V 4.825136e-03 seob6851 1V 6.032512e-03 seob4105 1V 4.825136e-03 seob7309 lV 6.107204e-03 seob7866 1V 4.825136e-03 fcrb7760 1V 6.130086e-03 seoa2899 1V 4.826733e-03 ncrc5088 1V 6.197231e-03 hfcr0489 1V 4.869783e-03 seob2661 1V 6.211987e-03 fcrb5527 1V 4.888032e-03 mioc2021 1V 6.236422e-03 seob5551 1V 4.908034e-03 seoc1804 1V 6.261128e-03 mioa4077 1V 4.909178e-03 miob8463 1V 6.391219e-03 ncrc9228 1V 4.93663e-03 fcrc0959 1V 6.416508e-03 fcrb3120 1V 5.081016e-03 ncr2484 1V 6.42957e-03 ncre4448 1V 5.161235e-03 fcrb6301 1V 6.441057e-03 hfcr5865 1V 5.255535e-03 fcr1347 1V 6.462965e-03 fcrb3715 1V 5.312325e-03 fcrb5241 1V 6.462965e-03 fcrb7852 1V 5.318621e-03 fcrc5391 1V 6.462965e-03 ncrb7027 1V 5.324125e-03 mioa0497 1V 6.462965e-03 fcr1312 1V 5.325183e-03 miob3531 1V 6.462965e-03 fcrb8542 1V 5.325183e-03 miob6713 1V 6.462965e-03 fcrc6465 1V 5.325183e-03 miod0456 1V 6.462965e-03 fcrc7046 1V 5.325183e-03 ncr2869 1V 6.462965e-03 mioa1473 1V 5.325183e-03 ncr7668 1V 6.462965e-03 mioa9891 1V 5.325183e-03 ncrc0249 1V 6.462965e-03 miob0167 1V 5.325183e-03 ncrc3596 1V 6.462965e-03 miob3247 1V 5.325183e-03 ncrc4654 1V 6.462965e-03 miob5751 1V 5.325183e-03 ncrc6417 1V 6.462965e-03 miod5258 1V 5.325183e-03 seoa1747 1V 6.462965e-03 ncrc6588 1V 5.325183e-03 seoa2822 1V 6.462965e-03 seoa4647 1V 5.325183e-03 seoa6754 1V 6.462965e-03 seob3064 1V 5.325183e-03 seoa9363 1V 6.462965e-03 seob6096 1V 5.325183e-03 seob0201 1V 6.462965e-03 seoc0369 1V 5.325183e-03 seob3170 1V 6.462965e-03 miob9163 1V 5.338918e-03 seob7729 1V 6.462965e-03 seoa6930 1V 5.338918e-03 seob9756 1V 6.462965e-03 seob2685 1V 5.358138e-03 ncrb7350 1V 6.507615e-03 miob0974 1V 5.376436e-03 seob9970 1V 6.507615e-03 fcrb6107 1V 5.429727e-03 seoa8300 1V 6.520751e-03 seob4689 1V 5.482935e-03 fcrc4663 1V 6.53971e-03 hfcr2686 1V 5.502936e-03 hfcr3404 1V 6.5438e-03 seoa9712 1V 5.541398e-03 fcrb6363 1V 6.573178e-03 mioc2403 1V 5.546828e-03 seob5889 1V 6.584665e-03 mioc2592 1V 5.591403e-03 fcrb1767 1V 6.5959e-03 seoa4066 1V 5.634661e-03 miob7109 1V 6.695564e-03 miob8274 1V 5.669235e-03 mioc2961 1V 6.726375e-03 5558 1V 5.842704e-03 ncrc3690 1V 6.760566e-03 i oa 1V 5.868299e-03 seob6272 1V 6.914972e-03 m ncr3614 fcr6748 1V 5.870026e-03 fcrb6776 1V 6.94408e-03 fcrb5449 1V 5.870026e-03 ncrb8802 1V 6.973937e-03 fcrc0180 1V 5.870026e-03 fcr5257 1V 7.107476e-03 mioa2038 1V 5.870026e-03 fcrb6279 1V 7.107476e-03 mioa6556 1V 5.870026e-03 fcrb7812 1V 7.107476e-03 miob2227 1V 5.870026e-03 mioa9246 1V 7.107476e-03 miob9678 1V 5.870026e-03 miob1814 1V 7.107476e-03 mioc5182 1V 5.870026e-03 miob4803 1V 7.107476e-03 miod4019 1V 5.870026e-03 miob6632 1V 7.107476e-03 ncr8111 1V 5.870026e-03 mioc0899 1V 7.107476e-03 ncrb0220 1V 5.870026e-03 mioc2619 1V 7.107476e-03 ~

seoa1834 1V 5.870026e-03 ncr1299 1V 3 7.107476e-0 seoa4264 1V 5.870026e-03 ncr4612 1V 7.107476e-03 seoa5138 1V 5.870026e-03 ncr7904 1V 7.107476e-03 seoc1764 1V 5.870026e-03 ncrb3424 1V 7.107476e-03 seoc3374 1V 5.870026e-03 ncrc2796 1V 7.107476e-03 fcrb5254 1V 5.873716e-03 ncrc3551 1V 7.107476e-03 hfcr3844 1V 5.898035e-03 ncrc4798 1V 7.107476e-03 fcr0999 1V 5.963463e-03 ncrc9044 1V 7.107476e-03 DEMANDE OU BREVET VOLUMINEUX
LA PRESENTE PARTIE DE CETTE DEMANDE OU CE BREVET COMPREND
PLUS D'UN TOME.

NOTE : Pour les tomes additionels, veuillez contacter 1e Bureau canadien des brevets JUMBO APPLICATIONS/PATENTS
THIS SECTION OF THE APPLICATION/PATENT CONTAINS MORE THAN ONE
VOLUME

NOTE: For additional volumes, please contact the Canadian Patent Office NOM DU FICHIER / FILE NAME
NOTE POUR LE TOME / VOLUME NOTE:

Claims

1. A method for screening molecular markers to identify classifiers, the method comprising:
a. obtaining, for members of a first training population, first molecular marker data reflective of the expression in blood of each of a plurality of molecular markers, wherein said first training population comprises a first trait subgroup and a second trait subgroup;
b. selecting a plurality of candidate molecular markers from among said plurality of molecular markers based on a determination of the ability of said first molecular marker data to discriminate between members of said first trait subgroup and members of said second trait subgroup;
c. obtaining, for members of a second training population, second molecular marker data reflective of the expression in blood of all or a portion of said plurality of candidate molecular markers, wherein said second training population comprises said first trait subgroup and said second trait subgroup;
d. generating a plurality of combinations of molecular markers from said candidate molecular markers;
e. generating a plurality of classifiers by applying a mathematical model to said second molecular marker data for each of said plurality of combinations of molecular markers; and f. selecting one or more classifiers from said plurality of classifiers based on a determination of the ability of each classifier of said plurality of classifiers to discriminate between members of said first trait subgroup and members of said second trait subgroup.

2. The method of claim 1, wherein a subset of said plurality of candidate molecular markers of (b) is selected based on a determination of the ability of molecular marker data of said candidate molecular markers to discriminate between members of said first trait subgroup and members of said second trait subgroup.

3. The method of any of claims 1 to 2, wherein said determination of the ability to discriminate is made on the basis of a measure of statistical significance.

4. The method of any of claims 1 to 2, wherein said determination of the ability to discriminate is made on the basis of differential fold change.

5. The method of claim 3, wherein said determination of the ability to discriminate is further made on the basis of differential fold change.

6. The method of any of claims 4 or 5, wherein said selected candidate molecular markers demonstrate a differential fold change of greater than 2Ø

7. The method of any of claims 4 or 5, wherein said selected candidate molecular markers demonstrate a differential fold change of greater than 3Ø

8. The method of claim 3, wherein said determination of statistical significance is a p value and said p value is set such that the number of selected candidate molecular markers is less than 100.

9. The method of claim 3, wherein said determination of statistical significance is a p value and said p value is set such that the number of selected candidate molecular markers is less than 50.

10. The method of claim 3, wherein said determination of statistical significance is a p value and said molecular markers are selected if the molecular marker data results in a p value of less than 0.05.

11. The method of claim 3, wherein said determination of statistical significance is a p value and said molecular markers are selected the molecular marker data results in a p value of less than 0.01.

12. The method of any of claims 1 or 2, wherein said determination of the ability to discriminate is made on the basis of a Wald-Wolfowitz runs test, a Mann-Whitney U test, a Kolmogorov-Smirnov two-sample test, a Significant Analysis of Microarrays technique, or Manduchis' algorithm for assigning confidence to differentially expressed genes.

13. The method of any one of claims 2 to 12, wherein said determination of the ability of each of said subset of candidate molecular markers to discriminate between said members of said first trait subgroup and said second trait subgroup is made using second molecular marker data.

14. The method of claim 1, wherein said first training population and said second training population have zero or more members in common.

15. The method of claim 1, wherein said plurality of combinations includes all possible combinations of said candidate molecular markers.

16. The method of claim 1, wherein said plurality of combinations includes all possible combinations of two of said candidate molecular markers.

17. The method of claim 1, wherein said plurality of combinations includes all possible combinations of three of said candidate molecular markers.

18. The method of claim 1, wherein said plurality of combinations includes all possible combinations of four of said candidate molecular markers.

19. The method of claim 1, wherein said selecting one or more classifiers from said plurality of classifiers comprises:
i obtaining for each member of a scoring population, third molecular marker data reflective of the expression in blood of molecular markers within said plurality of classifiers, wherein said scoring population comprises members of said first trait subgroup and said second trait subgroup;
ii assigning a score, for each classifier in said plurality of classifiers, based on an ability of the respective classifier to discriminate between members of said first trait subgroup and members of said second trait subgroup in said scoring population using said third data; and iii selecting one or more classifiers from among said plurality of classifiers based on the score assigned to the selected classifier.

20. The method of claim 19, wherein said scoring population and said first and second training populations have zero or more members in common and said third data corresponds with said first and second data accordingly.

21. The method of claim 19, wherein said selecting one or more classifiers based on score comprises:
i ranking each classifier in the plurality of classifiers on the basis of the score assigned to said classifier; and ii selecting the top 10 ranking classifiers.

22. The method of any of claims 19 and 21, wherein said score, for each respective classifier in said plurality of classifiers, is a receiver operator curve (ROC) score determined by an area under a receiver operator curve obtained by applying the respective classifier to said scoring population.

23. The method of claim 22, wherein said selecting based on score comprises selecting those classifiers in said plurality of classifiers that have an ROC
score greater than 0.5.

24. The method of claim 22, wherein said selecting based on score comprises selecting those classifiers in said plurality of classifiers that have an ROC
score greater than 0.65.

25. The method of claim 22, wherein said selecting based on score comprises selecting those classifiers in said plurality of classifiers that have an ROC
score greater than 0.80.

26. The method of claim 22, wherein said selecting based on score comprises selecting those classifiers in said plurality of classifiers that have an ROC
score greater than 0.90.

27. The method of any of claims 1 to 26, wherein said first and second molecular marker data is obtained from one or more available databases.

28. The method of claim 1, wherein said first molecular marker data is obtained using a data collection method that allows for the collection of expression data corresponding to molecular markers for a major portion of the genome.

29. The method of claim 1, wherein said first molecular marker data is obtained using a microarray.

30. The method of claim 1, wherein said second molecular marker data is obtained using quantitative RT-PCR.

31. The method of claim 1, wherein said mathematical model is one of a regression model, a neural network, a clustering model, principal component analysis, nearest neighbor classifier analysis, linear discrimination analysis, quadratic discriminant analysis, a support vector machine, a decision tree, a genetic algorithm, a projection pursuit, or weighted voting.

32. The method of claim 31, wherein said mathematical model is optimized using bagging, boosting, or the Random Subspace Method.

33. The method of claim 1, wherein the number of candidate molecular markers selected comprises less than 100 molecular markers.

34. The method of claim 1, wherein the number of candidate molecular markers selected comprises less than 50 molecular markers.

35. The method of claim 1, the method further comprising:
a. obtaining, for a test subject, fourth molecular marker data reflective of the expression in blood of candidate molecular markers of said one or more selected classifiers;
b. applying said one or more classifiers to said fourth molecular marker data to thereby classify said test subject into either said first trait subgroup or said second trait subgroup.

36. The method of claim 35, wherein said fourth molecular marker data is received over the Internet from a remote source.

37. A method for identifying classifiers for a trait, the method comprising:
a. obtaining, for members of a training population, molecular marker data reflective of the expression in blood of all or a portion of a plurality of candidate molecular markers, wherein said plurality of candidate molecular markers are those molecular markers in a Table selected from Tables 1A-7I; wherein said Table is selected based on said trait and said training population comprises at least a first trait subgroup and a second trait subgroup for said trait as disclosed in Table F.
b. generating a plurality of combinations of molecular markers from said candidate molecular markers;
c. generating a plurality of classifiers by applying a mathematical model to said molecular marker data for each of said plurality of combinations; and d. selecting one or more classifiers from said plurality of classifiers based on a determination of the ability of said one or more classifiers to discriminate between members of said first trait subgroup and members of said second trait subgroup.

38. The method of claim 37 wherein a subset of said plurality of candidate molecular markers of the Table selected in (a) is selected based on a determination of the ability of the molecular marker data of said subset of candidate molecular markers to discriminate between members of said first trait subgroup and members of said second trait subgroup.

39. The method of claim 38 wherein said molecular marker data is said second data.

40. The method of claim 37, wherein said plurality of combinations includes all possible combinations of molecular markers wherein said molecular markers are those identified in the selected Table.

41. The method of claim 37, wherein said plurality of combinations includes all possible combinations of pairs of molecular markers wherein said molecular markers are those identified in the selected Table.

42. The method of claim 37, wherein said plurality of combinations includes all possible combinations of three molecular markers wherein said molecular markers are those identified in the selected Table.

43. The method of claim 37, wherein said plurality of combinations includes all possible combinations of four molecular markers wherein said molecular markers are those identified in the selected Table.

44. The method of claim 38, wherein said plurality of combinations includes all possible combinations of said subset of molecular markers of said selected Table of molecular markers.

45. The method of claim 38, wherein said plurality of combinations includes all possible combinations of pairs molecular markers of said subset of molecular markers.

46. The method of claim 38, wherein said plurality of combinations includes all possible combinations of three molecular markers of said subset of molecular markers.

47. The method of claim 38, wherein said plurality of combinations includes all possible combinations of four molecular markers of said subset of molecular markers.

48. The method of any of claims 37 or 38, wherein said determination of the ability to discriminate is made on the basis of a measure of statistical significance.

49. The method of any of claims 37 or 38, wherein said determination of the ability to discriminate is made on the basis of differential fold change.

50. The method of claim 48, wherein said determination of the ability to discriminate is further made on the basis of differential fold change.

51. The method of any of claims 49 or 50, wherein said selected candidate molecular markers have molecular marker data which demonstrate a differential fold change of greater than 2Ø

52. The method of any of claims 49 or 50, wherein said selected candidate molecular markers have molecular marker data which demonstrate a differential fold change of greater than 3Ø

53. The method of claim 48, wherein said determination of statistical significance is a p value and said p value is set such that the number of selected candidate molecular markers is less than 100.

54. The method of claim 48, wherein said determination of statistical significance is a p value and said p value is set such that the number of selected candidate molecular markers is less than 50.

55. The method of claim 48, wherein said determination of statistical significance is a p value and said molecular markers are selected if they have a p value of less than 0.05.

56. The method of claim 48, wherein said determination of statistical significance is a p value and said molecular markers are selected if they have a p value of less than 0.01.

57. The method of any of claims 37 or 38, wherein said determination of the ability to discriminate is made on the basis of a Wald-Wolfowitz runs test, a Mann-Whitney U test, a Kolmogorov-Smirnov two-sample test, a Significant Analysis of Microarrays technique, or Manduchis' algorithm for assigning confidence to differentially expressed genes.

58. The method of any one of claims 38 to 57, wherein said determination of the ability of each of said subset of candidate molecular markers to discriminate between said members of said first trait subgroup and said second trait subgroup is made using said second data.

59. The method of claim 37, wherein said first training population and said second training population have zero or more members in common.

60. The method of claim 37, wherein said selecting one or more classifiers from said plurality of classifiers comprises:
i obtaining for each member of a scoring population, third data reflective of the expression in blood of molecular markers within said plurality of classifiers, wherein said scoring population comprises members of said first trait subgroup and said second trait subgroup;
ii assigning a score, for each classifier in said plurality of classifiers, based on an ability of the respective classifier to discriminate between members of said first trait subgroup and members of said second trait subgroup in said scoring population using said third data; and iii selecting one or more classifiers from among said plurality of classifiers based on the score assigned to the selected classifier.

61. The method of claim 60, wherein said scoring population and said first and second training populations have zero or more members in common and said third data corresponds with said first and second data accordingly.

62. The method of claim 60, wherein said selecting one or more classifiers based on score comprises:
i ranking each classifier in the plurality of classifiers on the basis of the score assigned to said classifier; and ii selecting the top 10 ranking classifiers.

63. The method of any of claims 60 and 62, wherein said score, for each respective classifier in said plurality of classifiers, is a receiver operator curve (ROC) score determined by an area under a receiver operator curve obtained by applying the respective classifier to said scoring population.

64. The method of claim 63, wherein said selecting based on score comprises selecting those classifiers in said plurality of classifiers that have an ROC
score greater than 0.5.

65. The method of claim 63, wherein said selecting based on score comprises selecting those classifiers in said plurality of classifiers that have an ROC
score greater than 0.65.

66. The method of claim 63, wherein said selecting based on score comprises selecting those classifiers in said plurality of classifiers that have an ROC
score greater than 0.80.

67. The method of claim 63, wherein said selecting based on score comprises selecting those classifiers in said plurality of classifiers that have an ROC
score greater than 0.90

68. The method of any of claims 37 to 67, wherein said molecular marker data is obtained from one or more available databases

69. The method of claim 37, wherein said molecular marker data is obtained using quantitative RT-PCR

70. The method of claim 37, wherein said mathematical model is one of a regression model, a neural network, a clustering model, principal component analysis, nearest neighbor classifier analysis, linear discrimination analysis, quadratic discriminant analysis, a support vector machine, a decision tree, a genetic algorithm, a projection pursuit, or weighted voting.

71. The method of claim 71, wherein said mathematical model is optimized using bagging, boosting, or the Random Subspace Method.

72. The method of claim 37, wherein the number of candidate molecular markers selected comprises less than 100 molecular markers.

73. The method of claim 37, wherein the number of candidate molecular markers selected comprises less than 50 molecular markers.

74. The method of claim 37, the method further comprising:
a. obtaining, for a test subject, second molecular marker data reflective of the expression in blood of candidate molecular markers of said one or more selected classifiers;
b. applying said one or more classifiers to said second molecular marker data to thereby classify said test subject into either said first trait subgroup or said second trait subgroup.

75. The method of claim75, wherein said second molecular marker data is received over the Internet from a remote source.

76. A system for analysing the blood of a test subject, the system comprising:
a. obtaining, for said test subject, data reflective of the expression in blood of each molecular marker related to a classifier generated according to the method of any of claims 1 or 37;
b. applying said classifier to said data to thereby classify said test subject into a first trait subgroup or a second trait subgroup.

77. A composition useful for diagnosing a trait of interest said composition comprising a plurality of isolated polynucleotides each of said plurality of isolated polynucleotides selectively hybridizing to a molecular marker product so as to permit said plurality of isolated polynucleotides to generate molecular marker data for a combination of molecular markers, wherein said combination of molecular markers are selected from a Table chosen from one of Tables 1A to 7I and wherein said combination of molecular markers are derived using the method of claim 63 and results in a ROC score of greater than 0.6.

78. The composition of claim 78 wherein said trait of interest is selected from those traits disclosed in Table F.

79. A system for analysing the blood of a test subject, the system comprising:
a. a biochemical device for obtaining, for a test subject, data reflective of the expression in blood of each molecular marker in a classifier derived according to the method of claim 1;
b. a computing device for applying said classifier to said data to thereby classify said test subject into either a first trait subgroup or a second trait subgroup;
and c. a display for indicating to a user the result of said classification.

80. A system for screening molecular markers to identify classifiers, the system comprising a processor and being characterized by:
a. means for obtaining, for members of a first training population, first data reflective of the expression in blood of each of a plurality of molecular markers, wherein said first training population comprises a first trait subgroup and a second trait subgroup;
b. means for selecting a plurality of candidate molecular markers from among said plurality of molecular markers based on a determination of the ability of said molecular markers to discriminate between members of said first trait subgroup and members of said second trait subgroup using said first data;
c. means for obtaining, for members of a second training population, second data reflective of the expression in blood of all or a portion of said plurality of candidate molecular markers, wherein said second training population comprises said first trait subgroup and said second trait subgroup d. means for generating a plurality of combinations of molecular markers from said candidate molecular markers;
e. means for generating a plurality of classifiers by applying a mathematical model to each of said plurality of combinations of molectular markers using said second data; and f. means for selecting one or more classifiers from said plurality of classifiers based on a determination of the ability of each classifier of said plurality of classifiers to discriminate between members of said first trait subgroup and members of said second trait subgroup.

81. A system for identifying classifiers for a trait, the system comprising a processor and being characterized by:
a. means for obtaining, for members of a training population, data reflective of the expression in blood of all or a portion of a plurality of candidate molecular markers, wherein said plurality of candidate molecular markers are those molecular markers in a Table selected from Tables 1A-7I; wherein said Table is selected based on said trait and said training population comprises at least a first trait subgroup and a second trait subgroup for said trait.
b. means for generating a plurality of combinations of molecular markers from said candidate molecular markers;
c. means for applying a mathematical model to each of said plurality of combinations, using said second data, to derive a plurality of classifiers;
d. means for selecting one or more classifiers from said plurality of classifiers based on a determination of the ability of said plurality of classifiers to discriminate between members of said first trait subgroup and members of said second trait subgroup.