CN101611313A - Mass spectrometry biomarker assay - Google Patents

Abstract

The invention provides the method for determining that one or more biomarkers exist in the sample, comprise step: (a) make sample carry out the retention time exponential sum respective quality that each detection signal was analyzed and write down to mass spectrum (MS); (b) make corresponding to the reference database of each quality of signals and be associated with the biomarker quality, forming each signal and with reference to the correlativity between the biomarker, and abandon those quality not with the relevant signal of reference biomarker quality; (c) store those quality with reference to the relevant signal of biomarker; (d) by utilize similarity measurement with in the MS spectrum of each signal and the database with reference to the MS Spectral matching of biomarker, determine each storage assembly and with reference to the correlativity between the biomarker, to define one group of positive coherent signal; (d) measure the intensity of each positive coherent signal, and utilize discriminant function that its absolute signal intensity or relative signal intensity are kept the score; (e) with threshold application in available from discriminant function score value, to determine existing or not existing of biomarker.

Description

Mass spectrometry biomarker assay

The present invention relates to mensuration to biomarker.Particularly, the invention describes the multiple assay that can pass through the existence of biomarker in the mass spectroscopy automatic screening sample.

The various biological mark is called biomarker, has been applied to medical science and toxicity state by biochemical and molecular biology and is identified and study.Biomarker can find in tissue and biological fluids that described blood is the most common biological fluids that is used for biomarker research.

Biomarker can have predictable energy, and thereby can be used for predicting or detecting the existence of particular disorder or disease, level, type or stage (existence or the level that comprise specified microorganisms or toxin), to the neurological susceptibility (comprising genetic predisposition) of particular disorder or disease, or to the reaction (comprising drug therapy) of particular treatment.It is believed that by improving the usefulness of research and research and development plan, biomarker will be in the important effect of the following play more and more of drug discovery and research and development.Biomarker can be used as diagnostic reagent, the monitoring thing of progression of disease, the monitoring thing of treatment and the prediction thing of clinical effectiveness.For example, multiple biomarker project is just attempting to identify the label of particular cancers and specific cardiovascular and immunological diseases.

Complete albumen can utilize based on the isolation technics of gel and liquid phase and measure in many ways.Used the two-dimensional gel electrophoresis of the albumen with dissolving, wherein albumen is separated according to electric charge and size.Albumen is separated, makes that the modification after isomeric forms and the translation separates.By the quantitative albumen of staining technique, wherein can utilize pre-staining and poststaining technology.Metabolic marker also allows the range of linearity to extend nearly 5 orders of magnitude, and the sensitivity in femto mole (femtomolar) scope is provided.Carry out Identification of Fusion Protein from the gel spot of excision.Digestible protein after chemical degradation and the modification.The peptide mixer that produces extracting from the gel sample that separates, and identify by mass spectroscopy then.

Multidimensional HPLC (high performance liquid chromatography) can be used as the good alternative method of protein isolate or peptide.Albumen or peptide mixer are continuously by producing more high-resolution chromatographic stationary phase or dimension.HPLC is flexibly for many experimental techniques, and can owing to their in separating interested specific protein or peptide class suitability and select various fixed and activity mutually each other and with the compatibility of detected downstream and evaluation mass spectrometry method.It is the best way that high performance liquid chromatography is separated for solute at present, and it also allows highly to reproduce the ground automation mechanized operation.The online structure of many devices separation platform of these types is applied in the proteomics research at large.

Mass spectroscopy (MS) also is the key element in a kind of proteomics field.In fact, MS is the main tool that is used to study and characterize purifying protein in this field.The interface of protein group and MS is connected, and shows hundreds of or thousands of albumen, produces by gel technique, wherein can obtain high resolving power on single gel.The researcher just successfully utilizes the ability of MS to replace originally giving the two-dimentional gel of protein group power.

The application of mass spectroscopy (MS) in protein and peptide expression analysis has produced remarkable technical progress to identify by liquid phase separation techniques and/or based on protein or peptide that the isolation technics of gel is separated with development.Mass spectral characteristi for protein and peptide has two kinds of main methods: substance assistant laser desorpted ionized (MALDI) and electrospray ionization (ESI).Utilize several different methods, MALDI and ESI ion source can make up with the mass spectrometry of flight time (TOF) or other types to determine the quality or the sequence of peptide.

In MALDI, peptide and matrix cocrystallization, and use laser pulse.This processing makes peptide evaporation and ionization.The molecular weight of charged peptide (quality) is determined in the TOF analyzer then.In this device, electric field has quickened charged molecule towards detecting device, and its difference (their flight time) that makes Ionized peptide arrive the duration of detecting device disclosed the molecular weight of peptide, and littler peptide arrives detecting device quickly.This method has produced the quality profile of peptide mixer-just, the profile of the molecular weight of peptide and quantity in the potpourri.This profile can be used for identifying known protein matter from protein sequence database then.

By with the ESI-MS interfacial preparation to liquid chromatography (LC/MS/MS), be introduced in the mass spectrometric ion gun from the wash-out peptide of LC-post.Very thin pin is applied voltage.Described then pin sprays into drop in the mass analyzer, and described therein droplet evaporation and peptide ion discharge corresponding to various state of charge and from the position of measuring sequence, and described state of charge is segmentation (fragmented).In LC/MS/MS, the researcher has utilized microscopic capillary LC device with the initial separation peptide.

Mass spectroscopy (MS) is a kind of valuable analytical technology, because it measures the inwardness of biomolecule highly delicately.Therefore MS can be used for measuring the molecule type (albumen, peptide, or any other biomolecule) of wide region and the sample type/biomaterial of wide region.Correct specimen preparation is known to produce and spectrally resolved and sensitivity are crucial for the MS signal.Therefore specimen preparation is crucial for total feasibility and the sensitivity analyzed.

Albumen is to be difficult to the biomacromolecule that separates by the liquid chromatography isolation technics because in column material, stationary phase, particle in disadvantageous substance transfer.Yet, can destroy making albumen become more junior unit (peptide or polypeptide) form by the peptide bond that will connect two adjacent amino acids.This can pass through proteinase, can interact and dissolve the albumen of the peptide bond on other albumen, enzyme cut and finish.Trypsase is the most frequently used proteinase, is used for the protein expression analysis and research.After the enzymatic degradation, the peptide compound mixture of generation will separate and fractionation by micellar electrokinetic capillary chromatography.All all peptides of digestible protein are unsegregated in this stage in the sample.The peptide that produces from corresponding protein will can not be separated into a unit the fractionation chromatography step, but the peptide that produces with every other albumen from sample separates.Split (resolved) and the wash-out peptide that separates from height capillaceous, the most normally based on electric charge and hydrophobicity and by fractionation.The peptide that separates is avoided possible pollution thus from the online introducing mass spectrometer of the chromatographic fraction of platform.Peptide carries out quality determination (m/z) then, to be captured in all peptides in window preset time.Next, for many peptide quality are selected in order-checking (MS/MS), based on they abundance (abundance) at given time window.This relies on electrospray ionization ion trap mass spectrometry system to be undertaken by new ion sampling interface.The interface utilize linear four utmost points as ion guide and ion trap to improve the performance of trap.Improved the load cycle of system in the linear four ion proofs of catching in extremely.Dipole excitation at linear four ions of catching in extremely is used to spray unwanted ion.

After nineteen ninety, successfully used instrument the first time, utilize the ion trap mass spectrometry method of electrospray ionization (ESI) to become the widely used instrument that is used for trace analysis.Electrospray is for example source of the mitigation of peptide and albumen of ionizable important analytes.The highly charged ion that produces among the ESI can be expanded the scope of mass analyzer.The trap mass spectrometer has the MS performance (MS n......) that good performance is for example connected flexibly.In this ioning method, activate precursor ion by in trap, quickening quality selectivity linear ion hydrazine under the unsettled condition in the fragment ions of some formation.After the time lag, the stability parameter that changes ion trap before is catching of unsettled fragment to allow.The result is the product ion spectrum that comes from the interior precursor ion that can distribute with change.Allow precursor ion to enter after the linear ion hydrazine, it is possible following the many milliseconds of the differentiation that can distribute in the precursor.Time-the fragmentation product ion spectrum of delay typically demonstrates the continuous fragments product of reduction, produces the spectrum of easier explanation.For time-the important several important experiment parameter of fragmentation delayed identified and come into question.These technology all are useful for little precursor ion and multi-charge peptide.

Tandem mass spectrometry (MS/MS) is the cores of great majority for the modern mass spectrum research of compound mixture.Fragmentation relate to by with the activation of the precursor ion of object gas collision, and can produce charged and neutral fragment.The character of fragment ions, and their intensity, indicated usually precursor ion structure and thereby can produce for the useful information of identifying unknown analyte, and be provided for the useful triage techniques of inhomogeneity analyte.Activation by multiple collision has prolonged soak time and has made higher store energy go in the precursor ion.Higher collision gas pressure has also hinted higher collision relaxation rate.

The combination of Protein Separation by the 2D gel electrophoresis and the analysis by mass spectroscopy be set up as be used for biomarker analysis in, the multisystem that can analyze several biomarkers is in the experimental phase at present.

It is relevant with the composite mode of biomarker that numerous disease has demonstrated, and it can be used to diagnose the illness or indicate the reaction of patient to drug therapy.These patterns often relate to several biomarkers, need the analysis of multiple while.Therefore, the system that need can test multiple biomarker simultaneously.Ideally, this system can be robotization.

Summary of the invention

The invention provides mensuration for biomarker in the sample, its be robotization and be accurate.This mensuration depends on mass spectroscopy with the identification of organism mark, and is referred to herein as mass spectrometry biomarker assay (MSBA).

The invention provides and be used for determining that one or more polypeptide biomarkers exist, preferably, the method of the existence in people's specimen, it can comprise inhuman specimen, it typically is limited in the biological fluids of certain volume, described biological fluids comprises naturally occurring albumen and is included in a certain amount of tissue, the peptide in the sample that blood or other obtain clinically.

This method preferably comprises following step:

(a) make sample carry out the corresponding quality of retention time exponential sum that each detection signal was analyzed and write down to mass spectrum (MS);

(b) make corresponding to each quality of signals and be associated with reference database, described reference database contains the biomarker quality from known disease or the biological standard group that changes, forming each specimen signal and from the correlativity between the biomarker of standard group biomarker in the reference database, and abandon those quality not with main data group in the relevant test signal of reference biomarker quality;

(c) store those quality and the relevant specimen signal of reference biomarker in the normal data;

(d) by utilize similarity measurement with in the MS spectrum of each signal and the database with reference to the MS Spectral matching of biomarker, determine each storage assembly and with reference to the correlativity between the biomarker, to define one group of positive coherent signal;

(d) measure positive relevant each and store test signal, and utilize discriminant function that its absolute signal intensity or relative signal intensity are kept the score;

(e) with threshold application in available from discriminant function score value, to determine existing or not existing of biomarker.

Preferably, method of the present invention is utilized the main data group (masterdata set) of specimen screening stage.

Advantageously, the quality and the sequence characteristic of this method filtration and garbled data group, described characteristic is based on the electric charge relevant with the protein sequence of certain evaluation in the main data group, quality, the unique nature of sequence spectrum.

Therefore,, the invention provides the method that is used for determining in the existence of one or more polypeptide biomarkers of sample, comprise step in first aspect:

(a) make sample carry out the quality that the retention time exponential sum correspondence of each detection signal was analyzed and write down to mass spectrum (MS);

(b) make corresponding to the reference database of each quality of signals and be associated with the biomarker quality, forming each signal and with reference to the correlativity between the biomarker, and abandon those quality not with the relevant signal of reference biomarker quality;

(c) store those quality with reference to the relevant signal of biomarker;

(d) by utilize similarity measurement with in the MS spectrum of each signal and the database with reference to the MS Spectral matching of biomarker, confirm each storage assembly and with reference to the correlativity between the biomarker, to define one group of positive coherent signal;

(d) measure the intensity of each positive coherent signal, and utilize discriminant function that its absolute signal intensity or relative signal intensity are kept the score;

(e) with threshold application in available from discriminant function score value, to determine existing or not existing of biomarker.

Method of the present invention allows hundreds of biomarker in user's while analytic sample.This method depends on the database of biomarker, and it has been shown as and disease association, and it has comprised the quality and the spectroscopic data of each biomarker, and allows described biomarker accurately to identify in given sample by MSBA software.Be present in the peptide in the sample and get rid of the sequence do not expect based on the retention time index by screening, it arrives the time correlation of MS detecting device with peptide, and surpassing 30,000 sequences can be analyzed in several minutes, and given biomarker height is put the evaluation of letter ground.This method is automatable, and is high-throughout, and can be by unskilled relatively technician's operation.

Sample can carry out MS not carrying out previous lock out operation and analyze.In said embodiment, preferably by utilizing the direct infusion of static nanometer electrospray principle, flow injection or example enrichment flow injection, and analytic sample.

Advantageously, before analyzing, MS handles sample, preferably sample separation composition before loading to them on the MS.For example, sample preparation comprises the sample separation by single-phase or heterogeneous high pressure lipuid chromatography (HPLC) (HPLC).

Preferably electrospray ionization of MS system self (ESI) MS, substance assistant laser desorpted ionized-flight time (MALDI-TOF) MS or surface-enhanced laser desorption ionization-flight time (SELDI-TOF) MS.

The method according to this invention advantageously robotization and under computer control, implement.By relatively identifying biomarker in the sample with the reference data of described biomarker; Preferably, the reference mass of multiple biomarker and MS spectroscopic data store on the computers.

For the reference MS spectrum of given biomarker preferably by cluster calculation (clustering calculation) and by averaged spectrum actual and the measurement data acquisition.

Method of the present invention can realize in two ways; Be marked with the reference that is provided for quantitative signal intensity in the utilization and do not utilize described standard.Thereby, in one embodiment, before analyzing, mark in one or more and add in the sample by MS.Preferably, interior mark is a mark.

In described enforcement of the present invention, compare by measurement biomarker signal intensity and with it and one or more known interior target signal intensities, and the absolute signal intensity of each biomarker signal is kept the score.

In substituting enforcement, mark is not handled sample in not adding.In said embodiment, by the ratio between the reference signal strength of individual organisms marking signal intensity and patient group among the measurement patient, and relative signal intensity is kept the score.

Biomarker can be described as " measuring and be evaluated as normal biological processes objectively, pathologic process, or a kind of feature of the indication of the pharmacological reaction that treatment is intervened ".Biomarker is to identify and measurable indicant with any of particular disorder or disease association, the correlativity that has some aspects of biomarker existence or level and described illness or disease under the described situation (comprises the existence of described illness or disease, the level of level or change, type, stage, neurological susceptibility, or to the reactivity of the medicine that is used for the treatment of described illness or disease).Correlativity can be qualitative, quantitatively, or qualitative and quantitative.Typically, biomarker is a compound, compound fragment or compound group.Described compound can be any compound that is found in or produces from biosome, comprises albumen (and peptide), nucleic acid and other compounds.

Sample can be any interested biological substance, but advantageously biological tissue and preferably for example blood or blood plasma of biological fluids.

Method of the present invention depends on the correlativity of the reference mass and the MS spectrum of observed MS signal and known organism mark.Reference data preferably is stored on the computer server, and it allows entire method to implement under computer control.

By relatively that signal is relevant with normative reference, for example utilize computer function as described herein.Preferably, signal is characterized by " positive " or " feminine gender " according to the threshold value that whether obtains similarity; Negative and signal that do not obtain the similarity threshold value is abandoned in the MSBA method, and those positive signals mate and cause diagnosis to the existence of biomarker described in the biological sample with biomarker simultaneously.

According to the enforcement that MSBA measures, the known reference standard in the biological sample is added in reference to, or compares by the referenced strength that calculates in organizing with the patient, and measure signal intensity.

With under the situation of standard implementation, by being present in the biomarker signal in the sample and adding the MSBA score that interior target ratio in the sample is calculated the biomarker signal to.All biomarkers in multiple assay will be analyzed in the same manner, cause final MSBA to get molecular group.

For in the MSBA method, set up absolute quantitatively, preferably utilize inner (calibrant) standard.Described standard for example is isotope-labeled, and it is accurate to make that mensuration reads out in protein sequence aspect height, and is favourable aspect absolute quantitation.The foundation of calibrating sequence will allow the measurement of absolute protein biology mark level in blood or any other clinical sample in the MSBA screening.

A kind of new method is provided, form by a plurality of Connection Steps that are used to detect, have multiple reading, wherein 2-100 with quantitative protein sequence biomarker, but be not limited to it, the expression of biomarker can in single MSBA reads, depict (mapped).The MSBA system builds on the liquid phase platform, and it can be handled the single line diagnosis and describe, or the multiple configuration flow process of the sample of parallel processing simultaneously, improves the capacity and the flux of system thus.The detecting pattern of MSBA method is to determine with ensuing quantitative by the accurate Quality Identification of mass spectroscopy and sequence.

This method can be applicable to the biological sample of any kind, and it maybe can be converted into liquid form for liquid form.The MSBA method also can be handled from the cell of any kind, or the sample of the method for biotechnology, wherein for example measures the dynamics profile with respect to the time.Along with the time sample is automatically introduced the MSBA platform, implement analysis by next with respect to the time.The whole analysis ability of MSBA diagnosis profile measurement is computer-controlled fully, comprises by weighing the quality signal evaluation of difference and final MSBA score diagnosis, sequential analysis, multiple quantitative.All intermediate steps in the MSBA that moves on this platform circulation are estimated by tailor-made algorithm, and described tailor-made algorithm is made accurate judgement, decision making from resulting from from each round-robin mass data that the MSBA of any given biological sample analyzes.

The MSBA method make identified particular peptide with and the evaluation of the variant modified of biological chemistry, its show as the entity of separation or be present in albumen and the compound mixture of peptide within.Every kind of peptide can pass through specific amino acid sequence and limit, and it can be by its accurate quality or the affine selectivity evaluation in conjunction with character of its unique immunity for given immunology reagent.

This method allows to identify the significant property of protein of statistics and its modified forms.And, the relative populations of every kind of biomarker in the possible measuring samples, even in not utilizing under the target situation.Perhaps, can carry out absolute quantitation respectively to every kind of biomarker in any given biological sample by mark in utilizing, wherein in mark (n=1-20) be the protein sequence of biomarker for example.Mark can be formulated the amino acid sequence into cold (cold) then in these, or is the isotope labeling amino acid sequence.Except possible mark exception, standard has and the same sequence of selecting of biomarker.

This method combines and produces specific processing, segregation, and separation and evaluation are present in several committed steps of the unique protein sequence in the biological material specimens.This method can be applicable to people's clinical sample.This method also can be applicable to come from non-human animal's sample.

We provide the rapid method of multistep of the character that is used to identify unique protein sequence, described character for example shows as the entity from the atomic mass unit of biological sample, it has been verified to have quantitative change in given multi-biological mark group, its size for example can change in the scope of 2-100 in given sample.

We also provide and have determined or confirm that biomarker in any particular organisms sample has the method that the multiple quantity of the biomarker group of protein sequence changes, and it is determined in advance in cell or any other type sample clinical.This quantity changes the final MSBA algorithm computation of passing through to produce the MSBA score, and it will become diagnostic and read.

And the significant similarity of statistics can be detected and be registered as the unique protein sequence character or the character of multiple peptide.Determine that the significant similarity of statistics relates to and utilize retrievable albumen of the public and geneseq database and be in particular the requirement of satisfying the MSBA method and the algorithm that develops.

The integration that is used for the method step of biomarker evaluation is favourable.The method of integrating depends on following principle: 1) high-quality biomedical clinical material, what 2) have an ensuing liquid phase separation reproduces and sample preparation at a high speed 3) the quantitatively accurate and qualitative biomarker and 4 of determining multiple group) successively control data produce with calculate with permission multiplexed protein sequence set in the algorithm that separates of biomarker.

The accompanying drawing summary

Fig. 1 has shown the synoptic diagram of MSBA principle.

Fig. 2 in more detail illustration the data processing method that relates among the MSBA.

Fig. 3 has shown the mass spectrum from blood sample, and described blood sample is from the consumptive.Identified the multiple biomarker in the sample.

Fig. 4 has shown the example of the biomarker note of making from multiple assay, show as MS spectrum, wherein passes through the MS/MS spectrum identification biomarker of the CVLFPYGGCQGNGNK biomarker of MSBA software and ensuing representative generation.

Fig. 5 has shown that described MSBA model has 11 kinds of biomarker signals for 19 patients' sample data as estimating the example of the predictability of MSBA model as described in the embodiment 2.10 kinds of examples and 9 kinds of contrasts of being used as blind sample have been utilized.The MSBA score of utilizing equation 5 to calculate for each experimenter.In this embodiment, the MSBA score equals 1 or be diagnosed as case (red circle) greater than 1 experimenter.Otherwise this experimenter is considered to contrast (blue circle).Prediction accuracy is 100%.

Fig. 6 shown as utilization as described in the embodiment 3 have for 96 patients' sample data 10 kinds of signals the MSBA model self distinguish the result.Each some represent a patient, and vertical axis representative distinguishes score (z), and it utilizes equation 6 calculating.If score＞0 then is interpreted as it case (red circle), otherwise this experimenter is considered to contrast (blue circle).Prediction accuracy is 83.3%.

Detailed Description Of The Invention

Biomarker

FDA is " measuring objectively and be evaluated as normal biological processes, pathologic process, or to a kind of feature of the indication of the pharmacological reaction of Results " for the definition of biomarker.

As used herein, term " biomarker " refers to can be used for the existence of monitoring of diseases or the polypeptide of progress, and is consistent with above-mentioned FDA definition.

Biomarker can be used as diagnostic reagent, the monitoring thing of PD, the monitoring thing for the treatment of and the prediction thing of clinical effectiveness. For example, the plan of multi-biological marker research is just attempting to identify the mark of particular cancers and specific cardiovascular and immunological diseases.

Some these disease related proteins can be accredited as new drug targets, and some can be used for the biomarker of PD. The new diagnosis that described biomarker can be used for improving the clinical development of new drug or is used for the research and development specified disease.

Disease related protein is known in the art, and has set up them as the purposes of the biomarker of disease. Described biomarker can be monitored by the inventive method. Yet, can identify new disease related protein. For example, can obtain by following method the monitoring of disease related protein. Protein sample is obtained from single patient or patient's group. These samples can be cells, tissue, or biological fluids, and it is processed with extraction and rich protein and/or peptide composition. Typically, this method need to be dispensed in the liquid phase, but also can comprise the albumen that is attached to solid matrix and/or the foundation of peptide components. Separation and analysis (proteomics, peptide group) have produced the protein expression finger-print by the quantitative and qualitative analysis measurement for disease and health volunteer afterwards. These finger-prints can be used as the difference individuality and/or set up and/or follow the trail of the evaluation thing of the uniqueness of certain nature or lysis. These prototype finger-prints are set up for every kind of independent sample/experimenter, and are recorded as numerical value in Computer Database. Then utilize this finger-print of biological information tool analysis, to identify and to select albumen or peptide, described albumen or peptide are present in the prototype finger-print and its expression can differentially exist in coming from healthy and disease experimenter sample or differentially not exist. Then these albumen/peptides have further been characterized and have been produced detailed profile, and it identifies the characteristic physical property of albumen or peptide. Single albumen/peptide or albumen/peptide group can be determined with certain nature or lysis significant correlation.

Mass spectrography

Mass spectrography is the system of selection for the analysis of albumen and peptide. Modern biomarker is found to study and has been used two kinds of main mass spectrography principle: MALDI-TOF (substance assistant laser desorpted ionized flight time) mass spectrography, wherein albumen is analyzed with crystal state, and ESI (electrospray ionization) mass spectrography, wherein albumen is analyzed with liquid condition. In addition, the surface of MALDI strengthens chip application, is called Protein-based tumor biomarker (SELDI), has been widely used in biomarker and has found in the research. For example see the people such as Petricoin, Lancet, 16,572-577,2002; The people such as Alexe, Proteomics.2004,4766-4783; With the people such as Liotta, Endocr Relat Cancer.2004 Dec; 11 (4): 585-7.

Surface-enhanced laser desorb/ionization (SELDI)-TOF-MS technology has been utilized and has been incorporated into the chromatographic surface of measuring target plate. Protein binding materials on the plate is then by the MALDI-MS Direct Analysis. It mainly is peptide and the albumen of low-molecular-weight scope that SELDI measures. This technology is applicable for main for medium abundant peptide and the albumen of not integrating out suitable the preceding purification scheme. The SELDI technology mainly causes a kind of pattern, from this pattern, can utilize the MALDI-TOF-TOF of peptide to identify order-checking.

Many-device separation platform has guaranteed to utilize electrospray ionization chromatography online, or utilizes the ionization principle for example substance assistant laser desorpted ionized off-line chromatographic, makes up high-resolution peptide separation method. For example see, Aebersold, R.﹠Goodlett, D.R.Chem.Rev. 2001,101,269-295; Mann waits the people, Annu.Rev.Biochem, 2001.70,437-473; Wolters waits people Anal.Chem.73,5683-5690 (2001); And Washburn, wait the people, Nat.Biotechnol.19,242-247 (2001).

Mass spectrography (MS) also is the key element in proteomics field. In fact, MS is that this area is used to study the main tool with profiling protein structure and sequence. See Aebersold, R.﹠ Mann, M.Mass spectrometry-based proteomics.Nature 422,198-207 (2003); Steen, H. and M.Mann (2004) .Nat Rev Mol Cell Biol 5 (9): 699-711; And Olsen, J.V. and M.Mann (2004) Proc Natl Acad Sci U S A 101 (37): 13417-22.

The researcher just successfully utilizes the effect of MS to replace giving at first the two-dimentional gel of protein group motive force. Utilize ESI and liquid chromatogram (LC)/MS/MS, a kind of voltage is applied to a kind of very thin pin, described pin comprises peptide mixer, produces peptide sequence, from LC-post wash-out. Then described pin is spurted into drop in the mass spectrometer, and therein drop evaporation and peptide ion are released. In LC/MS/MS, the researcher has utilized microcapillary LC device with the starting point isolated peptides.

Mass spectrography (MS) is a kind of valuable analytical technology, because it measures the inwardness of biomolecule highly delicately, and its quality. Therefore, MS can be used for measuring the molecule type (albumen, peptide, or any other biomolecule) of the wide region in the sample type/biomaterial of wide region.

Correct sample preparation can affect the generation of MS signal and spectral resolution and sensitivity. LC/LC is connected in high-resolution separation system for example one-dimensional high pressure lipuid chromatography (HPLC) (LC) with the multidimensional liquid chromatography) can directly be connected with mass spectrography. This connection allows to obtain rapid automatizedly and collect large data group, and it has represented the quantitative and sequence information in the mass spectrum that produces. The air gun proteomic techniques of this integration is called MudPIT (multidimensional Identification of Fusion Protein technology). See the people such as Eng, J Am Soc Mass Spectrom 1994,5:976-989; The people such as Link, Nat Biotechnol.1999 Jul; 17 (7): 676-82; The people such as Washburn, Nat Biotechnol.2001 Mar; 19 (3): 242-7; The people such as Lin, American Genomic/Proteomic Technology, 2,001 1 (1): 38-46; With the people such as Tabb, J.Proteome Res.2002 1:21-26.

In the air gun proteomics method, analyze by specific protein digestive ferment trypsase and in other-and the peptide of outer-peptase/protease-producing strain for example, rather than complete albumen. This fragmentation provides clear and definite advantage because even very large albumen, its Physical and chemical characteristics with change is for example very hydrophobic, or very alkaline albumen, can be analyzed. Described protide otherwise be very unmanageable. These albumen will produce the peptide mixer that size and number is enough to allow accurate albumen note and evaluation. Yet since produced several peptides from every kind of albumen separately, the complexity of mixture to be analyzed has improved. As a result, need sizable instrument time and computing capability for the air gun method. Yet the resource of protein expression information is widely, and full automatic middle generation is set, and side by side identifies in real time albumen.

For processing different patient's samples of the disease that has represented multiple degree, can use liquid chromatography chromatogram and mass spectrum that distinct methods adjustment and calibration produce. Can carry out this normalization by a kind of software approach, the total signal that results from thus whole experiment produces to be utilized and to produce with total signal of multiple analysis patient sample and compares. The mean value of all signals and community will be calibrated to allow difference quantitative.

In the second method, the peptide standard of scheduled volume is added in the sample. This interpolation will be before treatments of the sample and afterwards, before or after carry out. The standard of utilizing will be the actual biomarker sequence that synthesizes the tagging sequence, or not have tagging, and add sample (spiked with the sample).

Labelling technique is very general for the purposes of quantitative clinical albumen adjustment research in the proteomics field. Multiple in conjunction with chemistry by utilizing, multiple labelling technique is developed and is employed. The mark of the most generally using in the proteomics field is ICAT and ITRAQ mark. See the people such as Parker, Mol.Cell.Proteomics, 625-659,3,2004; The people such as Ross, Cell.Proteomics, 3,1153-1169,2004; With the people such as DeSouza, J.Proteome Res., 2005,4,377-386.

Sample separation

Single, or multidimensional HPLC (high performance liquid chromatography) will be as the method for preferential selection for separating of albumen or peptide. Albumen or peptide mixer are by continuous chromatographic stationary phases or dimension, and it produces higher resolution ratio. HPLC is applicatory for many experimental techniques, and can owing to they for the adaptability of separating interested specific protein or peptide class with select multiple fix and mutually mobile each other with downstream monitoring and the compatibility of identifying mass spectrometry method. HPLC by the clinical sample of proteolytic enzyme digest, has wherein produced corresponding enzyme product, peptide mixer for separating of. Sample preparation methods is applied to for example blood of protein sample, tissue, or the biological fluids of any other type. For example see the people such as Schulte, Expert Rev.Diagn., 5 (2), 2005,145-157; The people such as Chertov, Expert Rev.Diagn., 5 (2), 2005,139-145; The people such as Adkins, Mol.Cell.Proteomics, 1 (12), 2002 947-955; The people such as Pieper, Proteomics. 2003 Jul; 3 (7): 1345-64; And Anderson, N.L.﹠ Anderson, N.G.Mol.Cell Proteom.20021,845-867.

Corresponding peptide mixer is by continuous chromatographic stationary phases or dimension, and it has produced high-resolution. HPLC is flexibly for many experimental techniques; In of the present invention the setting, produced a kind of optimization of having eliminated especially the albumen of the high abundance component of in the human blood sample, expressing, thus medium, and protedogenous enrichment is produced in the abundance such as low zone. Separating of peptide and albumen is based on peptide sequence, the function group of peptide sequence, and physical property.

The MSBA-operating principle

Before sample is carried out MSBA, in most of the cases need sample treatment and preparation process. Introducing this step before the MSBA method is to disturb reagent and matrix components in order to eliminate, and promotes thus total detectability of improving, causes explaining and the increase of total sensitivity. Yet in certain embodiments, sample preparation can be exempted, and particularly is in higher abundance and sample complexity when very low when biomarker. Those skilled in the art can determine that preparation process is whether crucial.

The MSBA platform can multitude of different ways operate, and mainly determines by character and its complexity of sample.

Determine the biomarker protein sequence by multiple analysis quantitative and qualitative analysis in patient's sample. Can utilize mark and unmarked MSBA principle, utilize MSBA to measure configuration according to two kinds of possible principles: interior mark adds principle and does not have principle of internal standard.

Conventional method

After the sample preparation, sample is injected into the MSBA platform.

Next take following operation;

Step 1A

At first, biomarker MS-signal demand is identified in sample.

Screening and the retention time exponential sum biomarker list quality that the biomarker respective quality is relevant separately+/-1 daltonian list that pre-determines in the biological fluids sample. Determine in several minutes available from the correlation reserved time index that most of MSBA measure, and have the changeability of about+/-2%, yet this numerical value may change.

As shown in Figure 1, by real-time mass spectrum and MSBA reference spectra storehouse coupling are implemented these steps.

Next, the quality in the MS spectrum and the biomarker quality with reference to list is compared, to approximately+/-1 dalton.

Step 1B

When biomarker candidate quality be accredited as have with reference to list in+/-during the biomarker of the MS spectrum that mates within the 1Da, its information is kept on the MSBA-server. If this quality is incorrect, the MSBA screening is not stored into the spectrum file on the server.

These operations are implemented with process communication (for example user-type of server is communicated by letter) mechanism by file-sharing.

Step 2A

When the mass property in the MS-spectrum is identified, beginning Quality Identification and Series Characteristics Anal Ysis.

Pattern match step within the MSBA software will be identified certain similarity measurement, for example the cosine correlation. Utilize similarity measurement, determined correct protein sequence. Carry out this affirmation by Spectral matching. Carry out this Spectral matching by the reference spectra in comparative sample spectrum and the MSBA database. For the positive properties in this stage, need 0.8 or higher cosine correlation factors to confirm accurate protein sequence. Threshold value of equal value for the similarity measurement that substitutes is obvious for those skilled in the art.

The comparison of reference spectra and evaluation are carried out in the following manner.

The MS/MS spectral representation is one to be listed as into logarithm (doublets) (m, v), and wherein m represents mass-to-charge ratio, and v represents the ion signal intensity level. By using m_bThe interval m is carried out binary system expansion (binning), (m_bThe developed width of=bin) MS/MS spectrum also can be expressed as vector $\stackrel{\→}{v}=\{v_1,\·\·\·,v_n\},$ Wherein its length (n) equals the quantity of bin, and the value of each key element (element) is the intensity summation of all signals within each bin. This is profile (profile) performance of MS/MS spectrum.

Two kinds of different MS/MS spectrum

Cosine correlation (S) can be such as the cosine correlation calculations according to equation 1 (eq1):

The value of S changes between 0 to 1.The A value is that two vectors of 0 expression are fully independently, is 1 o'clock at the S vector value, and this represents that the direction of these two vectors is identical.

Notice that two MS/MS spectrum vectors must have identical scale-of-two expansion (binning), that is, if the scale-of-two of the m of vector expansion (binning) is 500-501,501-502, ... 1999-2000, then another spectrum scale-of-two expansion (binned) in the same manner.As a result, the length of two vectors must be identical.

Whether the MS signal of measuring for judgement is correct biomarker, and the MS/MS of measuring-signal partly is extracted out, and compares by the MS/MS reference spectra of utilizing for example above-mentioned cosine correlativity and sample.

If cosine relevance values S is equal to or greater than predetermined threshold value, for example 0.8, judge that then the signal of measuring comes from the biomarker of inferring in the reference group.

Following part has described how to make up the reference spectra that obtains as group-specific spectrum from many individual patient.

For every kind of candidate's biomarker,, should collect several MS/MS spectrum to make up with reference to the MS/MS spectrogram in case set up described biomarker.This is the averaged spectrum from actual and measurement data set, and passes through cluster calculation and obtain.

The example of the structure of described reference spectra is as follows:

(1) collects multiple MS/MS spectrum for the target organism mark.These MS/MS spectrum must pass through MASCOT, and SEQUEST or other programs are confirmed as available from the target organism mark so that given menstruation really is flat.

(2) studied the similarity of MS/MS spectrum with above-mentioned homophylic every kind of collection.This implements by the cluster calculation of utilizing similarity measurement.This cluster calculation is implemented into similarity measurement and is reduced to predetermined threshold.After these cluster calculation, produced the summary list of the protein sequence ion of being withed a hook at the end within MS/MS spectrum.Following step is the protein sequence ion of removing from the reservation in the summary list of cluster calculation.

(3) be used to summary MS/MS spectrum foundation and qualified, may calculate the arithmetical mean of every kind of key element of spectrum vector from cluster.This mean vector can be used as the MS/MS reference spectra now.

(4) in the clustering method process, may obtain to organize the result of generation more than a kind of reference from the patient.

A) in this example, have more than a kind of cluster to be included in difference among the MS/MS spectral profile figure.The standard that is arranged on these environment is that these groups need have reliable target organism mark evaluation.Might produce then and utilize be used for a kind of target more than a kind of reference spectra.If have MS/MS fragment ions different but relevant in group, described situation will occur with the albumen of identical note.

B) also may obtain to have the situation of cluster analysis data, wherein have the difference in the biomarker profile diagram.That will be represented and can set up several group of individuals from for example different phenotypes of patient group.In these cases, relatively multiplex mode will be that phenotype is specific.But, between phenotype, also may exist biomarker overlapping.

Following illustrated, these confirm algorithm application in real time and utilization within the MSBA platform of high flux screening operation.

Step 2B

In case identified positive biomarker quality, mass spectrometer (for example Finnigan LTQ) will rest on the quality objective, to carry out the multiple scanning of biomarker ion signal.The quantity of scanning will depend on the score coupling that every kind of specific protein sequence produces, but will be calibrated to the positive properties of biomarker.Scanning window will be determined automatically by MSBA software.

Standard for positive correlativity will be to be greater than or equal to 0.8 in cosine correlativity similarity measurement.

Ensuing later step be have an albumen database by utilization commercial search engine for example MSCOT or SEQUEST or any other search engine produce the significant characteristic of statistics of protein sequence, be correct biomarker characteristic to confirm it.

The MSBA system will only store and be archived in signal and the data file within biomarker quality and the sequence scope.Generation is not transferred to the MSBA database from the every other data of this mensuration.

Step 3

The calculating that the multi-biological marker determination is read

The calculating that the multi-biological marker determination is read is implemented by using the MSBA algorithm, and described MSBA algorithm is made up of the discriminant function that calculates diagnostic MSBA score.

Discriminant function is defined as function x ₁..., x _n, x wherein _iRepresent the absolute or relative signal intensity of n of i biomarker.The output of discriminant function must be the positive or negative value according to diagnostic result.For example, if the positive of being diagnosed as, the output valve of discriminant function must be positive, and vice versa.

For example, the discriminant function of utilization is equation 2:

$\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{a}_i{x}_i-a_{\mathrm{\θ}}{x}_{\mathrm{total}}---\left(\mathrm{Eq}2\right)$

Wherein n is the quantity of diagnostic multi-biological mark.x _iBe the absolute or relative signal intensity of i biomarker, and x _TotalIt is the total signal strength that MS measures.

Be included in the weight factor in the algorithm of MSBA software in addition.

Vector { a ₁..., a _n, a _θBe the weight vector of the direction of definite hyperplane method vector that separates, it is divided into two kinds with n dimensional signal Strength Space: diagnosis is positive and diagnose negative.Determine weight vector method an example after will be described, yet, can utilize polyalgorithm, for example, supporting vector machine (Support Vector Machine), artificial neural network (Artificial NeuralNetwork) and other are to determine weight vector.

Another example of discriminant function is included in the MSBA algorithm and is defined as follows:

αf _p(x ₁，...，x _n)-βf _n(x ₁，...，x _n)????(Eq3)

Function f _pAnd f _nIt is the arbitrary function that produces the tolerance of one group of biomarker of waiting to diagnose the measurement among the patient and similarity between the reference biomarker signal in the array MSBA server or distance.f _pExpression comes the similarity or the difference function of the positive reference of self diagnosis, and f _nExpression comes the similarity or the difference function of the negative reference of self diagnosis.

α and β are the coefficients that can be used for the diagnosis of the weighting unequally positive and the negative tolerance of diagnosis.

If function f _pAnd f _nProduced similarity tolerance, then patient's sample will be diagnosed as the positive when equation 3 produces positive value.If this function produces distance metric, the positive value representation diagnosis of equation 3 is negative.

An example of described function is Euclidean distance (Euclidean distance), wherein y _iBe to wait to diagnose i biomarker signal intensity among the patient,

$\sqrt{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{(y_i-x_i)}^2}---\left(\mathrm{Eq}4\right)$

And x _iBe i biomarker signal intensity of reference group.

Another example is the standard error with the predicted value of the regretional analysis of following formula:

$\sqrt{(n\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{y}_i^2-{\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{y}_i\right)}^2-\frac{{(n\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{x}_i{y}_i-\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{x}_i\right)\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{y}_i\right))}^2}{n\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{x}_i^2-{\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{x}_i\right)}^2})/n(n-2)}---\left(\mathrm{Eq}5\right)$

Wherein n is the quantity of biomarker signal, x _iBe i measure signal intensity of patient's sample, and y _iBe by utilizing linear regression line from each x _iThe value of middle prediction, its x by measuring _iSignal and reference signal between least square fitting and calculating.

Summarize the whole software scheme among Fig. 2, comprised the algorithm of each particular step in the control method.

Step 4

Biomarker is explained and is quantitative

MSBA measures platform and is based upon:

A) principle of Fen Liing

B) principle of non-separation

In the situation of non-separation principle, we can carry out biomarker note and quantitative, by:

(a) directly MS-analyzes

I) by utilizing static nanometer electrospray principle with the direct infusion of biological sample.Utilize disposable nanometer ejector pin, wherein each nanometer ejector pin will only be exposed to a biological sample, avoid sample overload and memory effect thus.

Ii) Flow Injection Analysis pattern, wherein sample is injected as connector (plug).The sample volume of selecting in the connector is directly related with the signal intensity of biomarker protein sequence separately.Also may utilize (a few ml) sample volume injected greatly for low abundance biomarker, arrive saturated (steady state (SS)) of mass spectrometric ion signal usefulness thus.

The (iii) example enrichment flow injection of analyzing by the MS that utilizes chromatogram solid phase extractions enriching column.This step allows to clean simultaneously, by the matrix components in the removing sample and the trace enrichment of biomarker.The advantage of this method is for example to analyze biomarker the tissue extract from the sample source of high complexity.

In addition, the example enrichment pattern (iii) in, we can be created in the 2-500 scope but the amplification of signal factor that is not limited thereto.In addition, this method will improve the detectability with the biomarker of low expression level, and the accuracy of protein sequence note.

B) in compartment analysis, the biomarker that liquid phase chromatography (LC) is integrated is identified the high resolving power that depends on LC, it can (see figure x) in single-column pattern operation, or in the multicolumn pattern, utilize the post conversion operations, wherein sample improves the sample flux thus with the continuous mode analysis.

The MSBA programming

It is the code sample that calculates weight vector (or model) and utilize the core of supporting vector machine algorithm predicts diagnosis herein.This code is write as with R-Language.A kind of model (model) makes up from training data group (train), and is utilized the prediction (pred) to produce given test data set (test) then.Data set train and test are the data frameworks that comprises multiple quantity data point, it is by the target diag that comprises diagnostic result (classification value: " positive " or " feminine gender ", this value is empty for the example of pred) and the vector that comprises the signal intensity of each biomarker and total signal strength form.The MSBA programming will depend on the data that result from the protein sequence screening of implementing on two patient's groups, and biomarker results from described patient's group.

Programming within MSBA software is implemented by following manner:

Storehouse (e1071)

Model＜-svm (Diag～., data=train)

But the above-mentioned equation computation model of ##, but too simply can not reflect equation 2.

The supporting vector that ## selects is: model$SV

## and weight vector are: model$coefs

pred＜-predict(model，test)

Embodiment

Following embodiment is the illustration from lung cancer research, and described research is implemented by the LC-MS albumen profile measurement in human blood sample.Analyzed two patient's groups, case and contrast cancer group with differential protein expression difference are analyzed.

Embodiment 1

Experimental detail

Approximately get 6ml blood to the sampling test tube that comprises Calciparine/sodium salt from each patient, and mix 2-3 time up and down.Then, under 4 ℃ with 2, centrifugal 10 minutes of 000xg.From supernatant, obtain the blood plasma of three ml.Sample is-80 ℃ of freezing preservations.Next, albumen extracting and after exhausting abundant albumin human and IgG, carry out trypsinization from blood plasma.By LC-MS, MS/MS analyzes then then for (WO06100446) as previously mentioned, the fractionation plasma sample of aliquot.

The MSBA design

The multiple representation data of patient's biomarker in lung cancer research that the multi-biological mark has been summarized design liaison.

The 10-that results from MSBA method substance markers diagnosis read respectively illustration each and the every kind of biomarker (see figure 3) of living again.The quantity difference, promptly difference (see figure 1) and the quantity difference one that multiple known and that store changes in the MSBA database is used from the mensuration biomarker.The biomarkcr data that results from lung cancer research correctly is accredited as the positive with all these patients from diagnosing multiple MSBA reading.

All these 10 patients' score is in the scope of 0.80-1.0.

Fig. 4 has shown the example that the biomarker that obtains is explained from multiple assay, its MS spectrum and ensuing MS/MS spectrum (see figure 4) of the CVLFPYGGCQGNGNK biomarker that obtains of having represented of being discerned by MSBA software by biomarker is wherein represented.The MSBA coupling is utilized the reference biomarker spectrum in the MSBA-database of having used the cosine correlativity, shows that the cosine correlativity is equal to or higher than 0.8.

Table 1 is represented the details of MSBA-data generation, has wherein analyzed the predetermined quality of the biomarker that is conditioned.

Table 1

The biomarker number of signals	Retention time/minute	MS-value constituent mass m/z	MS/MS-value constituent mass m/z	The average correlation score of MS/MS spectrum	Multiple changes
						??1	??10.433	??608.5	??276.6，363.2，439.2，449.2，457.8，??513.8，522.3，??552.3，572.8，577.3，664.4，740.4，??777.4，853.4，??914.5，940.5，1043.5，1069.5	??0.94	??3.682
??2	??19.94	??862.5	??229.1，286.2，298.2，303.2，325.2，??432.2，??506.8，515.3，550.3，560.3，567.4，??595.4，??603.3，611.8，620.3，621.4，666.4，??673.4，??795.4，802.4，856.4，873.4，908.5，??1008.6，??1012.5，1036.6，1137.6，1165.6??1222.6，??1265.7，1293.7	??0.96	??3.625
						??3	??25.744	??492.1	??320.2，401.2，418.2，475.2，545.3，??562.3，??563.3，644.3，661.3，721.4，749.4，??806.4，??807.4，824.4	??0.92	??0.336
??4	??32.057	??838.6	??245.1，499.2，727.4，760.3	??0.93	??0.442
						??5	??32.149	??682.5	??373.7，467.2，492.8，694.9，708.9，??798.9，809.4，967.0，1001.6，??1286.7，1301.7，1487.8	??0.91	??0.492
??6	??35.108	??767.5	??175.1，331.7，410.7，411.7，420.2，??507.2，549.2，598.8，707.3，792.4	??0.93	??3.250

??7	??36.113	??908.6	??426.7，531.2，532.3，661.4，778.9，??792.9，793.4，842.4，850.9，??1008.4，1023.5，1110.5，1155.5，??1284.6，1285.6，1341.6，1383.6，??1428.6，1499.6，1612.7，1669.8	??0.90	??3.571
						??8	??36.232	??1010.5	??447.3，550.3，562.3，562.8，636.3，??690.3，723.3，768.4，837.9，881.4，??1053.5，1124.5，1271.6，1342.6，??1445.6，1558.7，1615.7，1745.8	??0.87	??3.4
??9	??69.621	??851.1	??-	??0.89	??0.355
						??10	??69.71	??1236.5	??379.2，405.2，463.7，514.3，587.8，??631.4，638.3，692.4，737.8，760.4，??761.4，816.9，895.4，903.9，917.9，??956.5，1080.5，1137.1，1157.6，??1158.6，1193.6，1229.1，1236.1，??1243.1，1278.6，1286.6，1343.2，??1345.6，1615.8，1616.8，1632.8	??0.92	??0.357

Embodiment 2

Following another embodiment comes from the lung disease case-control study, and it is implemented by the LC-MS albumen profile measurement (profiling) in the human blood sample.Analyze two patient's groups, have the case and the contrast lung disease group of the protein expression analysis of difference.

Experimental detail

The program of plasma sample collection and preparation is same as the previously described embodiments.

MSBA modelling and evaluation

46 patient's sample datas being made up of 10 cases and 36 contrast are used to make up the MSBA model.We have made up respectively and have comprised 14,8,26,8, with 5 of 11 signals different MSBA models.Made up after 5 models, demonstrated last MSBA model (5 ^Th) have a significant resolving ability.We only use the 5th model in the step below thus.The LC-MS information of 11 signals of last model (retention time (minute)/MS-value (m/z)) as follows: 11.6/485.2,12.5/608.1,18.3/547.0,20.2/681.3,21.1/575.1,21.3/531.5,25.5/561.6,23.1/514.5,32.5/682.2,44.0/985.2,48.7/945.8.

Be the predictability of evaluation model, we utilize 10 cases and 9 contrasts to attempt prediction case/contrast, and it is used as blind sample.According to above-mentioned MSBA diagnostic routine (also being illustrated in Fig. 2), each specimen is identified 11-substance markers signal of living again, and carry out each peak-to-peak signal quantitatively.From the quantity of the heavy signal of all 11-, utilize following formula (equation 5) to calculate each experimenter's MSBA score.In this embodiment, the MSBA score is equal to or greater than 1 experimenter and is diagnosed as case (table 2, MSBA diagnosis).As a result, we can correctly predict all samples, that is, resolving ability is 100% (see figure 5).

Embodiment 3

Following embodiment also comes from the lung disease case-control study, and it is implemented by the LC-MS albumen profile measurement (profiling) in the human blood sample of two patient's groups (case and contrast).In this embodiment, the multi-biological mark of another group is used to make up the MSBA model, and it has much bigger different patient data groups.

Experimental detail

The program of plasma sample collection and preparation is same as the previously described embodiments.

MSBA modelling and evaluation

Be used as the training data group by 21 cases and 75 sample datas that contrast 96 patients that form.When sample size improved, the sample variability also improved.Therefore at first we have used the Smirnov test to remove abnormal signal.As a result, 5 samples that comprise a lot of exceptional values are also removed from analysis bank.Utilize the result of t-test, we have made up the initial MSBA model that comprises 100 candidate's biomarker signals.Remove minimum contributing signal by recursively using discriminant analysis from distinguish model then, we have obtained to have the MSBA model of 10 signals at last.See Table 3 10 heavy label signals tabulation.

In this embodiment, be computational discrimination formula score (z), the scoring function that we have utilized another following formula to represent.

z＝∑a _i·x _i+C????(Eq.6)

(x _i: signal intensity, a _i﹠amp; C: the described coefficient of table 3)

If score value is positive, it is interpreted as case, otherwise is contrast.

Be the predictability of evaluation model, we utilize identical data set (21 case+75 contrasts, 96 altogether) to attempt prediction case/contrast.According to above-mentioned MSBA diagnostic routine (also being illustrated in Fig. 2), to each sample identification the 10-substance markers signal of living again, each peak-to-peak signal is carried out quantitatively.From the quantity of the heavy signal of all 10-, utilize following formula (equation 6) to calculate each experimenter's MSBA score.In this embodiment, the MSBA score is equal to or greater than 0 experimenter and is diagnosed as case.Table 4 and Fig. 6 show from distinguishing the result.In Fig. 6, each point is represented a patient, and Z-axis is represented discriminant score (z).In this embodiment, sensitivity is 85.7%, and specificity is 82.7%, and accuracy of forecast is 83.3%.

Table 2

Table 3

Table 4

Claims (17)

1. determine the method that one or more polypeptide biomarkers exist in the sample, said method comprising the steps of:

(a) make sample carry out the retention time exponential sum respective quality that each detection signal was analyzed and write down to mass spectrum (MS);

(b) make corresponding to the reference database of each quality of signals and be associated with the biomarker quality, forming each signal and with reference to the correlativity between the biomarker, and abandon those quality not with the relevant signal of reference biomarker quality;

(c) store those quality with reference to the relevant signal of biomarker;

(d) by utilize similarity measurement with in the MS spectrum of each signal and the database with reference to the MS Spectral matching of biomarker, determine each storage assembly and with reference to the correlativity between the biomarker, to define one group of positive coherent signal;

(d) measure the intensity of each positive coherent signal, and utilize discriminant function that its absolute signal intensity or relative signal intensity are kept the score;

(e) with threshold application in available from discriminant function score value, to determine existing or not existing of biomarker.

2. the process of claim 1 wherein that making specimen carry out MS under the situation that does not have lock out operation in advance analyzes.

3. the method for claim 2, wherein direct infusion, Flow Injection Analysis or example enrichment Flow Injection Analysis specimen by utilizing static nanometer electrospray principle.

4. the process of claim 1 wherein that specimen is processed before MS analyzes.

5. the method for claim 4, wherein said sample preparation comprises the sample separation by single-phase or heterogeneous high pressure lipuid chromatography (HPLC) (HPLC).

6. the method for aforementioned each claim, wherein MS is electrospray ionization (ESI) MS, substance assistant laser desorpted ionized-flight time (MALDI-TOF) MS or surface-enhanced laser desorption ionization-flight time (SELDI-TOF) MS.

7. the method for aforementioned each claim, the reference mass and the MS spectroscopic data that wherein are used for multiple biomarker store with electronic form or paper spare form.

8. the method for aforementioned each claim, the reference MS spectrum that wherein is used for given biomarker are by averaged spectrum actual and the measurement data acquisition by cluster calculation.

9. the method for aforementioned each claim wherein joins in the sample before analyzing with reference to being marked on MS in one or more of peptide.

10. the method for claim 9, wherein in mark molecular label mark.

11. the method for claim 9, wherein interior mark is labeled and is included in the main data group.

12. the method for claim 11, wherein absolute signal intensity is by measuring the biomarker signal intensity and itself and one or more known interior mark relatively being kept the score.

13. each method of claim 1-9 is not wherein handled sample under the target situation in not adding.

14. the method for claim 13 is wherein kept the score to relative signal intensity by the ratio between the reference signal strength of individual biomarker signal intensity and patient group among the measurement patient.

15. the method for claim 13, it is full automatic.

16. the process of claim 1 wherein that calculating utilizes any clinical variable from the discriminant function of the score of MS signal intensity optional comprising, for example the phenotype of result of clinical detection and/or clinical observation and/or medical records.

17. determine the diagnostic method of the existence of disease, described method comprises the protein sequence biomarker of compare test sample and with reference to biomarker, wherein comprises the peptide of identifying in the table 1 with reference to biomarker.

Images