WO2007022625A1 - Peptide reversion methods and uses thereof - Google Patents

Abstract

A proteomic method for evaluating and comparing the efficacy of therapeutic treatments is described. Peptide expression profiles are determined by LC-MS analysis of biological samples and comparisons of peptide expression profiles across conditions such as diseased, healthy and treated are employed to monitor and evaluate changes in protein expression associated with each condition. Peptide expression profiles that are affected in the disease state and return to levels observed in the healthy condition upon treatment provide a sensitive and comprehensive measure of treatment efficacy. Peptide expression profiles that are affected only during treatment and not in the disease state (relative to healthy) are used as a measure of potentially non-therapeutic treatment effects.

Description

PEPTIDE REVERSION METHODS AND USES THEREOF

Background of the Invention

In general the invention features methods for identifying and using differentially expressed peptides to evaluate and predict drug compound efficacy, toxicity, and pharmacodynamic effects.

Current estimates for the costs of drug development exceed $800 million per drug The high cost is largely due to the high rate of failure m drug development Estimates place the failure rate at approximately 90% at the target discovery stage and approximately 66% at the Phase II clinical trial stage (FDA Center for Drug Evaluation and Research, 1999). There are many reasons for the high failure rate and cost of pharmaceutical drug development. In the initial stages of drug development, a unique drug target is generally identified and validated before a series of molecules are designed and screened against the target. However, a target is not fully "validated" until it has demonstrated efficacy in a human clinical trial implying that the relevance of a therapeutic target is uncertain until it has completed a Phase II human clinical trial. Companies are therefore faced with the difficult decision of screening molecules against a drug target before it is fully validated or validating the target first and then screening for drug molecules that have the desired effect on the target. Either scenaπo is a costly one.

Once a target is identified and compounds are screened to identify lead compounds, the development of the lead compounds is also fraught with risk. Lead compounds are usually optimized for affinity, solubility, and/or other desired pharmacokinetic features. One problem with optimization of the lead compound is that the optimization of one parameter, for example, affinity, can negatively affect another parameter, such as solubility. This makes the selection and optimization of a drug compound for human clinical use problematic. Ultimately, once a lead compound is selected and optimized, the compound is selected for further animal model testing or testing in human clinical trials. This selection is based on an examination of all pre-clinical data including the various features of the target (e g., the effect of transfection in cell lines, the effect of knock-outs in vivo, the evidence of modulation m disease, and genetic association studies) and the compound (e g., the size, affinity for target, selectivity for various related targets, solubility, and half-life). Since these parameters are currently measured independently, it is also exceedingly difficult to obtain a global view of the efficacy of a drug prior to the initiation of human clinical tπals.

Even when animal models are used to predict drug efficacy prior to human clinical tπals, these data are limited by the overall predictive value of many of the various animal models for human disease. For example, in general, animal models of oncology, neurodegenerative disorders, disorders of cognition and certain inflammatory or autoimmune disorders have been shown to be poor predictors of human clinical outcome. Thus, many compounds that work m animal models have not been found to be efficacious m humans and vice versa.

Once the drug compounds are in clinical trials there are several additional possible reasons for the failure of the compound. A compound may fail because it is ineffective or only moderately effective if it acts only on a single target in a signaling network containing multiple redundant pathways. Another related reason for the failure of certain late stage clinical compounds is that they induce adverse or life-threatening side-effects m humans. Yet another reason that certain drugs fail m clinical trials may be in part due to the selection of a disease population that is too heterogeneous for the clinical trial and that is not reflective of similar disease etiologies.

Drug compounds also fail in clinical trials because, for many diseases, the clinical endpomts are generally subjective or comprise disease activity indices which contain highly subjective parameters. Because of the subjectivity of the endpoint measurements in these disease activity indices, the measurements need to be validated for a variety of features including reliability, face validity, content validity, and sensitivity. This is usually achieved through concordance of scores between expert opinions, acceptable inter-observer variability among evaluators, correlation between individual patient scores on different indices and correlation between increases m individual patient's scores on indices and clinical decisions to change therapy.

One of the methods currently contemplated to reduce drug development costs and failure rates is the use of biomarkers of specific disease. Biomarkers specific to a disease can ideally be used as surrogate endpoints in the analysis of drug compound efficacy. However, many diseases are poorly characterized and lack biomarkers or surrogate end-points for drug efficacy measurement. Although new biomarkers of diseases are increasingly being identified, single protein markers are often inadequate due to lack of good correlation with disease outcome. In addition, certain biomarker assays are designed to specifically measure target binding or downstream target activation processes. These types of biomarkers are good indicators of target binding and/or activation but are not necessaπly useful for predicting drug efficacy. The utility of these types of biomarkers is dependent on the hypothesized etiology of disease. Thus, it is possible for these types of biomarkers to yield misleading false-positive or false-negative results for clinical efficacy. For example, estramusme was developed as an estrogen receptor alkylating agent. It was approved but later found to be an antimitotic which does not interact with the classical estrogen receptor. Had estrogen alkylation been used as a biomarker, a true negative result for the proposed mechanism of action may have become a false-negative prediction for clinical benefit. In addition, Cetuximab, which is indicated for EGF-receptor positive tumors has recently been shown to induce therapeutic responses in EGF-receptor negative tumors. In contrast, many drugs for sepsis which were aimed at blocking cytokine-mediated inflammatory responses, failed m late stage clinical tπals despite showing significant reductions in inflammatory molecules following therapeutic treatment These examples illustrate how biomarkers for target presence, binding, or activation can lead to false-negative or false-positive indicators of clinical benefit.

Despite the limitations of certain types of biomarkers, identification of new biomarkers and their ultimate development into surrogate endpoints is nonetheless widely appreciated as one of the best ways to reduce drug development costs today. Nevertheless, it remains difficult, time-consuming, and costly to demonstrate appropriate statistical correlation between specific protein changes in the biomarker and disease outcome. Most individual protein biomarkers in current use, such as PSA, CEA, or CA-125, lack the appropriate sensitivity and selectivity to be used as surrogate endpoints in clinical tπals.

A global method of measuring patient's health status which is not reliant on individual protein observations would be advantageous in the identification, validation, and prediction of clinical benefit for candidate drug compounds A global examination of many thousands of peptides in plasma can provide directional guidance as to whether a therapy is beneficial by comparison to the peptide abundance seen m healthy controls. It would also be advantageous to have an objective measure of drug action in the patient in a relatively short time-frame (such as days following treatment) and to be able to non-mvasively monitor the health status of the patient throughout treatment and before the measurement of clinically accepted endpoints.

Furthermore, given that blood is routinely drawn in the hospital for the evaluation of many different substances and blood components to gam insight into the health status of the patient, it would be desirable to have methods for measuring the efficacy of drug compounds using blood or plasma proteins as biomarkers. Most accepted surrogate endpoints such as glucose levels, cholesterol levels, and viral load measurements, are deπved from blood, plasma, or serum. Although the specific protein content of blood is not known, blood has recently been shown to possess a large number of proteins, many of them previously considered to be exclusively intracellular proteins. Given the many limitations and hurdles to drug compound development in the pharmaceutical industry, the availability of an objective, quantifiable, and widely applicable measure of drug efficacy is needed to reduce the costs and failure rates of drug development. Ideally, such a measure of drug efficacy would be reflective of a subject's global protein profile and not reliant on individual protein biomarkers or surrogate endpoint observations. Summary of the Invention

The invention provides improvements in the productivity of drug development, including reductions in the failure rate of compounds in clinical trials and improvements in the accuracy of predictions regarding which compounds will be efficacious and which will have the potential for adverse side effects. Amond the advantages provided by the invention are reduction of the high costs of drug development.

We have discovered methods for using differentially expressed peptides, that is to say peptides whose expression changes significantly in disease conditions versus controls, as predictors of drug compound efficacy by measuring the extent to which these peptides revert to control levels upon drug treatment. Such methods can be used as objective, quantifiable measurements in a variety of clinical or research decisions at any stage in drug development, drug evaluation, and patient population determination. For example, information obtained from the methods descπbed herein can be used to identify lead drug compounds, to determine relative therapeutic efficacy between two or more candidate compounds, to determine toxicity for a drug compound or relative toxicity between two or more candidate compounds, and to identify patients that are likely to respond to a particular compound and to identify patients that should not be given a particular compound because of toxicity and side-effects that are particularly harmful to a specific patient population. The methods can also be used to determine subtle differences between potential drug analogues or to compare generics to the brand-name equivalents, to determine the optimal dosages for a drug compound, and to monitor a patient's response to a drug compound at a particular dosage and to make modifications as needed. In addition, the methods described can be used to identify specific proteins that can be used as candidate biomarkers for particular diseases and to determine the mechanism of action for a particular drug compound. The methods can also be used for determining the health status of an individual or to monitor the treatment or progression towards a disease state for an individual undergoing treatment for a disease or at πsk for developing a disease. In addition, the methods of the invention can be used to identify a responder or a non-responder patient population that can be used to design clinical trials for particular drug compounds or to select appropriate animal models of disease for pre-climcal studies by comparing the protein profile of vaπous animal models to those of human disease

The present methods are particularly advantageous because they provide rapid, objective, global measures of drug efficacy, which are not dependent on a specific drug target and which may be obtained using readily accessible fluids such as blood, plasma, or urine. It will be understood by the skilled artisan that the methods descπbed herein refer specifically to peptides but can be used for any type of biomolecule including nucleic acids, metabolites, and oligosaccharides. Accordingly, in general, in a first aspect, the invention features a method for determining the therapeutic efficacy of a compound for the treatment of a disease in a subject The method includes several steps. First, the expression levels of a plurality of disease- related peptides from a normal reference subject, from a subject having the disease, and from a subject having the disease and treated with the compound are obtained.

These disease-related peptides can be obtained in a variety of ways including from a database, from the literature, and experimentally from a biological sample from each of the subjects. If the disease-related peptides are obtained experimentally, the following steps can be used. A first biological sample from the normal reference subject and a second biological sample from the subject having the disease are obtained and the expression levels of a plurality of peptides in the two samples are measured. The levels can be measured by any method known in the art including chromatography, mass spectrometry, or immunological methods. Generally, the method used to measure the expression levels will measure the abundance of the plurality of peptides. Optionally, the samples can be depleted of high abundance proteins or contacted with a protease or fractionated, or any combination therein, prior to measuring the expression levels. The expression levels of the plurality of peptides are compared, using statistical analysis or computer implemented methods, or both, and peptides that are differentially expressed in the first biological sample from the normal reference as compared to the second biological sample from the subject having the disease are identified as "disease-related peptides." Taken together, the total group of disease-related peptides are identified as a plurality of disease-related peptides. In general, a plurality of disease-related peptides includes at least 50 peptides, preferably at least 100 peptides, 150 peptides, 200, 300, 400, 500, 600, 700, 800, 900, 1,000, 2,000, 5,000 or at least 10,000 peptides.

By comparing the expression levels of the plurality of disease-related peptides in the biological sample from the subject having the disease and treated with the compound with the expression in the normal reference subject and the subject having the disease, the percentage of the plurality of disease-related peptides in the subject having the disease and treated with the compound whose expression reverts to substantially the same levels as for the normal can be determined. The compound is therapeutically effective for the treatment of the disease if a substantial percentage of the plurality of disease-related peptides reverts to substantially the same expression levels of the normal reference subject In general, a drug is considered highly effective if at least about 80% of the disease-related peptides revert to substantially the same level of expression as found in the normal reference sample; moderately effective if between about 50% and 80% of the disease-related peptides revert to substantially the same level of expression as found in the normal reference sample; and weakly effective if between about 20% and 50% of the disease-related peptides revert to substantially the same level of expression as found in the normal reference sample. The determination of therapeutic effectiveness can also be determined using a relative comparison. For example, using the methods descπbed herein, the extent of peptide reversion for a disease-related peptide, peptides, or peptide cluster from disease to normal can be compared for two drug compounds The drug compound that produces the relatively greater peptide reversion is considered to be therapeutically effective relative to the drug compound that produces the lesser peptide reversion.

In another aspect, the method descπbed in the first aspect is adapted for use m the determination of the therapeutic efficacy of one drug compound, to select therapeutically effective candidate compounds from a library of compounds, or to compare the relative efficacy of two or more drug compounds (e.g., where the compounds are analogs of each other or one of the compounds is a geneπc version of the other one of the compounds). For direct comparisons of two or more compounds, the same general method is followed using an additional one or more biological samples from one or more subjects having the disease and treated with the additional compounds are used m the method. In this aspect, the percentage of the plurality of disease-related peptides in the subject having the disease and treated with the compound whose expression reverts to substantially the same levels as for the normal is compared to the percentage of the plurality of disease-related peptides in the subject having the disease and treated with the additional compound whose expression reverts to substantially the same levels as for the normal. The drug compound that produces the relatively greater peptide reversion is considered to be therapeutically effective relative to the drug compound that produces the lesser peptide reversion.

In another aspect, the method descπbed in the first aspect is adapted for use in the determination of the optimum dosage for a compound for the treatment of a subject having a disease. For direct comparisons of two or more dosages of a compound, the same general method is followed. First, the expression levels of a plurality of disease-related peptides from a normal reference subject and from a subject having the disease are obtained The expression levels of the plurality of disease-related peptides in a biological sample from a subject having the disease and treated with a first compound and m a second biological sample from a second subject having the disease and treated with a second biological compound are measured. By comparing the expression levels of the plurality of disease- related peptides in the first and second biological samples with the expression in the normal reference subject and the subject having the disease, the percentage of the plurality of disease- related peptides in the first and second biological samples whose expression reverts to substantially the same levels as for the normal subject can be determined. In this aspect, the percentage of the plurality of disease-related peptides in the subject having the disease and treated with the compound at a first dosage whose expression reverts to substantially the same levels as for the normal is compared to the percentage of the plurality of disease-related peptides m the subject having the disease and treated with the compound at a second dosage whose expression reverts to substantially the same levels as for the normal. The dosage for the drug compound that produces the relatively greater peptide reversion is the optimum dosage. This method can be used to compare a number of possible dosages for a drug compound, for example, two dosages, three dosages, four dosages, five dosages, or more. If desired, a standard curve using the data for the plurality of disease-related peptides can be determined.

In another aspect, the invention features a method for comparing the toxicity of two or more compounds for the treatment of a subject having a disease. The method includes a number of steps. First, a peptide expression profile for a plurality of study peptides in a first biological sample from a first subject having the disease and treated with a first compound and a second biological sample from a second subject having the disease and treated with a second compound are obtained. A peptide expression profile for a plurality of study peptides from a third subject that is a normal reference subject and a fourth subject that is a subject having the disease are also obtained. The peptide expression profile for each of the subjects can be obtained from any source including the literature or a database, or can be obtained experimentally. For example, a biological sample from any of the subjects can be obtained and the expression levels of a plurality of peptides is measured m each biological sample using mass spectrometry, chromatography, or immunological methods. The expression levels for the plurality of peptides is then compiled into a peptide expression profile.

The peptide expression profiles from the first biological sample is then compared with the peptide expression profile from the normal reference subject and the peptide expression profile from the subject having the disease. Based on this comparison, a plurality of disease-independent peptides that are differentially expressed in the peptide expression profile from the first biological sample as compared to the peptide expression profile from the third subject and the peptide expression profile from the fourth subject is identified The peptide expression profile of the second biological sample is also compared to the peptide expression profile from the third reference subject and the peptide expression profile from the fourth subject and a plurality of disease-independent peptides that are differentially expressed in the peptide expression profile of the second biological sample as compared to the peptide expression profile from the third subject and the peptide expression profile from the fourth subject is identified. The first compound is more toxic than the second compound if the plurality of disease-independent peptides identified for the first biological sample is greater than the plurality of disease-independent peptides identified for the second biological sample Conversely, the second compound is more toxic than the first compound if the plurality of disease-independent peptides identified for the second biological sample is greater than the plurality of disease-independent peptides identified for the first biological sample. Such relative toxicity comparisons can be made for any number of compounds including two, three, four, five, six, seven, eight, nine, ten, or more drug compounds. The compounds can be analogs of each other or generic versions of each other. The method can also be used to select a relatively less toxic candidate compound from a library of compounds. The methods used to determine drug compound efficacy and drug compound toxicity can be used alone or in combination, for example, for the selection of a drug compound that produces the greatest relative efficacy with the lowest relative toxicity as compared to one or more additional compounds.

In another aspect the invention features a method for identifying an animal model of a human disease. The method includes several steps. First, the expression levels of a plurality of disease-related peptides for a first subject that is a normal reference subject and a second subject having the disease are obtained. The expression levels of a plurality of disease-related peptides from a third subject that is a normal reference animal and from a fourth subject that is a candidate animal model for the human disease, wherein the third and fourth subjects are the same species are also obtained. The plurality of disease-related peptides obtained for the first and second subjects are compared with the plurality of disease-related peptides for the third and fourth subjects. The animal is a model for said human disease if a substantial percentage of the disease-related peptides obtained for the fourth subject is expressed at substantially the same expression levels as the disease-related peptides obtained for the second subject. Desirably, the animal is a mammal such as a mouse, bovine, equine, canine, ovine, and feline. The disease can be any disease, such as cancer, immune diseases, haematopoietic diseases, reproductive system disorders, musculoskeletal diseases, cardiovascular diseases, mixed fetal diseases, excretory diseases, neural or sensory diseases, respiratory diseases, endocπne diseases, digestive diseases, connective tissue diseases, and epithelial diseases.

In another aspect the invention features a method for predicting the health status of an individual at risk for or suspected of having a disease. The method includes several steps. First, a peptide expression profile for a plurality of peptides from a first subject that is a normal reference subject, a peptide expression profile for a plurality of peptides from a second subject having the disease, and a peptide expression profile for a plurality of peptides from a third individual at risk for or suspected of having a disease is obtained. The peptide expression profile can be obtained using any of the methods descπbed herein. The peptide expression profiles for each of the subjects is the compared. The health status of the individual is then predicted based on the substantial identity of the peptide expression profile of the third biological sample to the peptide expression profile of the first and second biological samples. For example, the individual at πsk for or suspected of having a disease is diagnosed with the disease if the peptide expression profile of the third biological sample is at least 75% identical to the peptide expression profile of the second biological sample.

This method can used to diagnose progression towards a disease state by obtaining a biological sample from the individual at risk for or suspected of having a disease at two or more time points, where an increase over time m the identity between the peptide expression profile for the individual at risk for or suspected of having a disease and the peptide expression profile for the second subject having the disease indicates progression towards the disease state.

For any of the aspects of the invention, the subject can be one subject or a group of subjects, also known as a sample set, that includes more than one subject (e.g., 2 or more, 5 or more, 10 or more, 20 or more, or as many as needed for the study. The biological sample used in any of the methods is preferably a bodily fluid such as blood, plasma, serum, urine, bile, cerebrospinal fluid, aqueous or vitreous humor, an exudate, and fluid obtained from a joint. The methods described herein are useful for a variety of diseases including, but not limited to, cancer (e.g. leukemia, colon, bone, breast, liver and lung cancer), immune or haematopoietic diseases (e.g. anemia, Hodgkm's disease, acute lymphocytic anemia, multiple myeloma, arthritis, asthma, AIDS, autoimmune disease, inflammatory bowel disease, psoriasis and Lyme disease), reproductive system disorders (e.g. prostatitis, inguinal hernia, varicocele, penile carcinoma, ovarian adenocarcinoma and Sertoh-leydig tumours), musculoskeletal diseases (e.g. giant cell tumours, Paget's disease, systemic lupus erythematosus, gout, muscular dystrophy and cachexia), cardiovascular disease (e.g rhabdomyomas, heart disease, arrhythmia, cardiac arrest, heat valve disease, hypernatraemia and hyponatraemia), mixed foetal diseases (e.g. foetal alcohol syndrome, Down's syndrome, Patau syndrome, Turner's syndrome, Apert syndrome and Tay -Sachs disease), excretory diseases (e.g. urinary incontinence, urinary tract infections or renal disorders), neural or sensory disease (e.g. Alzheimer's disease, Parkinson's disease, cerebral malaria, meningitis, cerebellar ataxia, attention deficit disorder, autism and obsessive compulsive disorder), respiratory disease (e.g. emphysema, lung cancer and occupational lung disease), endocnne diseases (e.g. diabetes, Addison's disease and glomerulonephritis), digestive diseases (e g. portal hypertension, irritable bowel disease, gastric atrophy or pancreatitis),and connective tissue or epithelial diseases (e.g. Crohn's disease, scleroderma, wound healing and epidermolysis bullosa).

As used herein, by "biological sample" or "sample" is meant any solid or fluid sample obtained from, excreted by, or secreted by any living organism, including single-cell micro-organisms (such as bacteria and yeast) and multicellular organisms (such as plants and animals, such as a vertebrate or a mammal. A biological sample may be a biological fluid obtained from any location (such as blood, plasma, serum, urme, bile, cerebrospinal fluid, aqueous or vitreous humor, or any bodily secretion), an exudate (such as fluid obtained from an abscess or any other site of infection or inflammation), or fluid obtained from a joint (such as a normal joint or a joint affected by disease such as rheumatoid arthritis). Preferably, the biological sample used in the methods of the invention is a blood or plasma sample Alternatively, a biological sample can be obtained from any organ or tissue (including a biopsy or autopsy specimen) or may comprise cells (whether primary cells or cultured cells) or medium conditioned by any cell, tissue or organ. If desired, the biological sample is subjected to preliminary processing, including preliminary separation techniques (e g , fractionation or chromatography) or depletion of abundant peptides For example, cells or tissues can be extracted and subjected to subcellular fractionation for separate analysis of biomolecules in distinct subcellular fractions, e.g., proteins or drugs found in different parts of the cell. A sample may be analyzed as subsets of the sample, e.g., bands from a gel.

By "sample set" is meant a collection of one or more samples that are grouped together for analytical purposes and generally involves samples having one or more common properties. Common properties include, for example, the source of the samples or whether the sample was derived from control subjects, subjects diagnosed with a particular disease, or being treated with various drugs.

By "biomolecule" is meant any organic molecule that is present in a biological sample, including peptides, polypeptides, proteins, post-translationally modified peptides or proteins (e.g., glycosylated, phosphorylated, or acylated peptides), oligosaccharides, polysacchaπdes, lipids, nucleic acids, and metabolites. Preferably, the biomolecule is a peptide or a polypeptide.

By "biomarkers" is meant a biomolecule, preferably a peptide that is differentially expressed and used to distinguish one population from another. For example, a population biomarker for a disease is a peptide that is differentially expressed in diseased patients as compared to normal controls and can be used to diagnose or monitor disease progression or to determine the therapeutic efficacy of a drug compound for the treatment of the disease. By "compound" is meant any small molecule chemical compound, antibody, nucleic acid molecule, polypeptide, or fragments thereof. By "computer implemented methods" is meant any method that includes the use of software or hardware for a computer. Preferred computer implemented methods include Mass Intensity Profiling System (descπbed in U.S. Patent Application Publication No. 2003/0129760; hereinafter referred to as "MIPS") and Constellation Mapping (descπbed in PCT publication number WO 2004/049385; hereinafter referred to as "Constellation Mapping").

By "differentially expressed" is meant a relative difference (e.g., an increase or a decrease) in the expression of a biomolecule (e.g., a peptide) in one sample as compared to another sample. Differential expression can include a difference m intensity or abundance of the biomolecule as measured by methods known in the art such as immunological methods or computer implemented methods such as MIPS or Constellation Mapping. The differential abundance of a protein contained within a sample is linearly related to the differential intensity measurement of its peptide ions through mass spectrometry according the following formula dA=1.931 IdI-1.0523, hence differential intensity or abundance measures are used here interchangeably. The relative difference in expression can be any detectable difference in the expression of the biomolecule (e.g., intensity or abundance) using methods known in the art or described herein and can be expressed as a % difference or a fold difference (e g , a 10%, 30%, 50%, 70%, , 90%, 150%, 200%, 300%, 500% or greater difference in expression from one sample as compared to another sample or a 1.5-fold, 2-fold, 2.5-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold or greater difference in expression from one sample as compared to another sample). A biomolecule (e.g., a peptide) expressed at "substantially the same expression levels" as another biomolecule (e.g., a peptide) will have an expression level that is at least 75%, 80%, 90%, 95%, 96%, 97%, 98%, 99% or more identical to the expression the comparison biomolecule (e.g., a peptide). Methods for detection of peptide expression include mass spectrometry, virtual mass spectrometry, tandem mass spectrometry, MIPS, Constellation Mapping, any immunological detection methods, or any combinations thereof. Methods for comparing relative expression for the determination of substantially the same or different expression levels are known in the art and include parametric tests such as T-tests, ANOVAs and non-parametric tests such as Wilcox test and permutation analysis.

By "differentially expressed peptide" is meant a peptide that has a relative difference in its expression between samples or a set of samples. Subsets of differentially expressed peptides include disease-related peptides and disease-independent peptides.

By "differentially expressed peptide cluster" is meant a plurality of differentially expressed peptides that are differentially expressed in one sample set as compared to another sample set. In general, a "plurality" includes a minimum of 50 peptides.

By "diagnostically differentially expressed peptide" or "diagnostically differentially expressed peptide cluster" is meant a peptide or cluster of peptides that is determined to be differentially expressed from one sample set to another and subjected to diagnostic criterion such as thresholds for the intensity value T such that no more than 30%, 20%, 10%, 5%, or fewer of the patients in the up condition have an intensity below T and no more than 30%, 20%, 10%, 5%, or fewer of the patients in the down condition have an intensity above T. By "disease-related peptide" is meant a digestion fragment, a peptide, a polypeptide, or a source polypeptide which is differentially expressed or found to be present at significantly higher or lower abundance or intensity in a sample from a subject having a disease as compared to a normal reference sample. Disease-related peptides can be subjected to diagnostic criterion to identify a subset that is diagnostically differentially expressed. The subset can be the same size or smaller than the total size of disease-related peptides. It should be understood that not all disease-related peptides or proteins are causally involved in the disease process. Many disease-related proteins will be related to secondary or tertiary events in disease.

By "disease-independent peptide" is a digestion fragment, a peptide, a polypeptide, or a source polypeptide which is differentially expressed or found to be present at significantly higher or lower abundance or intensity in a drug-treated (diseased) sample as compared to disease (untreated) and as compared to a normal reference. A disease-independent peptide is not a disease-related peptide. Disease-independent peptides may be new peptides which were not observed in control and diseased sample sets or peptides which were previously unchanged between these groups which become significantly differentially expressed upon drug treatment. By "peptide expression profile" is meant a compilation of the expression levels of a plurality of peptides measured for a biological sample. A peptide expression profile can be obtained through mass spectrometπc measurement of peptides such as LC-MS from a various samples or through other immunological means which measure peptide abundance. A peptide expression profile can be expressed in table or graphic formats. It is frequently expressed as relative fold changes over a reference sample set and can also be expressed as percent similarity to a reference sample set according to statistical methods known in the art. In certain graphic presentations as illustrated here, individual peptide expression profiles can be expressed as intensity plots or heat maps. Intensity plots are images in which hue, a symbol, intensity, and/or color are functions of the intensity, or related statistics (such as mean, median, p-value of difference or vaπance). Figures 2, 4B, 5B, 7, and 9 are examples of heat maps In heat maps, colors represent vaπous biomolecule levels. Colors can also be replaced or augmented, for example by black and white, grey scale, or symbols If the desired information cannot be encoded by the intensity plot's available colors, hues, intensities, or symbols, multiple plots may be produced and placed near each other or overlaid using transparencies and/or electronic displays.

By "central tendency" is meant the mean or median or means or medians weighted by vaπous other quantities, or robust estimators of central tendency such as trimmed mean or mean of values lying in a specified percentile range. Other statistical measures of central tendency are not excluded By "disease" is meant any abnormal condition of the body or mind that causes discomfort, dysfunction, or distress to the subject having the disease. Disease particularly amenable to the methods of the invention include cancer (e.g. leukemia, colon, bone, breast, liver and lung cancer), immune or haematopoietic diseases (e.g. anemia, Hodgkm's disease, acute lymphocytic anemia, multiple myeloma, arthritis, asthma, AIDS, autoimmune disease, inflammatory bowel disease, psoπasis and Lyme disease), reproductive system disorders (e.g. prostatitis, inguinal hernia, varicocele, penile carcinoma, ovarian adenocarcinoma and Sertoh- leydig tumours), musculoskeletal diseases (e.g. giant cell tumours, Paget's disease, systemic lupus erythematosus, gout, muscular dystrophy and cachexia), cardiovascular disease (e.g. rhabdomyomas, heart disease, arrhythmia, cardiac arrest, heat valve disease, hypernatraemia and hyponatraemia), mixed foetal diseases (e.g. foetal alcohol syndrome, Down's syndrome, Patau syndrome, Turner's syndrome, Apert syndrome and Tay -Sachs disease), excretory diseases (e.g. urinary incontinence, urinary tract infections or renal disorders), neural or sensory disease (e.g. Alzheimer's disease, Parkinson's disease, cerebral malaria, meningitis, cerebellar ataxia, attention deficit disorder, autism and obsessive compulsive disorder), respiratory disease (e.g. emphysema, lung cancer and occupational lung disease), endocrine diseases (e.g. diabetes, Addison's disease and glomerulonephritis), digestive diseases (e g portal hypertension, irritable bowel disease, gastric atrophy or pancreatitis),and connective tissue or epithelial diseases (e.g. Crohn's disease, scleroderma, wound healing and epidermolysis bullosa).

By "fraction" is meant a portion of a separation. A fraction may correspond to a volume of liquid obtained during a defined time interval, for example, as in LC (liquid chromatography). A fraction may also correspond to a spatial location in a separation such as a band in a separation of a biomolecule facilitated by gel electrophoresis.

By "peptide," "polypeptide," "polypeptide fragment," or "protein" is meant any chain of more than two amino acids, regardless of post-translational modification (e.g., glycosylation or phosphorylation), constituting all or part of a naturally occurring polypeptide or peptide, or constituting a non-naturally occurring polypeptide or peptide. In addition, proteins comprising multiple polypeptide subunits (e.g., insulin receptor, cytochrome b/cl complex, and πbosomes) or other components (for example, an RNA molecule) will also be understood to be included within the meaning of "protein" as used herein Similarly, fragments of proteins and polypeptides are also within the scope of the invention and may be referred to herein as "proteins," "polypeptides," or "peptides," "tryptic peptides", or

"cleavage fragments." A "digestion fragment" is a portion of a polypeptide produced, at least theoretically, by the action of a protease that reproducibly cleaves the polypeptide. A "source polypeptide" for a digestion fragment is a polypeptide from which a specified digestion fragment is at least theoretically produced by the action of a protease that reproducibly cleaves the source polypeptide. A source polypeptide contains at least two digestion fragments. By "peptide reversion" or "protein reversion" is meant the phenomenon by which a disease-related peptide or peptides, peptide clusters, a peptide expression profile, or peptide expression profiles, return to control or reference levels upon treatment (either through therapeutic treatment, spontaneous remission or otherwise). Such peptide based measures of changes in protein expression related to a biological or physiological response to a treatment can be either continuous or discreet. Continuous, m that the degree of peptide reversion is a continuum that can be monitored by analysis 100's of distinct peptides (a plurality of peptides) present in biological samples such that the measures obtained accurately and comprehensively represent the biological, physiological and temporal response to a treatment or treatments under study. Discrete, in that peptides may be detected or not-detected in sample sets or conditions under comparison (i e. responder/non-responder). In determining the percentage of peptide reversion, which is significant for a given comparison of conditions, peptides may be weighted by importance Furthermore, the relative importance of a peptide reversion event may be derived from prior knowledge of the disease under study (for example disease etiology or treatment response) or by correlating peptide reversion with another indicator of response. Peptide reversion measures may be expressed as a percentage of reversion (linear) or as a non-linear function (e.g. sigmoid function) or any appropriate distance function.

By "record" is meant all of the information provided for a polypeptide, e g., digestion fragment, in a database. A record includes all fields for the polypeptide.

By "reference subject" is meant a subject whose condition is known and can be used as a comparison in the methods of the invention. By a "control" or "normal" reference is meant a subject whose condition is healthy and not suffeπng from any identifiable ailment or disease, or is not suffering or at πsk for a specific disease for which comparison is desired A reference can also be a subject who has a disease and has been treated with a compound that is known to be effective or not effective for the treatment of that disease. In another example, a reference subject is a subject who has a disease and has been treated with a compound that is known to have or to not have toxic side-effects when used for the treatment of that disease The skilled artisan will understand which reference subject is desired for each of the embodiments of the invention.

By "reference level" or "reference value" is meant a particular level of an identified peptide (e.g., a disease-related or disease-independent peptide) used as a benchmark for assessment, which may come from a single data point or be deπved from multiple data points, such as a cut-off median, and may be measured directly, indirectly, or calculated Typically the reference level will be used as a reference to a normal or control level allowing the identification of levels that deviate from the normal. An algorithm can be designed, such as by those with skill in the art of statistical analyses, which will allow the user to quickly calculate a reference level for use in making predictions or monitoring a particular state or condition. The algorithm and reference level can be used to generate a device that will allow the end user to input levels for a characteristic and quickly and easily determine the status or risk index of an individual through comparison of the level that was input and the reference level. Similarly, it is possible to provide a device that indicates the status of an individual relative to a reference level. One skilled in the art can determine an appropriate reference level for use in the methods of the invention.

By "reference range" is meant a particular range of an indicator used as a benchmark for assessment, such as a mean deviation cut-off multiple points range within which, for example, "normal" or "disease" is expected to fall. The reference range can be a range of test values expected for a designated population of individuals, e.g., 95% of individuals that are presumed to be healthy (or normal). A reference range may be useful in minimizing variation possible with a single reference sample. Generally, all reference ranges include a set of two values with one value designated as an upper reference range limit and another designated as a lower reference range limit. A range may be sub-divided into ranges of differing significance, hence where within a range a value falls may provide additional correlates or probabilities. For example, a range for normal expression of a peptide is 0.1 to 0.4 μg/1 of plasma, and above the reference level of 0.4 μg/1 a disease state is indicated, however, within the normal range a range of 0.3 to 0.4 μg/1 may indicate an 80% probability of disease or a stage early in the progression towards disease (e.g., mild cognitive impairment). An algorithm can be designed, such as by those with skill in the art of statistical analyses, which will allow the user to quickly calculate a reference range for use in making predictions or monitoring a particular state or condition. The algorithm and reference range can be used to generate a device that will allow the end user to input levels for a characteristic and quickly and easily determine the status or risk index of an individual through comparison of the level that was input and the reference range. Similarly, it is possible to provide a device that indicates the status of an individual relative to a reference range. One skilled in the art can determine an appropriate reference range for use in the methods of the invention.

By "study peptide" is meant a distinct peptide that is consistently detected across samples analyzed as part of a study. A peptide that is consistently and independently detected across a number of samples of the same or different sample sets is grouped as and refered to as a "study peptide." Typically, between 20,000 and 30,000 study peptides will be detected for a plasma study undertaken in rats, mice or humans. Typically, approximately 3,500 proteins will be detected in human urine. By "study' is meant the analysis and comparison of a number of sample sets each representing a particular condition, treatment group, disease state, or normal state (control). Studies could be designed as either longitudinal studies or cross-sectional studies, hi longitudinal studies each sample set is comprised of samples obtained from the same subjects for example a pre-disease healthy sample, samples collected during the onset of disease, during various stages of the disease and during treatment of the disease are collected from each subject. In cross-sectional studies samples are collected from distinct populations of healthy, diseased and disease treated subjects.

By "sample set" is meant a group of biological samples of the same type and collected under the same conditions from either the same subject or from different subjects. Each sample is analyzed independently as part of a study. Sample sets representing different biological conditions or states are compared to determine "peptide reversion". By "subject" is meant a mammal, including, but not limited to, a human or non- human mammal, such as a mouse, cat, dog, bovine, equme, or ovme.

By "substantially identical" is meant at least 51%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99% or more identical. In the methods of the invention, substantial identity can refer to the expression levels for a peptide in one sample compared to the expression levels for the same peptide in another sample. In another example, substantial identity refers to the percent identity between the peptide expression profile of one sample and another.

By "substantial percentage" is meant at least 51%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99% or more. By "treating" is meant administering a compound or a pharmaceutical composition for therapeutic purposes. To "treat disease" or use for "therapeutic treatment" refers to administering treatment to a subject already suffering from a disease to improve the subject's condition. To "prevent disease" refers to prophylactic treatment of a subject who is not yet ill, but who is susceptible to, or otherwise at πsk of, developing a particular disease A drug or compound is considered "therapeutically effective" if a substantial percentage of the disease-related peptides revert to substantially the same level of expression as found in the normal reference sample. In general, a drug is considered highly effective if at least about 80% of the disease-related peptides revert to substantially the same level of expression as found in the normal reference sample; moderately effective if between about 50% and 80% of the disease-related peptides revert to substantially the same level of expression as found in the normal reference sample; and weakly effective if between about 20% and 50% of the disease- related peptides revert to substantially the same level of expression as found in the normal reference sample. The determination of therapeutic effectiveness can also be determined using a relative comparison. For example, using the methods descπbed herein, the extent of peptide reversion for a disease-related peptide, peptides, or peptide cluster from disease to normal can be compared for two drug compounds. The drug compound that produces the relatively greater peptide reversion is considered to be therapeutically effective relative to the drug compound that produces the lesser peptide reversion.

By an "undeπvatized" biomolecule or fragment thereof is meant a biomolecule or fragment thereof that has not been chemically altered from its natural state. Deπvitization may occur during non-natural synthesis or during later handling or processing of a biomolecule or fragment thereof.

By an "unlabeled" biomolecule or fragment thereof is meant a biomolecule or fragment thereof that has not been deπvatized with an exogenous label (e.g., an isotopic label or radiolabel) that causes the biomolecule or fragment thereof to have different physicochemical properties to naturally synthesized biomolecules

Other features and advantages of the invention will be apparent from the following description and claims.

Brief Description of the Drawings FIGURE l is a schematic depiction of individual peptide or peptide cluster changes m control, disease, and disease-treated populations where the position of the points represent the relative abundance levels of a peptide or peptide cluster and the line represents the median abundance. The change in peptide expression level seen in disease reverts back to normal upon treatment with a drug compound. FIGURE 2 is a heat map representing the major plasma proteome changes observed between CFA-treated or diseased animals, control animals (untreated healthy animals), and diseased animals treated with ERB-041 (small-molecule ERβ selective agonist) and Drug B. Each vertical column represents the data acquired for one peptide (out of a total of 711). The MS signal intensity (abundance) of each peptide is represented m gray scale, with white indicating the highest intensity and black indicating the lowest). The peptides were ordered according to hierarchical clustering using Euclidean distances Box A indicates a region of the map where these data support a small number of off-target effects with administration of Drug B; changes m peptide intensities (protein expression) not observed in the disease state (CFA treated) or in the control. Box B indicates a region of the map where the data show that ERB-041 is more effective than Drug B at reversing down regulation of protein expression observed in the disease state. Box C indicates a region of the map where these data support a small number of off target effect with administration of ERB-041; changes in peptide intensities (protein expression) not observed m the disease state (CFA treated) or in the control. Box D indicates a region of the map where these data show that the majority of peptides abundant in the disease state, corresponding to up-regulated proteins, are reverted by treatment with either ERB-041 or Drug B. FIGURE 3 is a graph showing that ERB-041 reduces joint swelling in a rat model. The joint score determined by redness and swelling of hind-paws are plotted against the days of treatment with ERB-041 or vehicle control.

FIGURE 4A is a multidimensional scaling plot of 13 mild cognitive impairment (MCI) plasma profiles (cluster at right of plot) versus 17 age-matched controls, normal (cluster at left of plot). FIGURE 4B is a heat map illustrating 662 peptides up-regulated in MCI (lower panel) versus normal controls (upper panel). Here the gray scale is proportional to signal intensity or peptide abundance with white indicating the highest intensity and black indicating the lowest). Differentially expressed peptides were selected from each sample set based on the following parameters: peptides which displayed at >3 fold differential abundance versus the reference sample set, a p value of <= 0.05 (in at least 50% of the samples analyzed).

FIGURE 5 A is a multidimensional scaling plot of 18 Alzheimer's disease (AD) plasma profiles (cluster at πght of plot) versus 17 age-matched controls, normal (cluster at left of plot). FIGURE 5B is a heat map illustrating 331 peptides which were down-regulated in AD (upper panel) as compared to normal controls (lower panel). Differentially expressed peptides were selected from each sample set based on the following parameters: peptides which displayed at >3 fold differential abundance versus the reference sample set, a p value of <= 0.05 (in at least 50% of the samples analyzed). FIGURE 6 is a multidimensional scaling plot where 5 untreated AD patient plasma profiles were compared to those of 13 MCI and 17 age-matched controls (Normal). The plot shows three distinct clusters for Normal samples (bottom left cluster), MCI samples (central cluster) and AD samples (top πght cluster) suggesting not only a distinct plasma protein profile for each but also a progression from MCI to AD FIGURE 7 is a heat map illustrating 121 peptides which show a progression from age -matched controls (normal: upper panel) to MCI (middle panel) to AD (lower panel) either through peptides becoming increasingly down-regulated (as in approximately peptides 63- 100) or increasingly up-regulated (as in approximately peptides 108-121) from a mild condition to a more advanced chronic neurodegenerative one. FIGURE 8 is a multidimensional scaling plot of 5 untreated AD patient plasma profiles (upper left cluster), 17 age-matched controls (Normal, lower right cluster) and 13 AD patients being treated with acetylcholinesterase inhibitors (lower πght cluster).

FIGURE 9 is a heat map representation of the same 5 untreated AD patient plasma profiles (middle panel), 17 age-matched normal controls (upper panel) and 13 AD patients being treated with acetylcholinesterase inhibitors described for FIGURE 8. 383 peptides were selected on the basis of their being expressed at >1.5 fold differential abundance and p<0.05 among sample sets. This heat map now combines up- and down-regulated peptides which are expressed m red and blue respectively (as descπbed for FIGURE 2).

FIGURE 1OA is a multidimensional scaling plot of 13 MCI plasma profiles versus 18 AD samples showing the clustering of the two sample sets into distinct sample sets. FIGURE 1OB is a heat map illustrating 442 peptides which were >2-fold up-regulated in AD as compared to MCI and exhibited a significance of P<0.05.

Detailed Description of the Invention

We have discovered methods for using differentially expressed peptides whose expression differs in normal versus diseased versus treated diseased conditions to evaluate and predict drug compound efficacy, toxicity, and pharmacodynamic effects. Such methods can be used as objective, quantifiable measurements in a variety of clinical or research decisions including identification of lead compounds, comparison of therapeutic efficacy for two or more candidate compounds, comparison of toxicity for two or more candidate compounds, determination of optimal dosages for a drug compound, determination of the mechanism of action for a particular drug compound, and the identification of specific proteins that can be used as candidate biomarkers for diagnosis of disease, for measuring therapeutic efficacy, and for measuring toxicity of a drug compound. The methods can also be used for determining the health status of an individual, for monitoring the treatment or progression towards a disease state for an individual undergoing treatment for a disease or at risk for developing a disease. In addition, the methods of the invention can be used to identify a responder or a non-responder patient population that can be used to design clinical trials for particular drug compounds or to select appropriate animal models of disease for preclinical studies by comparing the protein profile of vaπous animal models to those of human disease.

In the methods of the invention, differentially expressed peptides that are expressed at different levels in one sample versus another are identified in biological samples from vaπous subject populations (e.g., untreated diseased, treated diseases, and normal subjects). Such differentially expressed peptides are considered population biomarkers. For example, population biomarkers can be identified as differentially expressed in normal versus untreated disease samples. Expression of the same population biomarkers in the treated disease sample can then be determined and compared to the normal and the untreated disease samples to determine if the differentially expressed population peptides revert to normal expression levels upon treatment with a candidate compound. This reversion phenomenon is known as "peptide reversion." The methods of the invention can also be used to screen candidate drug compounds for their ability to induce peptide reversion in a population without causing the differential expression of toxicity markers or "disease-independent" peptides. In another example, a diagnostic peptide expression profile for an "at risk" patient can be generated using the methods of the invention and compared to the peptide expression profile generated for one or more samples from diseased individuals and samples from normal control individuals. In this example, if the at πsk patient's profile is substantially identical to the peptide expression profile for the samples from diseased individuals, the patient can be tested further or monitored for additional disease symptoms. Such assays are particular useful for the diagnosis of diseases, such as Alzheimer's disease, which are otherwise difficult to diagnose.

In one preferred embodiment of the methods of the invention, plasma samples are collected from a human control sample, a disease sample, and a treated disease sample

Human samples are depleted of albumin IgG, IgA, transferrin, haptoglobin and anti-trypsin using standard immunodepletion techniques. Ahquots for each of the depleted plasma samples are then separated by SDS-PAGE or 2-dimensional liquid chromatography (2D-LC) Gels are then cut into 24 bands and placed into 96-well plates for oxidation and tryppsm digestion. Peptides are extracted from the gel bands and vacuum dried. Peptides are then resolubihzed in 10% acetonitrile, 0.2% TFA and combined into ten pools

Each pool is analyzed by LC-MS, preferably nano LC-MS (Waters capLC coupled to a Q-Tof Ultima). The detected peptides are characterized by their mass/charge ratio (m/z), retention time on the liquid chromatography column, their charge and intensity. The peptides that fall within comparable m/z and retention time windows across all samples are designated as "study peptides." Study peptides are the peptides that are consistently detected across all samples within the study and as such represent a specific type of data after normalization A typical set of study peptides ranges from 20,000 to 30,000 for a plasma study undertaken in rats, mice or humans or approximately 3,500 proteins for a study of human urine. Study peptides are matched across samples using Mass Intensity Profiling System (descπbed in U.S Patent Application Publication No. 2003/0129760; hereinafter referred to as "MIPS") and Constellation Mapping (described in PCT publication number WO 2004/049385; hereinafter referred to as "Constellation Mapping") to identify the differentially expressed peptides based on differences in their abundance and intensity. For each individual differentially expressed peptide it is determined, with statistical significance (e.g., using a t-test or other), if the peptide or peptide cluster separates the disease sample set from the control sample set based on the central tendency of the intensity distributions in the cluster. Peptides which separate the disease sample set from the control sample set with a statistical significance below 0.05 are selected as disease-related peptides. In addition, in the examples below, we have used an additional "diagnostic criterion" which requires that there is an intensity value T such that no more than 30% of the patients in the up condition have intensity below T and no more than 30% of the patients in the down condition are above T. Thus, this means that the peptide provides a biomarker that has sensitivity and specificity of 70% and 70%. The T value can be determined for each sample (e.g., 30% or lower, 20% or lower, 10% or lower, or 5% or lower), however, the stricter the cπteπon, the fewer the peptides that will satisfy it Peptides that are differentially expressed and subjected to the diagnostic cπteπon, known as "diagnostically differentially expressed peptides," can be grouped into a peptide cluster (e.g., at least 50, 100, 150, 200, 300, 400, 500, 1000, 2000, 5000, or 10000 or more peptides, and/or source polypeptides) that can be used to distinguish between the normal and disease state in this example. It will be understood by the skilled artisans that the methods descπbed above can be adapted to distinguish between any two or more states, including but not limited to treated/untreated; normal/diseased; normal/disease state I/disease state2; normal/disease stage I/disease stage 2; and normal/disease untreated/disease treated/disease treatment 2 (toxic).

By observing the differences m the diagnostically differentially expressed peptides or diagnostically differentially expressed peptide cluster, a "peptide expression profile" may be determined between and among conditions and/or states. Diagnostically differentially expressed peptides may be used to compare states and conditions and determine the overall relative differences between peptide expression profiles for one or more states. For example, disease state X (untreated) may have 100 digestion fragments (peptides) that are diagnostically differentially expressed as compared to controls while the drug-treated disease state Y has only 20 digestion fragments (peptides) that are diagnostically differentially expressed from normal, which in this case are shared with disease state X (untreated), and 80 which do not differ from normal but that are diagnostically differentially expressed from disease state X. In this example, the peptide reversion rate for the treated condition is 80% since 80 out a total of 100 disease-related peptides revert to control levels upon drug treatment.

By observing the relative reversion rate, or ability of one compound versus another to revert disease-related peptides back to normal levels, one can predict which compound will be more efficacious before observing any phenotypic changes or clinical endpomts indicative of efficacy. The compound with the highest peptide reversion rate is predicted to be the most efficacious.

Most drug treatments also induce plasma peptide changes on their own, that is to say diagnostically differentially expressed peptides are observed in drug treated, disease samples which differ from disease (untreated) samples and differ from controls. These are called "disease-independent peptides" which are related to off-target effects and potential toxic side- effects. Results of comparisons of diagnostically differentially expressed peptide profiles can also be used to rank or select states and or conditions. For example, comparing profiles for samples of unknown disease stage with reference samples of known disease stages, the unknown samples may be grouped and ranked by stage. In another example, a comparison of profiles for a disease state versus normal and a number of possible treatment conditions of the disease, such as different treatments or different treatment combinations may permit the selection of the best compounds, candidate treatments for further study and their ranking based on relative rates of reversion to normal, or relative rates of reversion to normal of individual population biomarkers uniquely diagnostically differentially expressed in the disease state versus normal as compared to a second disease state. In another example, comparison of treatment profiles with profiles of toxic treatments or a toxic state may permit the selection for elimination of certain treatments for consideration based on common diagnostic differential expression between individual peptides in a toxic profile and a treatment profile. Ranking and selection may also take into account factors independent of the profile, for example, the best associated percent reversion to normal may be used in combination with the best mode of delivery of the compound to rank treatment compounds The following depiction of peptide groupings in decreasing size is not intended to limit the invention in any way, but is provided to help illustrate the invention.

Sample set > Study peptides > Differentially expressed peptides

(Disease-related peptides/Disease-independent peptides) > Diagnostically differentially expressed peptides.

For the term differentially expressed peptides and diagnostically differentially expressed peptides, the term peptide encompasses single peptides or a cluster of multiple distinct peptides.

The methods of the invention and exemplary uses for the methods are described in detail below.

Methods for Analysis of Proteins Sample Preparation

Samples useful in the methods descπbed herein include any biological sample (e.g , a biological fluid such as blood, plasma, serum, urine, bile, cerebrospinal fluid, aqueous or vitreous humor). Alternatively, a biological sample can be obtained from any organ or tissue. Desirably, the sample is a blood or plasma sample, which is filtered to remove particulate matter. Plasma collection and depletion

For plasma analyses, blood is collected in PlOO Vacutamers (Becton Dickinson) that contain EDTA and protease inhibitors. These tubes were specifically designed for proteomic analysis (Hulmes et al , Clin Prot 1 : 017-032, (2004)) and are the industry standard Plasma is deπved from the blood samples by a well-documented protocol and frozen for shipping. Frozen samples are barcoded and entered into the LIMS system and stored upon arrival at Capπon. Samples are thawed and depleted of high abundance, using the commercially available Agilent® Multiple Affinity Removal System™ (MARS) on a high performance liquid chromatograph (HPLC) The MARS removes greater than 98% of six interfering high-abundant proteins (albumin, IgG, IgA, transferrin, haptoglobin, and antitrypsin) from human plasma or three proteins (albumin, IgG, and transferrin) from rodent plasma The MARS has been widely studied and implemented and has resulted in numerous publications (e.g. Bjorhall et al., Proteomics 5:307-317 (2005)). In much the same way that the BDPlOO Vacutamer provides a reliable and GMP -rated product, the Agilent MARS provides the same advantages Digestion to peptides

Following plasma depletion, the proteins are proteolyzed under denaturing conditions with LysC and then diluted and proteolyzed with trypsin. The use of denaturing conditions was pioneered by Yates and co-workers (Wu et al., J Am Soc Mass Spectrom. 16- 1231- 1238, (2005)), where it was shown that denatured proteins proteolyze more completely and reproducibly than non-denatured proteins. This process ensures a complete endpomt digestion and generates high-quality and reproducible samples.

Digestion is necessary to facilitate chromatographic and mass spectrometry analysis, but it also provides for a more πgorous quantitative and qualitative result. The process of proteolysis converts each protein into an array of peptides, typically on the order of one tryptic peptide per kiloDalton of protein mass All of these peptides would be expected to move in unison when measured through the expression profiling process. Therefore, the process of proteolysis creates a multiplicity of data points for the measurement of each protein, and these minimize the false positive rate. For example, in a cataloging experiment, a given peptide must be searched against an enormous peptide database and judged to be true or false on the basis of the MS/MS data alone However, in protein expression profiling, all of the peptides to a given protein move in unison, making the determination true positives much easier.

Peptide fractionation and analysis by mass spectrometry SCX Fractionation: Following proteolysis, the peptides are desalted and then subjected to HPLC via strong cation exchange (SCX) chromatography. The same column, gradient, and flow are used for all samples, regardless of origin (organelle or blood plasma) This standardized process provides for a very reproducible means of separation, and one which can be readily monitored for quality control. Not only does the SCX fractionation serve to reduce sample complexity, it also improves the lower limit of detection (LLOD), by allowing increased sample volumes to be subsequently loaded on the reversed phase column that is part of the LC-MS system. Loading more material increases the dynamic range of the analytical detection "system" and allows fora deeper and more comprehensive look into the proteome For example, a standard reversed phase column connected to the mass spectrometer (500 μm diam x 30 cm length) can accommodate approximately 40 μg of total protein before the chromatography begins to suffer ill effects caused by overloading This is the equivalent of 8 μL of depleted plasma. However, by using fractionation, the equivalent of 64 μL of plasma can be analyzed, because each of the eight fractions can contain 40 μg of digested total protein The difference is an eight-fold increased depth of penetration into the proteome.

Samples may also be separated into individual fractions using ID-SDS polyacrylamide gel electrophoresis (PAGE). Gel bands are then contacted with a protease (e g., trypsin). Optionally, each peptide fraction can be further fractionated, (e.g., using strong cation exchange chromatography) Peptides are extracted from the gel bands, vacuum dried, and then resolubihzed and combined into ten pools.

Each pool is analyzed by mass spectrometry using standard methods such as nano- liquid chromatography - mass spectrometry (LC-MS). Optionally, the sample is normalized prior to LC-MS to eliminate the hydration effects and lyophihzed to increase stability. In addition, samples can be randomized to remove instrument and day of the week effects

Using mass spectrometry, peptides for each sample are characterized by mass/charge ratio, retention time on the liquid chromatography column, charge, and intensity The peptides that fall within comparable m/z and retention time windows across all samples are designated as study peptides.

The peptide intensities for the study peptides for each sample set are then compared using the methods described below. Data Analysis Evaluating Differences

The significance of changes between peptide expression profiles, in light of signal variability withm profile sets, can be determined, for example, using standard statistical techniques. Non-hmitmg examples of standard statistical techniques include those descπbed in MIPS (described in U.S. Patent Application Publication No. 2003/0129760), Constellation Mapping (descπbed in PCT publication number WO 2004/049385), and U.S. Patent No

6,906,320. The statistical analysis (using both parametric and nonparametπc techniques) is used to estimate the significance of difference in peptide profiles from different sample sets and is expressed as a monotonic function of p-value. A p-value represents the likelihood that an observed change between a peptide intensity within the distribution of intensities associated with a particular protein could have arisen by chance in the absence of differences between the sample sets in the level of some protein or biomolecule. Ranks that come from statistical measures of the ability to correctly classify samples can also be used in combination with or in place of p-values.

In one example, LC-MS peptide data is analyzed using computer implemented methods for comparing protein abundance which generally includes the steps of inputting mass spectrometry data, centroidmg and reducing the noise, producing isotope maps, detecting and centering peptides, producing peptide maps, and aligning peptide maps, thereby allowing the determination of differential abundance of biomolecules in the samples. Examples of preferred computer implemented methods for the analysis of the peptide data include MIPS (descπbed in U.S. Patent Application Publication No. 2003/0129760) and Constellation Mapping (described in PCT publication number WO 2004/049385) methods or software for the tracking of peptides across samples and the quantitative measurement of differentially expressed peptides both by differential intensity and differential abundance.

MIPS is based on the analysis of data from mass spectrometry and is used to calculate absolute or relative peptide abundances by pair-wise comparisons of mass intensity profiles of individual biomolecules whose presence, absence, increased or decreased expression, or pattern of expression is associated with a disease or a condition of interest. MIPS is used to provide a map of peptides that indicates their intensity and their mass/charge and retention time

Intensity maps derived from MIPS analysis are then further aligned between samples using Constellation Mapping software. Constellation Mapping software can be used to identify differentially abundant biomolecules in different samples.

For each individual differentially expressed peptide it is determined, with statistical significance (e.g., using a t-test or other), if the peptide or peptide cluster separates the disease sample set from the control sample set based on the central tendency of the intensity distributions m the cluster. Peptides which separate the disease sample set from the control sample set with a statistical significance below 0.05 are selected as disease-related peptides. In addition, in the examples below, we have used an additional "diagnostic criterion" which requires that there is an intensity value T such that no more than 30% of the patients in the up condition have intensity below T and no more than 30% of the patients in the down condition are above T. Thus, this means that the peptide provides a biomarker that has sensitivity and specificity of 70% and 70%. The T value can be determined for each sample (e.g., 30% or lower, 20% or lower, 10% or lower, or 5% or lower), however, the stricter the cπteπon, the fewer the peptides that will satisfy it. Peptides that are differentially expressed and subjected to the diagnostic cπteπon are known as "diagnostically differentially expressed peptides," and can be grouped into a peptide cluster (e.g., at least 50, 100, 150, 200, 300, 400, 500, 1000, 2000, 5000, or 10000 or more peptides, and/or source polypeptides) that can be used to distinguish between one sample set (e.g., a control group) from another (e.g., a disease group). By observing the differences in the diagnostically differentially expressed peptides or diagnostically differentially expressed peptide cluster, a "peptide expression profile" may be determined between and among conditions and/or states. Diagnostically differentially expressed peptides may be used to compare states and conditions and determine the overall relative differences between peptide expression profiles for one or more states. The degree of differential abundance used to distinguish groups is based upon known statistical methods and determined empirically for each disease type. For example, in a disease such as inflammation, where a relatively large proportion of total peptides analyzed are differentially expressed as compared to controls (approximately 10% of -35,000 peptides are over- expressed by more than 3 fold), it is possible to use a higher threshold of false positive and false negative cπteπa such as 99% sensitivity for inclusion of peptides withm a cluster. In other diseases, such as certain diseases of the central nervous system and pain, less than 1% of the total plasma peptides vary in the disease samples as compared to control samples. In this case, a much lower threshold of false positive and false negative (30%, 30%) is useful in selecting peptides for further comparison. Peptides which are significantly up or down regulated in disease with a statistical significance below 0.05 are deemed to be differentially expressed peptides that are useful in the methods of the invention.

Polypeptide Identification

If desired, peptides identified as differentially expressed or as population biomarkers for a disease using the methods descπbed herein can be identified as corresponding to a particular peptide or source polypeptide using techniques known in the art, for example tandem mass spectrometry (MS/MS), peptide sequencing, and Virtual Mass Spectrometry (VMS; as described in U.S. Application No. 11/424,736). Once a peptide is matched to a source polypeptide in a database and the identity of an individual polypeptide biomarker is discovered, standard immunoassays can be used to measure the level of the polypeptide biomarker in a patient for diagnostic or therapeutic monitoring purposes.

Uses of differentially expressed peptides

Once a set of differentially expressed peptides are identified that can be used to distinguish one sample set (e g., a diseased group) from another sample set (e.g , a control group), the peptides are useful for several methods that are all contemplated by the invention

In one method of the invention, the differentially expressed peptides are used to identify statistically significant differentially expressed peptides from control sample sets, untreated disease sample sets, and drug-treated sample sets. The sample sets can be compared individually or simultaneously. The comparisons can be used, for example, to monitor disease progression, determine drug efficacy (both relative to another drug compound and absolute efficacy), monitor and determine the toxicity or side-effects of a drug compound, determine optimum dosing of a drug compound, and to determine the pharmacodynamic effects of a drug compound. The comparisons can also be used to diagnose or monitor disease or predict the overall heath status in an individual, to predict progression towards a more advanced disease state, and to determine patient stratification for clinical trials. The methods can also be used to identify animal models of disease for which the peptide expression profile or the peptide reversion of specific disease-related peptides closely resemble that of the samples from human subjects having the disease.

Results of comparisons of differentially expressed peptides m various sample sets may also be used to rank or select states or conditions For example, comparing peptide expression profiles for samples from an individual with an unknown disease stage with reference samples from individuals with known disease stages, the unknown samples may be grouped and ranked by stage based on their identity to the samples from the known disease stages. In another example, a comparison of peptide expression profiles for a disease state versus control and a number of possible treatment conditions of the disease, such as different treatments or different treatment combinations, may permit the selection of candidate treatments for further study and their ranking based on relative rates of reversion of expression levels of disease-related peptides to control levels. Or, for example, comparison of treatment profiles for varying doses may permit the selection of the best dose by optimizing the dose which exhibits maximum peptide reversion and minimum induction of disease- independent changes. Ranking and selection may take into account factors independent of the profile, for example, the best associated percent peptide reversion to control may be used in combination with the best mode of delivery of a treatment compound to rank treatment compounds.

The comparisons can also be used to identify population biomarkers that are differentially expressed between any two or more states, for example, treated/untreated, wild- type/gene knockout, control/diseased, and control/disease state 1/ disease state 2. Of particular interest is the ability to identify population biomarkers that can predict a given outcome following drug treatment.

These uses are not intended to limit the invention in any way and are descπbed in detail below. Determining drug efficacy for a drug compound

Once a set of disease-related peptides that are differentially expressed are identified and can be used to distinguish one sample set (e.g., a disease group) from another sample set (e.g., a normal reference group), a similar process is used to compare the expression of the disease-related peptides in samples from the untreated disease sample set and the treated disease sample set to make relative predictions about drug efficacy based on the relative number of disease-related peptides that revert to the levels of expression found in the normal reference sample set. This phenomenon is diagrammed in Figure 1. A drug is considered therapeutically effective if a substantial percentage of the disease-related peptides revert to substantially the same level of expression as found in the normal reference sample In general, a drug is considered highly effective if at least about 80% of the disease-related peptides revert to substantially the same level of expression as found in the normal reference sample; moderately effective if between about 50% and 80% of the disease-related peptides revert to substantially the same level of expression as found m the normal reference sample, and weakly effective if between about 20% and 50% of the disease-related peptides revert to substantially the same level of expression as found in the normal reference sample. In addition, relative peptide reversion can also be determined in order to identify one drug compound that is more effective than another drug compound. In this example, a drug that reverts more than 80% of the peptides back to normal levels may be deemed more efficacious than a drug that reverts less than 80% of disease-related peptides back to normal levels

The methods descπbed can be used to assay biological samples, preferably blood or plasma samples, from subjects diagnosed with or at risk for virtually any disease or condition Disease particularly amenable to the methods of the invention include cancer (e.g. leukemia, colon, bone, breast, liver and lung cancer), immune or haematopoietic diseases (e.g anemia, Hodgkm's disease, acute lymphocytic anemia, multiple myeloma, arthritis, asthma, AIDS, autoimmune disease, inflammatory bowel disease, psoriasis and Lyme disease), reproductive system disorders (e.g. prostatitis, inguinal hernia, varicocele, penile carcinoma, ovarian adenocarcinoma and Sertoh-leydig tumours), musculoskeletal diseases (e.g. giant cell tumours, Paget's disease, systemic lupus erythematosus, gout, muscular dystrophy and cachexia), cardiovascular diseases (e.g. rhabdomyomas, heart disease, arrhythmia, cardiac arrest, heat valve disease, hypernatraemia and hyponatraemia), mixed foetal diseases (e.g. fetal alcohol syndrome, Down's syndrome, Patau syndrome, Turner's syndrome, Apert syndrome and Tay -Sachs disease), excretory diseases (e.g. urinary incontinence, urinary tract infections or renal disorders), neural or sensory disease (e.g. Alzheimer's disease, Parkinson's disease, cerebral malaria, meningitis, cerebellar ataxia, attention deficit disorder, autism and obsessive compulsive disorder), respiratory diseases (e.g. emphysema, lung cancer and occupational lung disease), endocrine diseases (e.g. diabetes, Addison's disease and glomerulonephritis), digestive diseases (e.g. portal hypertension, irritable bowel disease, gastric atrophy or pancreatitis),and connective tissue or epithelial diseases (e.g. Crohn's disease, scleroderma, wound healing and epidermolysis bullosa). One specific peptide reversion assay used for Alzheimer's disease (AD) is descπbed in Example 3. In this example, 331 proteins were overexpressed in AD as compared to normal controls and 574 proteins were under-expressed in AD samples as compared to normal controls. Using these data, candidate drug compounds can then be analyzed to determine which drug is effective at reverting at least 50%, preferably 60%, 70%, 80%, 90%, 95%, 96%, 97%, 98%, 99% or more of the disease-related peptides that are differentially expressed in AD samples back to control levels.

Another example of a peptide reversion assay used for rheumatoid arthritis (RA) is described in Example 2. In this example, 462 peptides were found to be overexpressed in RA as compared to normal controls and 249 peptides were found to be under-expressed m RA samples as compared to normal controls. Further, in this example, two specific drug candidates were analyzed (ERB-041 and Drug B) and both drugs showed reversion of up- regulated polypeptides back down to normal levels. ERB-041 proved to be more effective at reversing polypeptides that were downregulated in the RA disease state. This example demonstrates the use of the methods to determine relative drug efficacy of one compound compared to another drug compound. Additional candidate drug compounds can also be analyzed to determine which drug is effective at reverting at least 50%, preferably 60%, 70%, 80%, 90%, 95%, 96%, 97%, 98%, 99% or more of the disease-related peptides that are differentially expressed (either up-regulated or down-regulated) in RA samples back to control levels.

Determining the toxicity or disease independent changes induced by a drug compound

Disease-independent drug-induced changes may also be observed using the peptide reversion methods of the invention. In this example, a group of peptides may not show differential expression between normal and diseased populations but, once a drug compound is administered, the proteins are differentially expressed indicating that the drug is affecting polypeptides not related to the diseases stated, also known as "off-target" or "disease- independent peptides." Such information is very useful for understanding the potential toxicity or unwanted side-effects of a drug compound. For example, a very effective drug may revert 80% of disease-related peptides back to normal levels but may also induce the differential expression of several hundred peptides which are unrelated to disease. If these peptides are not differentially expressed after treatment with a second equally effective drug, then the second drug is deemed to be more effective. Alternatively, if a very effective drug reverts 80% of disease-related peptides back to normal levels but may also induce the differential expression of several hundred peptides which are unrelated to disease, the disease- unrelated peptides may be sequenced (by re-mjecting appropriate samples) through LC-MS- MS or VMS and the specific proteins induced by the drugs identified and evaluated for their potential to induce toxicity through inferences deπved from their known biological function In this example, given two analogues of a drug which exhibit equivalent efficacy based on peptide reversion, it is useful to examine the disease-independent changes induced by each. If, for example, one compound induces the expression of series of proteins that, once identified, are known to be involved in the iron metabolism pathway and the other does not, it is desirable to choose the compound which has the least disease-independent changes and where these changes do not affect proteins known to be a source of potential toxicity. In addition, these methods can be used to determine a dosage of a drug compound for which the toxic side-effects reach unacceptable levels. In this example, the peptide reversion of disease- related peptides and the presence of disease-independent peptides are determined for a compound at varying dosages. The optimum dosage is chosen based on the maximal peptide reversion of disease-related peptides achieved with a minimal presence of disease- independent peptides

Toxicity can also be determined within a sub-population of patients. For example, a drug may be very effective but may show liver toxicity within a sub-population of patients treated with the drug. Peptide reversion methods can be used to determine if the sub- population of patients show a polypeptide differential expression profile that is distinct from the remainder of the patients. Such information is useful for predicting outcomes following drug treatment. The methods of the invention can be used to compare patient samples immediately following drug treatment and before the appearance of toxicity to identify differentially expressed peptides which may predict which patients are susceptible to exhibiting toxicity and which are not pπor to the onset of symptoms In addition, the "disease-independent" peptides that are differentially expressed can be identified using mass spectrometry or VMS and the identity of the proteins will shed insight into the molecular basis for the difference in toxicity. For example, in the case of liver toxicity, one population may have very low CYP3A levels in plasma versus another. An understanding of the "disease-independent" peptides that are affected by a particular drug can provide information as to the toxic side-effects of a drug and may suggest additional drugs which can be administered simultaneously to reduce the toxic side-effects or indicate which patient sub- populations should not be given a particular drug. For example, if a drug is known to produce "off-target" effects on the estrogen-signaling pathway, such a drug may not be given to a woman having or at πsk for having breast or ovarian cancer.

Determining the effective dosing of a drug compound

Differential expression profiles can also be used to rank or select dosages for a given compound that produces the most peptide reversion in combination with the least disease- independent effects. For example, peptide differential expression profiles may be determined for a normal and an untreated disease group for a given disease. These profiles are then compared to the treated population using varying doses of the drug compound (each dose having its own profile for comparison). Treated disease patient profiles are then compared to the untreated and the normal at each dose to determine which dose produces maximum peptide reversion and minimum induction of disease-independent changes. Such a dose is then selected as the optimal dose for the use of the drug compound to treat the disease tested Ranking and selection may take into account factors independent of the profile, for example, the best associated percent reversion to control may be used in combination with the best mode of delivery of a treatment compound to rank treatment compounds.

Monitoring the health of an individual/Predicting progression towards a more advanced disease state

The differential expression methods descπbed herein can also be used to monitor the health of an individual or to predict disease progression or staging. In this example, the methods of the invention are used to determine the differential expression profile of a patient at risk for a given disease or known to have a mild form or precursor of a disease (e.g., mild cognitive impairment (MCI) as a precursor to AD). The patient sample is compared to a population of control samples that are known to be healthy and free of disease and a population of samples known to have the disease or a later stage of the disease (e.g., fullblown AD) and the patient profile can be ranked m terms of differential expression of the peptides relative to both the normal and the diseased patient populations. For example, as shown in Figures 5 and 6 and described in Example 3, the group of patients suffering from MCI showed a peptide expression profile that was in between that of normal controls and untreated AD. In other words, for the peptides that were differentially expressed (up- regulated in this case) between normal and AD groups, their levels in the MCI sample set tended to be greater than the normal sample set but less than the full-blown AD sample set A patient known to have MCI could then be monitored regularly for progression of the disease based on the changes in disease-related peptide to more closely resemble the AD population

Facilitating clinical and research decisions for drug compounds

As described above, the methods of the invention can be used to make a number of clinical and research decisions regarding one or more drug compounds. The methods of the invention can be used to obtain information for every step of the discovery process beginning with the identification of differentially expressed polypeptides that can be used m drug target screens. For example, if a peptide is overexpressed in AD samples, the peptide can be sequenced and its identity determined based on comparisons to a database of peptides and source polypeptides. Once identified, the peptide can be used in a standard drug compound library screen to identify potential candidate compounds that downregulate the target.

Appropriate animal models for any disease can also be identified using the methods of the invention. For example, the methods of the invention can be used to identify differentially expressed peptides in the plasma of normal and diabetic humans. Plasma samples from normal animals and species matched animal models of diabetes can also be assayed and the differential expression profiles for the animals compared to the human samples. The animal model peptide profile that shows substantial identity to the plasma peptide expression profile of the human samples is then chosen as the appropriate animal model for diabetes.

Candidate drug compounds can be tested in the animal models descnbed above to determine both efficacy and toxicity or disease-independent or "off-target" effects. Such information can be used to make determinations as to which drug compounds should be pursued further. Drug compounds that appear to be effective and do not show high off-target effects can then be pursued and tested in clinical trials. In addition, a compaπson of profiles for a disease state versus control and a number of possible treatment conditions of the disease, such as different treatments or different treatment combinations using the methods of the invention in the animal models may permit the selection of candidate treatments for further study and their ranking based on relative rates of reversion to control. Compaπson of the treatment profiles can also be used, as described above, for the optimization of dosages m the animal models by determining the dose which exhibits maximum peptide reversion and minimum induction of disease-independent changes. Ranking and selection may take into account factors independent of the differential expression profile, for example, the relative determination of the best associated percent reversion to control for a given compound may be used in combination with the best mode of delivery of a treatment compound to rank treatment compounds to select a compound that has the best combination. Determine patient stratification for clinical trials Using the methods descnbed herein, it is possible to determine which subjects are responding or not responding to specific drug treatments based on the degree of peptide reversion detected in their peptide reversion profile, that is to say the degree to which disease- related peptides in that indication revert back to normal upon drug treatment. Patients which exhibit greater peptide reversion following drug treatment are deemed to be responders, while patients who do not exhibit any significant peptide reversion are considered non-responders Hence, it is possible, using the methods descnbed herein, to eliminate patients from continued treatment where they will not benefit and to make this determination much earlier than otherwise would have been possible. By extension, it is also possible to increase enrollment of patients who are responding to treatment. Ideally, specific differentially expressed peptide clusters could be distinguished between responder and non-responder sample sets which exist in plasma prior to drug treatment. These specific protein differences could be related to the disease process (i.e. presence or absence of target) or be unrelated such as reflecting different levels of metabolizing enzymes. EXAMPLES Example 1. Plasma proteome changes in disease and drug-treated animals

We have observed in several animal models of disease, including pain and inflammation, and in several human conditions, including Alzheimer's disease (AD) and hypertension, that distinct proteomic profiles can be obtained m each of these disease conditions depending on the specific cπteπa used to select differentially expressed peptides In diseases where an abundance of protein changes are expected to occur m plasma, such as in systemic diseases or endocπne disorders, up to 10% of the observable proteome can be observed to change whereas in more localized disease such as AD significantly fewer plasma proteome changes are observed (i.e., <1% of total study peptides detected) A typical plasma proteomic study conducted at Caprion tracks 30,000 study peptides that are reproducibly detected in each individual plasma sample. Once disease-related peptides have been identified, it is possible to predict the efficacy of compounds on this disease by measuring the degree to which disease-related peptides revert to normal levels, i.e., peptide reversion. This can be done using a variety of well known statistical approaches.

Also, since the goal is frequently to compare one drug against another, it is possible simply to measure the proportion of disease-related peptides which revert to control levels upon treatment at a given dose to establish which drug is more efficacious. For example, a drug capable of reversing 80% of disease-related changes will be more efficacious than a drug which reverses only 30% of disease-related changes in that same disease. Furthermore, it is possible to predict which drugs may induce more off-target (or side effects) by measuring the number of peptides which are induced by the drug treatment but which are unrelated to the disease (i.e., disease-independent peptides). These may be new peptides which were not observed in control and diseased sample sets or peptides which were previously unchanged between these groups which become significantly differentially expressed upon drug treatment Hence, a drug which induces very few up or down-regulated peptides versus controls and disease is a relatively better choice for development than one which induces thousands of disease-independent peptides or "off-target" effects. The degree to which drugs revert a disease-related proteomic profile back to control levels is a measure of its efficacy and can be used to select drugs for patient treatment or to rank compounds for their advancement in the drug development process. Likewise, the degree to which the drug induces off-target effects is an indicator of potential side effects of the drug and can be used to rank different compounds in terms of their side-effect profile In combination, the degree to which a drug is relatively better at reverting disease-related proteins back to control without inducing a series of other off-target peptide changes is a relative measure of the drug's overall likelihood of success in the clinic. Example 2. Plasma proteome analysis of drug-related effects in a rodent model of inflammation

Plasma samples were collected from 18 rats (4 controls treated with vehicle only, 4 animals treated with Complete Freund's Adjuvant (CFA) and vehicle, 5 animals treated with CFA followed by ERB-041 and 5 animals treated with CFA followed by Drug B. CFA induces a well-characterized model of polyarthritis. ERB-041 is an experimental compound being developed by Wyeth Pharmaceuticals and was dosed daily at 5mg/kg. Drug B is another experimental compound bemg considered for further development by Wyeth Pharmaceuticals. All samples were collected at a week post-treatment and depleted of albumin and immunoglobulins using standard immuno-depletion techniques. Forty μg of depleted plasma was run in duplicate on 12% Bis-Tπs NuPAGE mini gels (Invitrogen) using MOPS Running buffer (Invitrogen). Gels were cut into 24 bands using custom-made gel cutting equipment. The bands were placed into 96-well plates for oxidation and tryspin digestion. Peptides were extracted from the gel bands with 0.2 M urea/50% acetomtπle, and vacuum dried. Peptides denved from the 24 bands were resolubihzed m 10% acetomtπle, 0.2% TFA and combined into 10 pools.

Each pool was analyzed by nano LC-MS (Waters capLC coupled to a Q-Tof Ultima) The detected peptides were characteπzed by their mass/charge ratio (m/z), retention time on the liquid chromatography column, their charge and intensity. The peptides that fell within comparable m/z and retention time windows across the samples were designated as Study Peptides. The intensities of the study peptides for each sample set (inflamed, vehicle-, or drug-treated) were compared using MIPS and Constellation Mapping (described in U.S. Patent Application Publication No. 2003/0129760 and PCT publication number WO 2004/049385 respectively) to identify the differentially expressed peptides The results reported were manually validated by verifying the differential intensity of randomly selected targeted ions.

In total, 36,583 study peptides were reproducibly tracked in this study. Study peptides were considered to be differentially expressed when at least 3 of the animals within a sample set contained peptides with intensities 5 times greater than the average intensity of the same study peptide m any other sample set. By this definition, 711 study peptides were differentially expressed between the sample sets. Bioinformatic analysis of results obtained from the rodent inflammation study clearly showed that the two Wyeth drugs were capable of reverting the vast majority of disease-related peptides back to control levels suggesting that they were both highly efficacious compounds. This was later confirmed by comparison of the activity of drugs on the observed phenotype of animals whose joint swelling scores returned to virtually normal levels upon treatment with either ERB-041 or Drug B. Nonetheless, several subtle differences between the two compounds could be detected which are useful in prioritizing compounds for further development. ERB-041 was slightly better at reverting disease-related peptides back to control than Drug B and showed less off- target effects as shown in Figure 2. Figure 2 is a heat map representing the major plasma proteome changes which can be observed between control animals, CFA-treated or diseased animals, as well as diseased animals treated with ERB-041 and Drug B. The heat map demonstrates that the control group of animals form a distinct plasma proteome profile when compared to the diseased group. Upon treatment with either drug, the plasma proteome profile of diseased animals reverted to a profile resembling that of control animals, thus predicting that these compounds were highly efficacious. Indeed, when animals were treated with either drug, most of the inflammation-induced changes disappeared and reduction of the joint swelling and redness was observed (see Figure 3). CFA-iηjected vehicle-treated rats had joint scores of 12 throughout the study, whereas CFA-mjected ERB-041 -treated rats' joint scores were reduced to ~2 after 4 days of treatment and stabilized at ~1 after 10 days of treatment. Normal rats have joint scores of 0. However, ERB-041 reverted slightly more disease-related proteome changes back to control levels than Drug B. Furthermore, treatment with ERB-041 induced the up-regulation of a very small group of proteins which were unrelated to the disease profile (disease-independent peptides) suggesting minimal potential for drug side-effects Drug B induced slightly more of these off-target effects than ERB-041.

Example 3. Plasma proteome analysis of drug-related effects in human neurodegenerative disease.

It is often reported anecdotally that AD patients fare much better when treated with acetylcholinesterase inhibitors. Nonetheless, the difficulty in measuring efficacy m this disease indication as well as its long time course have made the development of drugs for AD among the most expensive drug development avenues in the industry and among those with the highest failure rates. We believe that plasma protein profiling as illustrated in the following example will accelerate the selection of appropriate patients for enrollment in a clinical trial, accelerate and improve the monitoring of treatment effects in this population, and help select the drugs with the highest likelihood of success based on their ability to revert disease-related changes back to normal and not induce other off-target effects.

The following experiments were conducted using human plasma samples were obtained through a collaboration with researchers at the Lady Davis Institute in Montreal and were all derived from out-patients being examined at the McGiIl Memory Clinic. AU patients were evaluated through medical histories and neuropsychological testing and diagnosed as either MCI, AD or normal age -matched controls. In total 60 patients consented to have their plasma analyzed; twenty age-matched controls, 20 MCI, and 20 AD patients were enrolled A number of patients were removed from the study prior to analysis due to the presence of important confounding factors such as having co-morbid conditions such as high blood pressure, diabetes, arthritis, or being smokers. Also removed were those younger than 60 years or older than 85 years or having a Mini-Mental State Examination score which was outlying from the group. The AD sample set was comprised of 18 patients, 7 who were untreated for their condition and thus represented a true diseased group and 11 patients who were undergoing treatment with acetylcholinesterase inhibitors (Aricept™ and Cognex™)

Blood was drawn and processed by a standard procedure. Plasma was stored at -8O⁰C in 0 1 ml ahquots. Plasma was depleted of albumin and immunoglobulins by through the use of Agilent columns. Forty μg of depleted plasma was run in duplicate on 12%Bis-Tπs NuPAGE mini gels (Invitrogen) using MOPS Running buffer (Invitrogen). Gels were cut into 24 bands using custom-made gel cutting equipment. The bands were placed into 96-well plates for oxidation and tryspm digestion, peptides were extracted from the gel bands with 0.2M urea/50% acetomtπle, and vacuum dried. Peptides derived from the 24 bands were resolubihzed in 10% acetonitrile, 0.2% TFA and combined into 10 pools.

Each pool was analyzed by nano LC-MS (Waters capLC coupled to a Q-Tof Ultima) The detected peptides were characteπzed by their mass/charge ratio (m/z), retention time on the liquid chromatography column, their charge and intensity. The peptides that fell withm comparable m/z and retention time windows across the samples were designated as Study Peptides. The peptide intensities for each sample set (age-matched controls, MCI, and AD) were compared using MIPS, Constellation Mapping and VMS (described in U.S. Patent Application Publication No. 2003/0129760, PCT publication number WO 2004/049385, and U S. Provisional Patent Application Number 60/691,414, respectively) to identify the differentially expressed peptides. Plasma samples were treated as descπbed above and injected into the mass spectrometer for liquid chromatography-mass spectrometry analysis

Each sample set m the study exhibited statistically significant changes with respect to the other. However, these differences were much less pronounced than in the animal inflammation study probably owing to the nature of the disease and its localization to the bram but also due to the increased biological vaπation inherent in the plasma proteome of individuals as compared to inbred rats. Despite this inherent biological variability, it was possible to distinguish each sample set from each other and from age-matched controls using multi -dimensional scaling (see Figures 4A, 4B, 5 A, 5B, 1OA, and 10B). Figure 4A is a multidimensional scaling plot of 13 MCI plasma profiles versus 17 age-matched controls showing that the two sample sets form distinct clusters along one axis indicating that MCI patients can be distinguished from controls based on the protein plasma profile. Figure 4B is a heat map demonstrating the up-regulation of 662 peptides in MCI versus controls samples Figure 5 A is a multidimensional scaling plot of 18 AD plasma profiles versus 17 age -matched controls showing that the two sample sets form distinct clusters along the x axis indicating that AD patients can be distinguished from controls based on the protein plasma profile. Figure 5B is a heat map illustrating that 331 peptides were down-regulated in AD as compared to controls. Figure 1OA is a multidimensional scaling plot of 13 MCI plasma profiles versus 18 AD samples showing the clustering of the two sample sets into distinct groups. The heat map in Figure 1OB illustrates the 442 peptides which were >2-fold up- regulated in AD as compared to MCI and exhibited a significance of P<0.05.

Furthermore, a seπes of peptides were discovered which showed a trend toward increasing differences between MCI and AD suggesting that plasma protein profiles were reflective of disease severity (Figure 6-7). Figure 6 is a multidimensional scaling plot where 5 untreated AD patient plasma profiles were compared to those of 13 MCI and 17 age- matched controls. The plot shows three distinct clusters for age-matched controls, MCI, and AD samples suggesting not only a distinct plasma protein profile for each but also a progression from MCI to AD. Thus, it can be predicted from this plot that MCI patients have a phenotype intermediate between controls and AD. Figure 7 is a heat map illustrating 121 peptides which show a progression from age-matched controls to MCI to AD either through peptides becoming increasingly down-regulated (as in approximately peptides 63-100) or increasingly up-regulated (as in approximately peptides 108-121) from a mild condition to a more advanced chronic neurodegenerative one.

Despite the fact that the diagnosis of Alzheimer's disease is difficult (and can only be confirmed post-mortem), it was possible to observe peptide reversion in the drug-treated AD patients as compared to those who were untreated (Figures 8-9). Figure 8 is a multidimensional scaling plot of 5 untreated AD patient plasma profiles, 17 age-matched controls and 13 AD patient being treated with acetylcholinesterase inhibitors. In this figure, the AD untreated groups appears clearly distinct from the age -matched controls and AD- treated groups indicating that AD patients undergoing treatment with acetylcholinesterase inhibitors display peptide reversion of disease-related peptides back to control levels.

Figure 9 is a heat map representation of the same sample sets described in Figure 8 in which 383 peptides were selected on the basis of their being expressed at >1.5 fold differential abundance and p<0.05 among sample sets. This heat map now combines up- and down-regulated peptides, which are expressed in red and blue respectively (as for Figure 2) Overall, the AD untreated group appears distinct from both the age -matched controls and AD- treated groups. Individual peptides can be seen to be differentially regulated in AD versus controls or AD-treated.

The efficacy of compounds in AD has historically been extremely difficult to measure. These experiments demonstrate the use of protein plasma profiling for determining whether drugs are effective in certain patients and which drugs are more efficacious in this disease and have a higher chance of success in the clinic. These observations confirm our hypothesis that plasma protein profiles reflect the health status of individuals and that peptide reversion can be used to measure the relative therapeutic efficacy of compounds in disease. A more detailed examination of the differences between the MCI and AD groups revealed a sub-group of MCI patients with a distinct protein profile from the rest of MCI patients. The heat map shown in Figure 1OB clearly demonstrates that the MCI group appeared to be equally divided by a sub-group expressing a distinct pattern of peptide expression (for example in peptides 360-442). This subgroup represented approximately 50% of all MCI patients examined. This proteomic profile difference may reflect the well-known fact that 50% of MCI patients go on to develop AD whereas the other 50% remain stable or do not. By following which patients go on to develop AD, it will be possible to correlate these distinctive proteomic profiles with the likelihood of developing AD or not and thus be able to initiate more aggressive or appropriate therapeutic treatment earlier. Such profiles can also be used to diagnose AD at an early stage and using non-invasive procedures.

OTHER EMBODIMENTS

The description of the specific embodiments of the invention is presented for the purposes of illustration. It is not intended to be exhaustive or to limit the scope of the invention to the specific forms described herein. Although the invention has been described with reference to several embodiments, it will be understood by one of ordinary skill in the art that various modifications can be made without departing from the spirit and the scope of the invention, as set forth in the claims. All patents, patent applications, and publications referenced herein are hereby incorporated by reference.

Other embodiments are in the claims.

Claims

Claims

1. A method for determining the therapeutic efficacy of a compound for the treatment of a disease in a subject, said method comprising:

(a) obtaining the expression levels of a plurality of disease-related peptides from a normal reference subject, from a subject having said disease, and from a subject having said disease and treated with said compound;

(b) determining the percentage of said plurality of disease-related peptides obtained from the subject having said disease and treated with said compound whose expression levels revert to substantially the same expression levels obtained from the normal reference subject; wherein, a compound is therapeutically effective for the treatment of said disease in said subject if a substantial percentage of said plurality of disease-related peptides from said subject having said disease and treated with said compound revert to substantially the same expression levels obtained for the normal reference subject.

2. The method of claim 1, wherein the biological sample from said subject having said disease and treated with said compound is a bodily fluid.

3. The method of claim 2, wherein said bodily fluid is selected from the group consisting of blood, plasma, serum, urine, bile, cerebrospinal fluid, aqueous or vitreous humor, an exudate, and fluid obtained from a joint.

4. The method of claim 1, wherein said disease is selected from the group consisting of cancer, immune disease, haematopoietic diseases, reproductive system disorders, musculoskeletal diseases, cardiovascular diseases, mixed fetal diseases, excretory diseases, neural or sensory diseases, respiratory disease, endocrine diseases, digestive diseases, connective tissue diseases, and epithelial diseases.

5. The method of claim 1, wherein the compound is a highly effective compound for the treatment of said disease if the percentage of the plurality of disease-related peptides from said subject having said disease and treated with said compound that revert to substantially the same expression levels obtained for the normal reference sample is at least about 80%.

6. The method of claim 1, wherein the compound is a moderately effective compound for the treatment of said disease if the percentage of the plurality of disease-related peptides from said subject having said disease and treated with said compound that revert to substantially the same expression levels obtained for the normal reference sample is between about 50% and about 80%.

7. The method of claim 1, wherein the compound is a weakly effective compound for the treatment of said disease if the percentage of the plurality of disease-related peptides from said subject having said disease and treated with said compound that revert to substantially the same expression levels obtained for the normal reference sample is between about 20% and about 50%.

8. The method of claim 1, wherein said method is used to select a candidate compound from a library of compounds.

9. The method of claim 1 , wherein the expression levels of said plurality of disease- related peptides in the biological sample from a normal reference subject or the plurality of disease- related peptides in the biological sample from the subject having said disease are obtained from a database.

10. The method of claim 1, wherein said plurality of disease-related peptides comprises at least 50 peptides.

11. The method of claim 1 , wherein said plurality of disease-related peptides are identified by the following steps:

(i) obtaining a first biological sample from a first subject that is a normal reference subject and a second biological sample from a second subject having a disease;

(ii) measuring the expression levels of a plurality of peptides in the first, and second biological samples; and

(iii) comparing the expression levels of the plurality of peptides from the first and second biological samples; wherein a plurality of peptides that are differentially expressed in the first biological sample as compared to the second biological sample are identified as a a plurality of disease-related peptides.

12. The method of claim 11, wherein the comparing of step (iii) comprises the use of a computer implemented method.

13. A method for comparing the therapeutic efficacy of two or more compounds for the treatment of a subject having a disease, said method comprising:

(a) obtaining the expression levels of a plurality of disease-related peptides from a normal reference subject, from a subject having said disease, from a first subject having said disease and treated with a first compound; and from a second subject having said disease and treated with a second compound; (b) determining the percentage of said plurality of disease-related peptides from said first subject whose expression levels revert to substantially the same expression levels obtained from the normal reference subject; and

(c) determining the percentage of said plurality of disease-related peptides from said second subject whose expression levels revert to substantially the same expression levels obtained from the normal reference subject; wherein the first compound is therapeutically more effective than the second compound for the treatment of a subject having said disease if said percentage determined m step (b) for the first subject is greater than the percentage determined in step (c) for the second subject; or wherein the second compound is therapeutically more effective than the first compound for the treatment of a subject having said disease if said percentage determined m step (c) for the second subject is greater than the percentage determined in step (b) for the first subject.

14. The method of claim 13, wherein the first and second compounds are analogs of each other.

15. The method of claim 13, wherein one of the first or second compounds is a generic version of the other of one of the first or second compounds.

16. The method of claim 13, wherein any of the biological samples is a bodily fluid

17. The method of claim 16, wherein said bodily fluid is selected from the group consisting of blood, plasma, serum, urine, bile, cerebrospinal fluid, aqueous or vitreous humor, an exudate, and fluid obtained from a joint.

18. The method of claim 13, wherein said disease is selected from the group consisting of cancer, immune disease, haematopoietic diseases, reproductive system disorders, musculoskeletal diseases, cardiovascular diseases, mixed fetal diseases, excretory diseases, neural or sensory diseases, respiratory diseases, endocπne diseases, digestive diseases, connective tissue diseases, and epithelial diseases.

19. The method of claim 13, wherein the expression levels of said plurality of disease- related peptides from the normal reference subject or from the subject having the disease are obtained from a database.

20. The method of claim 13, wherein said plurality of disease-related peptides comprises at least 50 peptides.

21. The method of claim 13, wherein the plurality of disease-related peptides of step (a) are obtained by the following steps:

(i) obtaining a first biological sample from a first subject that is a reference subject and a second biological sample from a second subject having a disease;

(ii) measuring the expression levels of a plurality of peptides in the first, and second biological samples; and

(iii) comparing the expression levels of the plurality of peptides from the first and second biological samples; wherein a plurality of peptides that are differentially expressed in the first biological sample as compared to the second biological sample are identified as a plurality of disease-related peptides.

22. The method of claim 21, wherein said measuring of step (ii) comprises the use of chromatography, mass spectrometry, or immunological methods.

23. The method of claim 21 , wherein the comparing of step (iii) comprises the use of a computer implemented method.

24. A method for comparing the toxicity of two or more compounds for the treatment of a subject having a disease, said method comprising:

(a) obtaining a peptide expression profile for a plurality of study peptides in a first biological sample from a first subject having said disease and treated with a first compound and a second biological sample from a second subject having said disease and treated with a second compound;

(b) obtaining a peptide expression profile for a plurality of study peptides from a third subject that is a normal reference subject and a fourth subject that is a subject having said disease;

(c) comparing the peptide expression profile of the first biological sample of step (a) with the peptide expression profile from the third subject and the peptide expression profile from the fourth subject;

(d) identifying a plurality of disease-independent peptides that are differentially expressed in the peptide expression profile of the first biological sample as compared to the peptide expression profile from the third subject and the peptide expression profile from the fourth subject;

(e) comparing the peptide expression profile of the second biological sample of step (a) with the peptide expression profile from the third reference subject and the peptide expression profile from the fourth subject;

(f) identifying a plurality of disease-independent peptides that are differentially expressed in the peptide expression profile of the second biological sample as compared to the peptide expression profile from the third subject and the peptide expression profile from the fourth subject; and

(g) comparing the plurality of disease-independent peptides identified in step (d) for the first biological sample with the plurality of disease-independent peptides identified in step (f) for the second biological sample; wherein the first compound is more toxic than the second compound if the plurality of disease-independent peptides identified in step (d) for the first biological sample is greater than the plurality of disease-independent peptides identified in step (f) for the second biological sample; or wherein the second compound is more toxic than the first compound if the plurality of disease-independent peptides identified in step (f) for the second biological sample is greater than the plurality of disease-independent peptides identified in step (d) for the first biological sample.

25. The method of claim 24, wherein the first and second compounds are analogs of each other.

26. The method of claim 24, wherein one of the first or second compounds is a generic version of the other one of the first or second compounds.

27. The method of claim 24, wherein the first biological sample or the second biological sample is a bodily fluid.

28. The method of claim 27, wherein said bodily fluid is selected from the group consisting of blood, plasma, serum, urine, bile, cerebrospinal fluid, aqueous or vitreous humor, an exudate, and fluid obtained from a joint.

29. The method of claim 24, wherein said disease is selected from the group consisting of cancer, immune diseases, haematopoietic diseases, reproductive system disorders, musculoskeletal diseases, cardiovascular diseases, mixed fetal diseases, excretory diseases, neural or sensory diseases, respiratory disease, endocrine diseases, digestive diseases, connective tissue diseases, and epithelial diseases.

30. The method of claim 24, wherein the peptide expression profile for the plurality of study peptides from the normal reference subject or the subject having said disease are obtained from a database.

31. The method of claim 24, wherein said plurality of study peptides comprises at least 50 peptides.

32. The method of claim 24, wherein the peptide expression profile for said plurality of study peptides from the first subject, the second subject, the third subject, the fourth subject, or any combination thereof, is obtained from a method comprising:

(i) obtaining a first biological sample from the first subject, a second biological sample from the second subject, a third biological sample from the third subject, or a fourth biological sample from the fourth subject,

(ii) measuring the expression levels of a plurality of peptides in the first biological samples, the second biological sample, the third biological sample, or the fourth biological sample, wherein said measuring comprises the use of chromatography, mass spectrometry, or immunological methods; and

(iii) compiling said expression levels for the plurality of peptides in the first biological samples, the second biological sample, the third biological sample, or the fourth biological sample into a peptide expression profile for the first biological samples, the second biological sample, the third biological sample, or the fourth biological sample.

33. The method of claim 32, wherein the first biological sample, the second biological sample, the third biological sample, or the fourth biological sample is a bodily fluid.

34. The method of claim 33, wherein said bodily fluid is selected from the group consisting of blood, plasma, serum, urine, bile, cerebrospinal fluid, aqueous or vitreous humor, an exudate, and fluid obtained from a joint.

35. A method for determining the optimum dosage for a compound for the treatment of a subject having a disease, said method comprising

(a) obtaining the expression levels of a plurality of disease-related peptides from a first subject that is a normal reference subject, from a second subject having said disease, from a third subject having said disease and treated with said compound at a first dosage, and a fourth subject having said disease and treated with said compound at a second dosage;

(b) determining the percentage of the plurality of disease-related peptides in the biological sample from the third subject that revert to substantially the same expression levels obtained for the first subject; and

(c) determining the percentage of the disease-related peptides in the biological sample from the fourth subject that revert to substantially the same expression levels obtained for the first subject; wherein the optimum dosage for the compound for the treatment of said disease is the first dosage if the percentage determined in step (b) for the third subject is greater than the percentage determined in step (c) for the fourth subject; or wherein the optimum dosage for the compound for the treatment of said disease is the second dosage if the percentage determined in step (c) for the fourth subject is greater than the percentage determined in step (b) for the third subject.

36. The method of claim 35, further comprising:

(i) obtaining the expression levels of the plurality of disease-related peptides from at least one additional subject having said disease and treated with said compound at a third dosage; and

(ii) determining the percentage of the disease-related peptides from the at least one additional subject that reverts to substantially the same expression levels obtained for the first subject; wherein the optimum dosage for the compound is the third dosage if the percentage determined in step (ii) for the at least one additional biological sample is greater than the percentage determined in steps (b) and (c).

37. The method of claim 35, wherein said disease is selected from the group consisting of cancer, immune diseases, haematopoietic diseases, reproductive system disorders, musculoskeletal diseases, cardiovascular diseases, mixed fetal diseases, excretory diseases, neural or sensory diseases, respiratory diseases, endocrine diseases, digestive diseases, connective tissue diseases, and epithelial diseases.

38. The method of claim 35, wherein the expression levels of a plurality of disease- related peptides for a first subject that is a normal reference subject and a second subject having said disease are obtained from a database.

39. The method of claim 35, wherein said plurality of disease-related peptides comprises at least 50 peptides.

40. The method of claim 35, wherein said plurality of disease-related peptides are identified by the following steps:

(i) obtaining a first biological sample from a first subject that is a reference subject and a second biological sample from a second subject having a disease;

(ii) measuring the expression levels of a plurality of peptides in the first, and second biological samples; and

(iii) comparing the expression levels of the plurality of peptides from the first and second biological samples; wherein a plurality of peptides that are differentially expressed m the first biological sample as compared to the second biological sample are identified as a a plurality of disease-related peptides

41. The method of claim 40, wherein said measuring of step (ii) comprises the use of chromatography, mass spectrometry, or immunological methods.

42. The method of claim 40, wherein the comparing of step (in) comprises the use of a computer implemented method.

43. A method for identifying an animal model of a human disease, said method comprising:

(a) obtaining the expression levels of a plurality of disease-related peptides for a first subject that is a normal reference subject and a second subject having said disease;

(b) obtaining the expression levels of a plurality of disease-related peptides from a third subject that is a normal reference animal and from a fourth subject that is a candidate animal model for said human disease, wherein said third and fourth subjects are the same species; and

(c) comparing the plurality of disease-related peptides obtained m step (a) with the plurality of disease-related peptides in step (b); wherein the animal is a model for said human disease if a substantial percentage of the disease-related peptides obtained m step (b) for the fourth subject is expressed at substantially the same expression levels as the disease-related peptides obtained in step (a) for the second subject in step (a).

44. The method of claim 43, wherein said animal is a mammal.

45. The method of claim 44, wherein said mammal is selected from the group consisting of a mouse, bovine, equine, canme, ovine, and felme.

46. The method of claim 43, wherein said disease is selected from the group consisting of cancer, immune diseases, haematopoietic diseases, reproductive system disorders, musculoskeletal diseases, cardiovascular diseases, mixed fetal diseases, excretory diseases, neural or sensory diseases, respiratory diseases, endocnne diseases, digestive diseases, connective tissue diseases, and epithelial diseases.

47. A method for predicting the health status of an individual at πsk for or suspected of having a disease comprising: (a) obtaining a peptide expression profile for a plurality of peptides from a first subject that is a normal reference subject, a peptide expression profile for a plurality of peptides from a second subject having said disease, and a peptide expression profile for a plurality of peptides from a third individual at risk for or suspected of having a disease; and

(b) comparing the peptide expression profiles for the first, second, and third subjects; wherein the health status of the individual is predicted based on the substantial identity of the peptide expression profile of the third subject to the peptide expression profile of the first and second subjects.

48. The method of claim 47, wherein said individual is diagnosed with said disease if the peptide expression profile of the third subject is at least 75% identical to the peptide expression profile of the second subject.

49. The method of claim 47, wherein said method is used to diagnose progression towards a disease state by obtaining a peptide expression profile from said individual at risk for or suspected of having a disease at two or more time points, wherein an increase over time in the identity between the peptide expression profile for the individual at risk for or suspected of having a disease and the peptide expression profile for the second subject indicates progression towards the disease state.

50. The method of claim 47, wherein the peptide expression profile for the plurality of study peptides from the first subject, the second subject, and the third subject is obtained from a method comprising:

(i) obtaining a first biological sample from said first subject, a second biological sample from said second subject, and a third biological sample from said third subject;

(ii) measuring the expression levels of a plurality of peptides in the first biological sample, the second biological sample, and the third biological sample, wherein said measuring comprises the use of chromatography, mass spectrometry, or immunological methods; and

(iii) compiling said expression levels for the plurality of peptides in the first biological sample, the second biological sample, and the third biological sample into a peptide expression profile for the first biological sample, a peptide expression profile for the second biological sample, and a peptide expression profile for the third biological sample.