Peptidomics: Mass Spectrometry for Disease
Peptide Diagnostics
Thousands of peptides per sample
Modern mass spectrometry can identify and quantify thousands of endogenous peptides from a single blood or urine sample, creating a molecular fingerprint that reflects disease states invisible to conventional tests.
Cunningham et al., Front Biol, 2012
Cunningham et al., Front Biol, 2012
View as imageThe human body produces thousands of peptides that circulate in blood, accumulate in tissues, and filter into urine. These endogenous peptides are not random debris. They are products of specific enzymatic processes, signaling cascades, and cellular events that change when disease is present. A tumor secretes different peptides than healthy tissue. A failing heart releases natriuretic peptides at elevated levels. A brain accumulating amyloid plaques shifts its peptide processing patterns. Peptidomics is the systematic study of these endogenous peptide populations using mass spectrometry, and its central premise is straightforward: if disease changes the peptides a body produces, then measuring those peptides can reveal disease. The field has already delivered clinically validated biomarkers (BNP and NT-proBNP for heart failure are peptide biomarkers measured in every emergency department). It is now producing tools for cancer vaccine design, neurodegeneration detection, and organ transplant monitoring. For how synthetic peptides are being designed into diagnostic tests, see Peptide-Based Diagnostic Tests: How Synthetic Peptides Improve Disease Detection. For point-of-care applications that bring peptide measurements to the bedside, see Point-of-Care Peptide Diagnostics: Bringing Lab Tests to the Bedside.
Key Takeaways
- Modern LC-MS/MS peptidomics can identify and quantify thousands of endogenous peptides from a single biological sample, creating disease-specific molecular fingerprints (Cunningham et al., Front Biol, 2012)
- BNP and NT-proBNP are clinically validated peptide biomarkers used in every emergency department worldwide; in 415 elderly patients, cardiac natriuretic peptides predicted cardiovascular mortality over 6 years of follow-up (Alehagen et al., J Card Fail, 2007)
- Immunopeptidomics identified HLA class I-presented peptides in melanoma that revealed immunotherapy targets invisible to genomic analysis alone (Qi et al., J Proteome Res, 2021)
- A warehouse-based immunopeptidome approach enabled personalized peptide cancer vaccines for CLL patients, demonstrating feasibility in a clinical trial with measurable T-cell responses (Heitmann et al., Front Immunol, 2024)
- Proteogenomic immunopeptidomics of ovarian tumors identified shared peptide vaccine candidates across patients, potentially enabling off-the-shelf cancer vaccines for a disease with high unmet need (Chiaro et al., NPJ Vaccines, 2025)
- Combined metabolomics, peptidomics, and proteomics extraction from single samples is now technically feasible, reducing sample requirements and enabling multi-omic disease profiling (Keller et al., Metabolites, 2021)
What Peptidomics Measures and Why It Matters
Peptidomics is distinct from proteomics, though both use mass spectrometry. Proteomics typically involves digesting intact proteins into peptide fragments with enzymes like trypsin, then identifying the parent proteins from these fragments. Peptidomics skips the digestion step entirely. It measures the endogenous peptides that already exist in a biological sample, the natural products of cellular processing, secretion, and degradation.
This distinction matters because endogenous peptides carry information that tryptic peptides do not. When a protease cleaves a protein in a disease-specific way, the resulting peptide fragments become biomarkers for that enzymatic activity. When a tumor cell processes antigens through its HLA machinery, the peptides displayed on its surface become targets for immune recognition. When an organ is damaged, the peptides released into circulation become signals of that specific injury.
Cunningham et al.'s 2012 review in Frontiers in Biology documented the evolution of mass spectrometry-based peptidomics from proof-of-concept studies to systematic biomarker discovery platforms.[1] The field progressed from identifying individual peptides to profiling entire peptidomes, the complete set of endogenous peptides in a sample, to comparing peptidomes between healthy and diseased states. The key technological enabler was the development of liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS), which separates peptides by physical properties before fragmenting them for sequence identification. Modern instruments can identify thousands of unique peptides from a single sample in under two hours.
The Mass Spectrometry Toolbox
MALDI-TOF: The First Generation
Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF) was the first platform widely applied to peptide biomarker discovery. It provides rapid mass measurements of peptides in a sample, generating a "fingerprint" of peaks at different mass-to-charge ratios. Changes in this fingerprint between disease and control samples can indicate candidate biomarkers.
MALDI-TOF is fast and inexpensive per sample, making it suitable for high-throughput screening. Its limitations are lower sensitivity and resolution compared to LC-MS/MS, and it provides mass information without direct sequence identification, making biomarker identification a secondary step.
LC-MS/MS: The Current Standard
Liquid chromatography coupled with tandem mass spectrometry has become the dominant peptidomics platform. Peptides are first separated by liquid chromatography (typically reversed-phase chromatography), then ionized by electrospray ionization, mass-selected in a first mass analyzer, fragmented by collision with inert gas, and the fragment ions measured in a second mass analyzer. The fragmentation pattern reveals the amino acid sequence, providing definitive peptide identification rather than just mass measurement.
Keller et al.'s 2021 study in Metabolites demonstrated that extraction protocols can now be optimized to recover peptides, metabolites, and proteins simultaneously from a single sample aliquot.[2] This combined multi-omic extraction reduces sample volume requirements and enables integrated analysis where peptidomic data is interpreted alongside metabolomic and proteomic data from the same sample. For complex diseases where multiple molecular layers contribute to pathology, this integrated approach provides a more complete picture than any single-omic analysis.
Capillary Electrophoresis-MS: The Urinary Specialist
Capillary electrophoresis coupled with mass spectrometry (CE-MS) has emerged as a specialized platform for urinary peptidomics. Urine is an ideal biofluid for peptidomics because it contains relatively few high-abundance proteins (unlike blood, which is dominated by albumin and immunoglobulins) and reflects kidney, cardiovascular, and systemic disease processes. CE-MS platforms have been developed for chronic kidney disease staging, transplant rejection monitoring, and cardiovascular risk assessment based on urinary peptide patterns.
The CKD273 classifier, developed from CE-MS urinary peptidomics, uses a panel of 273 urinary peptides to predict chronic kidney disease progression. It has been validated in multiple independent cohorts and is the most clinically advanced multi-peptide diagnostic panel produced by peptidomics. The panel includes fragments of collagen, fibrinogen, and uromodulin that reflect fibrotic, inflammatory, and tubular processes in the kidney. Its development illustrates the peptidomics workflow: discovery in a training cohort, validation in independent cohorts, and ongoing prospective evaluation for clinical utility.
Data-Independent Acquisition: The New Frontier
Data-independent acquisition (DIA) mass spectrometry represents the latest technical advance in peptidomics. Traditional data-dependent acquisition (DDA) selects the most abundant peptide ions for fragmentation, meaning low-abundance peptides may never be sequenced. DIA fragments all ions in predefined mass windows, providing a more comprehensive and reproducible peptide census. When combined with spectral library matching, DIA can quantify thousands of peptides with high reproducibility across samples, which is essential for biomarker studies that require consistent measurements across large clinical cohorts.
Peptide Biomarkers Already in Clinical Use
The most successful peptidomics story predates the field's formal naming. B-type natriuretic peptide (BNP) and its N-terminal fragment (NT-proBNP) are endogenous peptides released by cardiac myocytes in response to ventricular wall stress. They are now measured in every emergency department to help diagnose and manage heart failure.
Alehagen et al.'s 2007 study in the Journal of Cardiac Failure compared four cardiac natriuretic peptide biomarkers head-to-head in 415 elderly patients followed over 6 years.[3] All four peptides predicted cardiovascular mortality, but NT-proBNP showed the strongest prognostic value. Patients in the highest NT-proBNP quartile had substantially higher 6-year cardiovascular mortality than those in the lowest quartile. The clinical utility was clear: a single peptide measurement from a blood draw provided prognostic information that improved treatment decisions.
This is what peptidomics aims to replicate at scale: identifying disease-specific peptides, validating them as biomarkers, and deploying them in clinical settings. BNP/NT-proBNP succeeded because the biology was well-understood (cardiac stress causes peptide release), the measurement technology was reliable (immunoassays, later mass spectrometry confirmation), and the clinical need was urgent (ruling out heart failure in patients presenting with shortness of breath). For how other peptide biomarkers signal metabolic dysfunction, see Chromogranin A: The Neuroendocrine Tumor Peptide Marker. For peptide markers of liver injury, see Collagen Peptide Markers in Liver Fibrosis: Diagnostic Applications.
Immunopeptidomics: Mapping Cancer's Vulnerabilities
The most rapidly advancing application of peptidomics is immunopeptidomics: profiling the peptides presented on cell surfaces by HLA (human leukocyte antigen) molecules. Every nucleated cell displays a sample of its internal peptide content on HLA class I molecules, creating a molecular "display window" that cytotoxic T cells survey. Cancer cells, with their mutated proteins and aberrant gene expression, display abnormal peptides that can be recognized as foreign by the immune system.
Immunopeptidomics uses mass spectrometry to identify exactly which peptides tumor cells present. This information is used to design peptide vaccines that train T cells to attack those specific tumor-displayed peptides.
Melanoma
Qi et al.'s 2021 proteogenomic study unveiled the HLA class I-presented immunopeptidome in melanoma, combining genomic sequencing with mass spectrometry-based peptide identification.[4] The proteogenomic approach identified tumor-specific peptides that genomic analysis alone predicted but could not confirm as actually presented on HLA molecules. Confirming actual presentation is critical because only a fraction of potentially mutated peptides are processed and displayed by the HLA machinery. Mass spectrometry provides that confirmation by directly detecting the peptides eluted from HLA molecules on tumor cell surfaces.
Koumantou et al.'s 2019 study took a complementary approach by demonstrating that the melanoma immunopeptidome can be actively edited.[5] By inhibiting endoplasmic reticulum aminopeptidase 1 (ERAP1), a key enzyme in HLA class I antigen processing, they altered which peptides were displayed on melanoma cell surfaces. This pharmacological editing of the immunopeptidome could enhance tumor immunogenicity by shifting the displayed peptide repertoire toward sequences more readily recognized by cytotoxic T cells.
Ovarian Cancer
Chiaro et al.'s 2025 study in NPJ Vaccines applied proteogenomic immunopeptidomics to ovarian tumors and identified shared peptide vaccine candidates across patients.[6] Epithelial ovarian cancers are immunogenic, with CD8+ T cell infiltration being prognostic of clinical outcome. The study identified tumor antigens that were presented across multiple patients' HLA types, potentially enabling "off-the-shelf" peptide vaccines rather than requiring fully personalized manufacturing for each patient. This shared antigen approach could dramatically reduce the cost and complexity of peptide cancer vaccines.
Chronic Lymphocytic Leukemia
Nelde et al.'s 2021 study in Frontiers in Immunology established an immunopeptidomics-guided warehouse concept for peptide-based immunotherapy in chronic lymphocytic leukemia (CLL).[7] Rather than manufacturing a unique vaccine for each patient, the warehouse approach pre-manufactures a library of validated tumor-associated peptides. Each patient's tumor is profiled by immunopeptidomics, and the appropriate combination of pre-made peptides is selected from the warehouse to create a personalized vaccine.
Heitmann et al.'s 2024 clinical trial evaluation demonstrated that this warehouse-based approach was feasible in actual CLL patients.[8] Personalized peptide vaccines composed from the warehouse induced measurable T-cell responses against the selected tumor antigens. This proof-of-concept bridges the gap between discovery peptidomics (identifying what peptides tumors display) and clinical application (using that information to treat patients). For high-throughput alternatives to mass spectrometry, see Peptide Microarrays: Profiling Disease with Thousands of Peptides at Once.
Feola et al.'s 2022 study extended immunopeptidomics into oncolytic virus design, creating a pipeline that combines immunopeptidome profiling with personalized oncolytic vaccine generation.[9] The approach uses mass spectrometry to identify the tumor immunopeptidome, selects optimal peptide antigens, and engineers oncolytic viruses to express those peptides during viral replication within the tumor. This multiplies the immune-stimulating effect by combining viral oncolysis with targeted peptide antigen presentation.
Beyond Cancer: Emerging Peptidomic Applications
Infectious Disease
West et al.'s 2024 study explored endogenous peptide compounds as potential therapeutics for severe influenza, identifying naturally occurring peptides with antiviral or immunomodulatory activity through peptidomic screening.[10] The approach inverts the typical biomarker discovery workflow: rather than looking for peptides that indicate disease, it looks for endogenous peptides that could treat disease. The peptidome becomes a library of candidate therapeutics rather than just diagnostic markers.
Neurodegeneration
Amyloid-beta peptides in cerebrospinal fluid and plasma are among the most intensively studied peptidomic biomarkers. The ratio of amyloid-beta 42 to amyloid-beta 40, measured by immunoprecipitation-mass spectrometry, can detect Alzheimer's disease pathology years before clinical symptoms appear. A 2026 study reported a streamlined IP-MS method that reduced antibody and sample volume requirements by approximately 75% while improving assay performance, moving this peptidomic approach closer to routine clinical deployment. For the full biology of amyloid-beta as a disease peptide, see Amyloid-Beta: The Peptide Fragment at the Heart of Alzheimer's.
Gut Microbiome
The gut microbiome produces bioactive peptides that enter circulation and influence host physiology. Peptidomic profiling of microbial peptides in stool and blood is an emerging approach to diagnosing dysbiosis, inflammatory bowel disease, and metabolic disorders. The challenge is distinguishing microbially-derived peptides from host-derived peptides in complex biological samples. For the intersection of peptide science and microbiome diagnostics, see Microbiome Peptide Profiling: A New Diagnostic Frontier.
Challenges and Limitations
Despite its technical sophistication, peptidomics faces several obstacles to broader clinical adoption.
Sample preparation variability. Endogenous peptides are vulnerable to degradation by proteases in the sample itself. Blood samples begin degrading peptides within seconds of collection. Standardizing collection, processing, and storage protocols across clinical sites remains difficult, and small differences in sample handling can produce large differences in peptide profiles.
Dynamic range. Blood peptide concentrations span many orders of magnitude. High-abundance peptides (albumin fragments, fibrinopeptides) can mask low-abundance disease biomarkers. Depletion strategies remove high-abundance species but can also remove disease-relevant peptides that are bound to carrier proteins.
Validation gap. Hundreds of candidate peptide biomarkers have been discovered through peptidomics. Very few have been validated in independent cohorts large enough to support clinical use. The gap between discovery (identifying a peptide that differs between disease and control) and clinical validation (proving that peptide measurement improves patient outcomes) remains the field's central bottleneck.
Cost and throughput. Mass spectrometry instruments cost $500,000 to $2 million and require trained operators. While sample throughput has improved dramatically, peptidomics remains substantially more expensive per sample than immunoassays or genomic tests. Translating peptidomic discoveries into cheaper, faster clinical assays (typically immunoassays or lateral flow tests) adds another development step.
Biological complexity. The endogenous peptidome is not static. It varies with time of day, recent food intake, physical activity, medication use, and individual genetics. These sources of biological variation create noise that can obscure disease-related signals. Large cohort studies with careful phenotyping are required to distinguish disease biomarkers from normal biological variation, but such studies are expensive and logistically demanding.
Standardization across laboratories. Different mass spectrometers, sample preparation methods, and data analysis pipelines can produce different results from the same samples. International efforts to standardize peptidomic protocols (instrument calibration, quality control samples, data reporting formats) are underway but incomplete. Until standardization is achieved, peptidomic biomarkers discovered in one laboratory may not replicate in another, slowing clinical translation.
Where the Field Is Heading
Three converging trends are reshaping peptidomics. First, machine learning is transforming data analysis. Peptidomic datasets are high-dimensional (thousands of peptide features per sample) and noisy (substantial biological and technical variation). Deep learning models trained on large peptidomic datasets can identify multi-peptide signatures that outperform individual biomarkers, detecting disease patterns that are invisible to human analysts or traditional statistical methods. The integration of AI with peptidomics is particularly powerful for immunopeptidomics, where predicting which peptides will be presented by specific HLA alleles enables in silico vaccine design before any mass spectrometry is performed.
Second, single-cell peptidomics is emerging. Conventional peptidomics analyzes bulk samples containing millions of cells, averaging the peptide content across all cell types. Single-cell mass spectrometry technologies are beginning to resolve peptide content at the level of individual cells, revealing cell-type-specific peptide signatures that bulk analysis obscures. This is relevant for tumor immunopeptidomics, where different subpopulations within a tumor may present different peptide repertoires.
Third, the integration of peptidomics with other omics layers (genomics, transcriptomics, metabolomics, proteomics) is becoming routine. Multi-omic integration provides a more complete molecular picture of disease than any single technology, and peptidomics contributes a layer of information, the endogenous peptide landscape, that genomic and transcriptomic approaches cannot access directly. As Keller et al. showed, combined extraction protocols now make this integration technically straightforward from a single sample aliquot.[2]
The Bottom Line
Peptidomics uses mass spectrometry to profile the thousands of endogenous peptides in biological samples, revealing disease-specific patterns that conventional tests miss. The field has already delivered clinically validated biomarkers (BNP/NT-proBNP for heart failure). Its most active frontier is immunopeptidomics, where mass spectrometry identifies tumor-displayed peptides to guide personalized cancer vaccine design. Clinical trials in CLL and ovarian cancer have demonstrated that immunopeptidome-guided vaccines are feasible and induce measurable immune responses. Challenges in sample standardization, dynamic range, validation throughput, and cost continue to limit broader clinical adoption.