Plasma Proteomics: From Sample Collection to Biomarker Discovery

Plasma proteomics is the large-scale study of proteins circulating in blood plasma, and for researchers pursuing biomarker discovery, the central challenge is that plasma contains a handful of highly abundant proteins that dominate the protein mass while thousands of biologically informative proteins exist at concentrations many orders of magnitude lower. This article provides a practical workflow for plasma proteomics, covering pre-analytical variables, sample preparation strategies, mass spectrometry approaches, and a decision framework for selecting between discovery and targeted methods. The guidance is intended for students, researchers, analysts, and life-science professionals who need concrete decisions about sample handling, depletion, digestion, and data interpretation.

At a Glance

Plasma proteomics requires deliberate choices at every stage, from blood collection tubes to data imputation. The table below summarizes the primary decision points and their practical implications.

Pre-Analytical Variables and Sample Collection

The quality of plasma proteomics data is determined before the sample reaches the mass spectrometer. Plasma and serum are rich sources of information about an individual's health state, and protein tests inform medical decision making, yet few new biomarkers have reached the clinic despite major investments. One reason is that sample-related biases can masquerade as biological findings. Mass spectrometry-based proteomics now allows highly specific and quantitative readout of the plasma proteome, and quality marker panels can assess whether suggested biomarkers are artifacts related to sample handling and processing.

Blood Collection Tubes and Anticoagulants

The choice of blood collection tube affects the proteome you measure. In a large-scale evaluation of plasma handling for clinical trial specimens, researchers compared EDTA and citrated anticoagulant blood collection tubes and found only minor variations associated with tube type. This finding suggests that either anticoagulant is acceptable, provided the same tube type is used consistently across all samples in a study. Mixing tube types within a cohort introduces a systematic variable that can confound biomarker comparisons.

Serum is an alternative to plasma, but the coagulation process removes fibrinogen and other clotting factors, and the cellular components release additional proteins during clot formation. Comparative studies of serum and plasma proteomes have identified sample quality-associated proteins, many of which have been reported as biomarker candidates in the literature. If your study aims to measure coagulation-related proteins, plasma is the appropriate choice. If your study aims to measure proteins released during platelet activation, serum may introduce artifacts that are difficult to distinguish from true biological signals.

Freeze-Thaw Cycles and Storage

Plasma samples are typically stored frozen, but each freeze-thaw cycle can degrade labile proteins and alter measured concentrations. In the FIELD trial evaluation, freeze-thaw cycles produced only minor variations in the measured proteome, but the study used a workflow optimized for abundant plasma proteins. For low-abundance proteins, repeated freeze-thaw cycles are more likely to cause measurable changes. The practical rule is to aliquot plasma into single-use tubes before freezing, so that each aliquot is thawed only once. Record the number of freeze-thaw cycles for each sample and include this variable in your quality assessment.

Hemolysis and Cellular Contamination

Hemolysis releases erythrocyte proteins into plasma, and these contaminants can dominate the measured proteome. Deep reference proteomes of erythrocytes, platelets, plasma, and whole blood have been generated from healthy individuals, enabling the identification of proteins that indicate cellular contamination. If your plasma samples show elevated levels of hemoglobin or other erythrocyte markers, the samples may be unsuitable for biomarker discovery because the measured differences may reflect hemolysis instead of disease biology. The online resource at plasmaproteomeprofiling.org provides a means to assess overall sample-related bias in clinical studies and to prevent costly misassignment of biomarker candidates.

Sample Preparation Strategies

Sample preparation is the most critical step in proteomics because it directly affects the subset of proteins and peptides that can be reliably identified and quantified. Plasma is highly heterogeneous, and protein expression can vary greatly from sample to sample, even for healthy controls. Detection of true biological changes requires that variation from sample preparation steps and downstream analytical detection methods remains low.

Depletion of High-Abundance Proteins

The protein concentration dynamic range in blood plasma exceeds ten orders of magnitude, and high-abundance proteins dominate the analysis, often rendering the analysis of low-abundance proteins impossible. Depleting high-abundance proteins is one strategy to solve this problem.

Immunodepletion removes specific abundant proteins such as albumin and immunoglobulin G using antibodies. In the FIELD trial evaluation, immunodepleting albumin and IgG provided few additional identifications compared to undepleted plasma, while adding cost and time. Cibachrome-blue-based depletion yielded additional proteins but with cost and time expenses. These findings suggest that for studies focused on the moderately abundant proteome, depletion may not be necessary when using modern mass spectrometry instrumentation.

Alternative depletion strategies are emerging. Nanodiamond-based depletion relies on selective binding of serum proteins to the surface of nanodiamonds. In proof-of-principle experiments, this simple approach enabled detection of proteins present at concentrations of 1 ng/mL, and remarkably, proteins down to 400 pg/mL after only one depletion step. Numerous disease biomarkers were detectable, including markers for multiple cancer forms, cardiovascular diseases, and Alzheimer's disease. Many of these biomarkers could not be detected with a state-of-the-art ultrahigh-performance liquid chromatography column that depletes the 64 most-abundant serum proteins.

Nanoparticle protein corona technology represents another frontier approach. Nanoparticles introduced into plasma become coated with a corona of proteins, and this corona enriches low-abundance plasma proteins while the abundant proteins remain in the supernatant. This approach can unveil previously inaccessible layers of the proteome, though it currently has limitations in standardization and throughput.

Digestion Methods

Protein digestion converts the proteome into peptides suitable for mass spectrometry analysis. Two common approaches are filter-aided sample preparation (FASP) and in-solution digestion.

FASP uses an ultrafiltration unit to perform buffer exchange, protein denaturation, reduction, alkylation, and digestion in a single device. In a comparative evaluation of human milk and plasma proteomes, FASP appeared to be the most time-efficient procedure with a low miscleavage rate when used for a biological sample aliquot, but quantitation was less reproducible. A prior protein precipitation step improved the quantitation by FASP due to significantly higher peak areas for plasma and much better reproducibility for milk. The miscleavage rate for milk, the identification rate for plasma, and the carbamidomethylation efficiency were also improved with prior precipitation.

In-solution digestion is simpler but produced higher miscleavage rates for both milk and plasma in the same evaluation, making it less suitable for targeted proteomics. For seminal plasma, a protocol using filter-aided sample preparation has been described for creating a high-quality peptide mixture for bottom-up proteomic analysis.

Automation and Throughput

Automation is necessary to increase sample processing throughput for large-scale clinical analyses. Replacement of manual pipettes with robotic liquid handler systems is especially helpful in processing blood-based samples such as plasma and serum. These samples are very heterogeneous, and protein expression can vary greatly from sample to sample, even for healthy controls. Automation reduces variation from sample preparation steps and enables detection of low-abundant proteins while providing low sample error and increased sample throughput.

A robust plasma analysis workflow combining automated sample preparation with micro-flow LC-MS can analyze dozens of samples simultaneously with high reproducibility. Without protein depletion and prefractionation, more than 300 protein groups can be identified in a single analysis using a micro-flow LC-MS system on an Orbitrap Exploris 240 mass spectrometer, including quantification of 35 FDA-approved disease markers. The quantitative precision of the entire workflow was acceptable with a median coefficient of variation of 9%. This workflow is suitable for the analysis of large-scale clinical plasma samples with simple and time-saving operation.

Mass Spectrometry Approaches

Mass spectrometry has emerged as the preferred analytical tool for plasma proteomics due to its nonbiased ability to characterize thousands of proteins in an experiment and its ability to identify low-abundance proteins. The choice between discovery and targeted approaches depends on the research question, sample cohort size, and available instrumentation.

Discovery Proteomics with Data-Dependent Acquisition

Data-dependent acquisition selects the most abundant precursor ions for fragmentation, providing broad proteome coverage. For plasma samples, this approach can identify hundreds to thousands of proteins depending on the depth of fractionation and the instrumentation used. Discovery experiments are appropriate when the goal is to generate hypotheses about candidate biomarkers without prior knowledge of which proteins are likely to change.

Data-Independent Acquisition

Data-independent acquisition systematically fragments all precursor ions within a defined mass range, providing more reproducible quantification across samples. In the FIELD trial evaluation, LC-MS with data-independent acquisition of undepleted plasma conducted over a 45-minute gradient yielded 172 proteins after excluding immunoglobulin isoforms. From 65 batches involving over 1500 injections, the median intra-batch quantitative differences in the top 100 proteins of the plasma external standard were less than 2%. This reproducibility makes data-independent acquisition well suited for large-scale clinical studies where samples are analyzed across multiple batches.

Targeted Mass Spectrometry

Targeted approaches such as selected reaction monitoring or parallel reaction monitoring focus the mass spectrometer on specific peptides corresponding to candidate proteins. These methods provide high sensitivity, specificity, and quantitative accuracy, making them appropriate for validation studies. A two-tiered proteomics strategy that combines global discovery with targeted validation was used to identify a protein signature of remission in anti-neutrophil cytoplasmic autoantibody-associated vasculitis. The resulting 7-protein panel was implemented in a targeted mass spectrometry-based assay and confirmed in an independent patient cohort, consistently outperforming routine markers such as C-reactive protein and ANCA titer.

Affinity-Based Platforms

Affinity-based platforms such as Olink proximity extension assays and SomaScan provide an alternative to mass spectrometry for measuring specific proteins. A comparative evaluation of the Olink Explore 3072 proximity extension assays and a peptide fractionation-based mass spectrometry method (HiRIEF LC-MS/MS) on 88 plasma samples found that the platforms exhibited complementary proteome coverage, high precision, and concordance in estimating sex differences in protein levels. Quantitative agreement between platforms was moderate, with a median correlation of 0.59, mainly influenced by technical factors. This finding suggests that combining mass spectrometry and affinity-based approaches can provide more comprehensive and reliable plasma proteome profiling than either platform alone.

Decision Tree for Method Selection

The choice between discovery and targeted methods depends on several factors that should be evaluated before beginning a study.

Step 1: Define the Research Question

If the goal is to discover novel biomarkers without prior hypotheses, a discovery approach using data-dependent or data-independent acquisition is appropriate. If the goal is to validate specific candidate proteins in a large cohort, a targeted approach using mass spectrometry or affinity-based assays is more suitable. If the goal is to measure a defined panel of proteins across many samples, affinity-based platforms may offer higher throughput.

Step 2: Assess Sample Cohort Size

For small cohorts of fewer than 100 samples, discovery proteomics with extensive fractionation can provide deep proteome coverage. For large cohorts of hundreds or thousands of samples, throughput and reproducibility become critical. Automated sample preparation combined with micro-flow LC-MS can analyze dozens of samples simultaneously with high reproducibility, making it suitable for large-scale clinical samples. The FIELD trial workflow was developed specifically to allow LC-MS analysis of more than 1500 samples.

Step 3: Evaluate Required Sensitivity

If the proteins of interest are expected to be present at very low concentrations, consider depletion or enrichment strategies. Nanodiamond-based depletion enabled detection of proteins down to 400 pg/mL. Nanoparticle protein corona enrichment can unveil previously inaccessible layers of the proteome. If the proteins of interest are moderately abundant, undepleted workflows may suffice.

Step 4: Consider Available Resources

Depletion strategies add cost and time. Immunodepleting albumin and IgG provided few additional identifications in the FIELD trial evaluation while adding expense. Cibachrome-blue-based depletion yielded additional proteins but with cost and time expenses. Automation requires capital investment in robotic liquid handlers but reduces labor costs and improves reproducibility for large studies.

Step 5: Plan for Validation

Biomarker discovery experiments generate candidate lists that require validation in independent cohorts. A scalable blueprint for biomarker discovery in complex inflammatory diseases uses a two-tiered strategy that combines global discovery with targeted validation. Plan the validation phase when designing the discovery phase so that sufficient samples are available for both stages.

Records and Measurements

Maintaining detailed records is essential for reproducible plasma proteomics. The following measurements should be documented for every sample.

Sample Collection Records

Record the collection date and time, the anticoagulant type, the collection site, and the time between collection and processing. Record the centrifugation parameters used to separate plasma from cellular components. Record the storage temperature and the number of freeze-thaw cycles for each aliquot.

Sample Quality Metrics

Measure and record hemolysis indicators such as hemoglobin levels. Assess cellular contamination by monitoring erythrocyte and platelet marker proteins in the proteomics data. The quality marker panels defined by Plasma Proteome Profiling can assess plasma samples and the likelihood that suggested biomarkers are artifacts related to sample handling and processing.

Instrument Performance Records

For mass spectrometry experiments, record the instrument settings, the chromatography gradient, and the batch number for each set of samples. Include a pooled plasma external standard in each batch to monitor quantitative reproducibility. In the FIELD trial evaluation, the median intra-batch quantitative differences in the top 100 proteins of the plasma external standard were less than 2% across 65 batches.

Data Processing Records

Record the software versions used for peptide identification, quantification, and statistical analysis. Document the parameters used for missing value imputation, normalization, and differential abundance testing. These records enable other researchers to reproduce the analysis and allow for reanalysis when software improves.

Common Failure Patterns

Several recurring problems undermine plasma proteomics studies. Recognizing these patterns early can prevent costly mistakes.

Batch Effects

When samples are analyzed across multiple batches, systematic differences in instrument performance, reagent lots, or environmental conditions can create artificial differences between groups. The FIELD trial evaluation demonstrated that careful workflow optimization can keep intra-batch quantitative differences below 2%, but inter-batch variation requires randomization of samples across batches and inclusion of bridging standards.

Hemolysis Contamination

Hemolyzed samples introduce erythrocyte proteins that can dominate the measured proteome. If hemolysis is more common in one study group than another, the resulting differences may be misinterpreted as disease-related biomarkers. The comprehensive literature survey conducted by Plasma Proteome Profiling revealed that many reported biomarker candidates are actually sample quality-associated proteins.

Overinterpretation of Low-Abundance Proteins

Proteins present at very low concentrations are more susceptible to technical variation and missing values. Imputation techniques provide a means to replace missing measurements with a value and are used in almost all downstream analysis of mass spectrometry-based proteomics data using label-free quantification. Deep learning approaches for imputation can recover biologically meaningful proteins that would otherwise be excluded from analysis. In an alcohol-related liver disease cohort with blood plasma proteomics data available for 358 individuals, removing 20 percent of the intensities allowed recovery of 15 out of 17 significant abundant protein groups using variational autoencoder imputations. When analyzing the full dataset, 30 additional proteins were identified as significantly differentially abundant across disease stages compared to no imputation, and some of these were predictive of disease progression in machine learning models.

Platform Discrepancies

Different measurement platforms can produce different quantitative results for the same proteins. The moderate quantitative agreement between mass spectrometry and affinity-based platforms means that candidate biomarkers identified on one platform may not replicate on another. Cross-platform validation is essential before committing to a large-scale validation study.

Quality Controls and Reproducibility

Reproducibility in plasma proteomics requires deliberate quality control at every stage of the workflow.

Pre-Analytical Quality Control

Use the same blood collection tubes throughout the study. Process samples using the same centrifugation parameters. Aliquot plasma into single-use tubes before freezing. Record the time between collection and freezing for every sample.

Analytical Quality Control

Include a pooled plasma external standard in every batch to monitor instrument performance. In the FIELD trial evaluation, the median intra-batch quantitative differences in the top 100 proteins of the plasma external standard were less than 2% from 65 batches involving over 1500 injections. Monitor the coefficient of variation for the external standard and investigate any batch where the variation exceeds acceptable thresholds.

Post-Analytical Quality Control

Assess sample quality using marker panels that detect hemolysis, platelet activation, and other pre-analytical artifacts. The online resource at plasmaproteomeprofiling.org provides a means to assess overall sample-related bias in clinical studies. Exclude or flag samples that show evidence of contamination before performing statistical analysis.

Data Sharing and Reproducibility

Deposit raw data and processed results in public repositories to enable independent verification. The National Center for Biotechnology Information provides data resources for depositing and accessing proteomics data. The FAIR Guiding Principles provide a framework for making data findable, accessible, interoperable, and reusable. The National Institutes of Health Genomic Data Sharing Policy describes expectations for data sharing in NIH-funded research.

Limitations and Interpretation

Plasma proteomics has inherent limitations that should be acknowledged when interpreting results.

Dynamic Range Constraints

The protein concentration dynamic range in blood plasma exceeds ten orders of magnitude. Even with depletion and enrichment strategies, the deepest measurements capture only a fraction of the theoretical proteome. Proteins present at very low concentrations may be missed entirely, and their absence from the data does not mean they are absent from the sample.

Correlation with Disease

Proteins measured in plasma may reflect the disease process, the body's response to the disease, or unrelated physiological variation. Adjusting for patient characteristics and clinical variables is essential to avoid confounding. The confounder-controlled proteomics approach used in the vasculitis remission study provides a scalable blueprint for biomarker discovery in complex inflammatory diseases.

Causality

Proteomics measurements are correlational. A protein that differs between disease and control groups may be a cause, a consequence, or a bystander in the disease process. Proteomics bridges the gap between genomics and phenotype by providing function-proximal readouts of protein abundance, post-translational modifications, pathway activity, and intercellular signaling, but causal inference requires additional experimental approaches.

Generalizability

Biomarker panels discovered in one cohort may not perform in other populations. The 7-protein panel for vasculitis remission was confirmed in an independent patient cohort, but broader validation across diverse populations is needed before clinical implementation. Proteomic evidence in myeloproliferative neoplasms spans four disease axes, including clonal fitness, bone marrow microenvironmental remodeling, chronic inflammation and thrombosis, and leukemic transformation, but translating these findings into clinical panels requires multicenter standardization.

Safety and Regulatory Context

Plasma proteomics research involving human subjects is subject to ethical and regulatory requirements. Researchers must obtain informed consent from participants, protect participant privacy, and comply with applicable regulations for human subjects research. The National Institutes of Health Genomic Data Sharing Policy describes expectations for data sharing and privacy protection in NIH-funded research.

When working with human blood samples, follow institutional biosafety guidelines for handling potentially infectious materials. Use appropriate personal protective equipment and follow standard precautions for bloodborne pathogens.

For studies that may lead to clinical tests, be aware that diagnostic tests are subject to regulatory oversight. The 35 FDA-approved disease markers quantified in the automated plasma proteomics workflow illustrate the potential for plasma proteomics to inform clinical decision making, but new biomarker panels require regulatory approval before clinical use.

Professional Escalation Criteria

Researchers should seek expert consultation when encountering specific challenges in plasma proteomics.

When to Consult a Biostatistician

Consult a biostatistician when designing the statistical analysis plan, particularly for studies with complex designs involving multiple batches, repeated measures, or confounding variables. The confounder-controlled approach used in the vasculitis remission study required careful adjustment for patient characteristics and clinical variables.

When to Consult a Mass Spectrometry Specialist

Consult a mass spectrometry specialist when instrument performance degrades, when developing new acquisition methods, or when troubleshooting poor reproducibility. The development of the automated plasma analysis workflow required systematic evaluation of stability and reproducibility.

When to Consult a Bioinformatics Specialist

Consult a bioinformatics specialist when processing large datasets, implementing imputation methods, or integrating data from multiple platforms. The deep learning imputation methods for label-free quantification require specialized computational expertise.

When to Consult a Regulatory Specialist

Consult a regulatory specialist before planning studies that may lead to clinical tests or when handling data subject to privacy regulations. The Genomic Data Sharing Policy describes expectations for data sharing in NIH-funded research.

Frequently Asked Questions

What is the difference between plasma and serum for proteomics?

Plasma is the liquid component of blood after removal of cells, obtained by centrifuging blood collected with an anticoagulant. Serum is the liquid remaining after blood has clotted, which removes fibrinogen and other clotting factors. Plasma preserves coagulation-related proteins, while serum may contain proteins released from platelets and other cells during clot formation. The choice depends on the research question, but consistency within a study is essential.

Why is plasma considered difficult for proteomics analysis?

Plasma contains a few highly abundant proteins that dominate the total protein mass, while thousands of biologically informative proteins exist at concentrations many orders of magnitude lower. The protein concentration dynamic range exceeds ten orders of magnitude, making it difficult to detect low-abundance proteins without depletion or enrichment strategies.

Should I deplete high-abundance proteins from my plasma samples?

Depletion strategies can improve detection of low-abundance proteins, but they add cost and time. In the FIELD trial evaluation, immunodepleting albumin and IgG provided few additional identifications compared to undepleted plasma when using modern mass spectrometry instrumentation. Nanodiamond-based depletion and nanoparticle protein corona enrichment can detect proteins at much lower concentrations. The decision depends on the abundance of the proteins of interest and the available resources.

What is label-free quantification in plasma proteomics?

Label-free quantification is a mass spectrometry approach that measures protein abundance without using isotopic labels. Peptide intensities are compared across samples to determine relative protein abundance. This approach is well suited for discovery experiments but requires careful normalization and imputation of missing values. Deep learning imputation methods can recover biologically meaningful proteins that would otherwise be excluded from analysis.

How do I choose between mass spectrometry and affinity-based platforms?

Mass spectrometry provides nonbiased characterization of thousands of proteins and can identify low-abundance proteins. Affinity-based platforms such as Olink and SomaScan measure specific proteins with high sensitivity and throughput. A comparative evaluation found complementary proteome coverage and moderate quantitative agreement between platforms. Combining both approaches can provide more comprehensive and reliable plasma proteome profiling.

What is the role of automation in plasma proteomics?

Automation is necessary to increase sample processing throughput for large-scale clinical analyses. Robotic liquid handler systems reduce variation from sample preparation steps and enable detection of low-abundant proteins while providing low sample error and increased sample throughput. Automated sample preparation combined with micro-flow LC-MS can analyze dozens of samples simultaneously with high reproducibility.

How should I handle missing values in plasma proteomics data?

Missing values are common in label-free quantification because low-abundance proteins may fall below the detection limit in some samples. Imputation techniques replace missing measurements with estimated values. Deep learning approaches such as variational autoencoders can recover biologically meaningful proteins that would otherwise be excluded from analysis. The choice of imputation method should be documented and justified.

What quality controls should I include in a plasma proteomics study?

Include a pooled plasma external standard in every batch to monitor instrument performance. Assess sample quality using marker panels that detect hemolysis, platelet activation, and other pre-analytical artifacts. Record the anticoagulant type, collection time, processing time, storage conditions, and freeze-thaw cycles for every sample. The online resource at plasmaproteomeprofiling.org provides a means to assess overall sample-related bias in clinical studies.

Related Bioinformatics Guides

• Network Pharmacology Approaches to Multi-Target Drug Discovery
• The Bioinformatics Revolution in Structural Proteomics and Computational Drug Discovery: A Unified Paradigm
• Genomic Selection in Animal Breeding
• Multi-Omics Integration Strategies
• The Rise of Omics: Genomics, Proteomics, and Metabolomics

References and Further Reading

• EMBL-EBI Training. European Bioinformatics Institute.
• NCBI Data Resources. National Center for Biotechnology Information.
• Genomic Data Sharing Policy. National Institutes of Health.
• The FAIR Guiding Principles. Scientific Data.
• Advancements in automation for plasma proteomics sample preparation.. Molecular omics, 2022.
• Plasma Proteome Profiling to detect and avoid sample-related biases in biomarker studies.. EMBO molecular medicine, 2019.
• Seminal Plasma Proteomics Using Filter-Aided Sample Preparation.. Methods in molecular biology (Clifton, N.J.), 2025.
• Nanodiamond for Sample Preparation in Proteomics.. Analytical chemistry, 2019.
• Evaluation of Sample Preparation Strategies for Human Milk and Plasma Proteomics.. Molecules (Basel, Switzerland), 2021.
• Combination of automated sample preparation and micro-flow LC-MS for high-throughput plasma proteomics.. Clinical proteomics, 2023.
• Plasma-Derived Extracellular Vesicle Proteomics.. Journal of proteome research, 2025.
• Optimised plasma sample preparation and LC-MS analysis to support large-scale proteomic analysis of clinical trial specimens: Application to the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) trial.. Proteomics. Clinical applications, 2023.
• Towards a proteomic plasma biomarker panel for diagnosing vasculitis remission.. 2026.
• Leveraging nanoparticle protein corona to advance plasma proteome profiling.. 2026.
• Proteomics as a theranostic compass in BCR::ABL1-negative myeloproliferative neoplasms: Integrating biomarker discovery with therapeutic stratification.. 2026.
• Comparative evaluation of Olink Explore 3072 and mass spectrometry with peptide fractionation for plasma proteomics. Communications Chemistry, 2025.
• Imputation of label-free quantitative mass spectrometry-based proteomics data using self-supervised deep learning. Nature Communications, 2024.
• Recent progress in mass spectrometry-based urinary proteomics. Clinical Proteomics, 2024.
• Integrated proteomics sample preparation and fractionation: Method development and applications. Trac Trends in Analytical Chemistry, 2019.

This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.