Advances in Biomarker Discovery and Analysis
Learn how proteomics, AI and microsampling are transforming the way we identify and analyze biomarkers.
Complete the form below to unlock access to ALL audio articles.
Biomarkers can offer critical information about an individual’s health status, disease prognosis or how they might respond to a treatment.
Developments in high-throughput technologies that enable the efficient, sensitive and large-scale analysis of biological samples have significantly advanced biomarker research over recent years.
In this article, we highlight just some of the research trends impacting biomarker discovery and analysis, including advances in proteomics research, artificial intelligence (AI) and microsampling.
Proteomics powers biomarker discovery
Looking to the future of biomarker research, Dr. Irene De Biase, professor of clinical pathology at the University of Utah, expects to see an exponential growth in studies that use omic approaches to evaluate large datasets to identify correlations between biomarkers and clinical outcomes: “For instance, three months into 2025, and there are already almost 300 publications using proteomics in various cohorts of patients with diabetes mellitus,” she told Technology Networks.*
Proteins drive cellular functions, earning their nickname as the “workhorses” of the cell. Proteomics, the large-scale study of the proteome, has emerged as a powerful field for biomarker discovery in the post-genomic era.
What is the proteome?
The proteome refers to the entire set of proteins found in a cell, tissue or organism at a specific point in time. While an individual’s genome is relatively static throughout their lifetime, the proteome is dynamic and actively responds to both external and internal signals.
Liquid chromatography-mass spectrometry (LC-MS) is the primary method used for protein biomarker discovery due to its high level of sensitivity and specificity. LC-MS has enabled researchers to detect disease-specific protein signatures, analyze post-translational modifications (PTMs) that affect disease mechanisms and discover novel biomarkers for conditions such as cancer and neurodegenerative diseases.1,2,3
As the demand for higher throughput and scalability in biomarker research grows, recent trends in LC-MS proteomics analysis are focused on improving and automating workflows, as well as refining sample preparation and analytical processes.4 Kverneland et al. recently published a fully automated workflow that combines sample digestion, cleanup and loading, and demonstrated its use for processing 192 HeLa cell samples in 6 hours.5 “The workflow is optimized for minimal sample starting amount to reduce the costs for reagents needed for sample preparation, which is critical when analyzing large biological cohorts,” the researchers said.
While blood is the most commonly analyzed biofluid due to its accessibility, valuable insights can also be gained from other fluids such as urine, cerebrospinal fluid (CSF), serum and breast milk. A recent paper by Zhang et al. outlined an automated, scalable workflow for preparing hundreds to thousands of samples, including milk and urine, with only minor modifications to the initial steps of a workflow originally designed for blood plasma, serum and CSF.6 This approach expands the scope of biomarker discovery by enabling the automated analysis of a broader range of biological samples.
Advances in protein enrichment and depletion methods are also enhancing MS-based proteomics. High abundance proteins in samples such as plasma can mask the signal of low-abundance – but biologically interesting – proteins. A variety of enrichment methods are being implemented to support the discovery and identification of low-abundance proteins that carry biomarker potential.7 For example, Palstrøm et al. recently performed a MS-based proteomic analysis of plasma samples that were enriched using magnetic p-aminobenzamidine (ABA) affinity probes.8 The samples were gathered from 45 patients with abdominal aortic aneurysm (AAA) – a life-threatening disorder that sees improved outcomes with a fast diagnosis – and 45 matched controls. The researchers identified plasma proteome alterations linked to AAA and, using machine learning, developed a potential biomarker panel for earlier and more accurate diagnosis.
Beyond MS-based approaches, affinity-based techniques are increasingly being used to analyze the proteome and uncover clinically relevant biomarkers. These techniques rely on specific interactions between proteins and binding agents, including aptamers and/or antibodies.
The UK Biobank recently announced the launch of the world’s “most comprehensive study of the proteins circulating in our bodies”. The project will harness an antibody-based platform to measure up to 5,400 proteins in 600,000 samples, half a million of which will be obtained from UK Biobank participants, while 100,000 samples will be taken from the same volunteers up to 15 years later.
"Adding proteomic data for the full UK Biobank cohort will be an absolute game changer for prediction of disease onset and prognosis, particularly for the many neglected diseases for which good prospective data are lacking,” Professor Claudia Langenberg, director of the Precision Healthcare University Research Institute at Queen Mary University of London, said. “These include debilitating and life-threating diseases, such as polycystic ovary syndrome and motor neurone disease. Just imagine if we could detect these and many other conditions much earlier than is currently possible."
A recent pre-print study by Kirsher et al. compared different plasma proteomics platforms – including aptamer-, antibody- and MS-based approaches by applying these methods to the same cohort of samples covering 13,000 proteins.9** Though the study is yet to be peer-reviewed, it suggests that trade-offs in coverage across these platforms have implications for the future of biomarker discovery and translational research.
Beyond proteomics, other omics fields – such as genomics, transcriptomics and metabolomics – continue to drive biomarker discovery, with growing efforts now focused on integrating these data through multiomics approaches to gain a more comprehensive view of disease biology.
AI and machine learning enhances biomarker research
AI-based approaches, such as deep learning and machine learning, are finding a variety of applications in biomarker discovery, development and validation. Such technologies can process and integrate vast, complex multiomics datasets more efficiently and at a higher capacity than traditional analytical tools.
In 2022, researchers published a novel pan-cancer proteomic map of 949 human cell lines across over 40 types of cancer.10 They used a deep learning-based computational pipeline – Deep DeeProM – to integrate large volumes of data. “DeeProM enabled the full integration of proteomic data with drug responses and CRISPR-Cas9 gene essentiality screens to build a comprehensive map of protein-specific biomarkers of cancer vulnerabilities that are essential for cancer cell survival and growth,” co-author Associate Professor Qing Zhong of the Children’s Medical Research Institute, University of Sydney, told Technology Networks. The analyses identified biomarkers detectable only at the proteomic level, showing enhanced predictive accuracy compared to models that rely solely on gene expression data.
Deep learning has also shown success in analyzing histopathology slides to facilitate the discovery of novel biomarkers. A recent preprint by Shulman et al. presents Path2Space, a deep learning model that predicts spatial gene expression from histopathology slides of breast cancer tumors.11** The researchers aimed to identify microenvironment characteristics and other spatial biomarkers associated with treatment response.
When applied to two large cohorts of breast cancer patients treated with trastuzumab and chemotherapy, Path2Space was able to identify spatial biomarkers and predicted therapy response with high accuracy.
While these examples showcase AI’s utility in oncology biomarker research, such technologies are also facilitating new biomarker discoveries across diseases including Alzheimer’s, diabetes and cardiovascular disease, among others.12,13,14
Remote sampling devices: The future of biomarker analysis?
Measuring blood biomarkers in clinical and research settings presents challenges, including geographical limitations, the cost and inconvenience of in-clinic venipuncture and infrequent sampling.
Microsampling is an emerging, minimally invasive alternative approach that enables the collection of samples in a remote location for subsequent laboratory analysis using remote sampling devices, of which there are now several types.15
Dr. Michael Snyder, Stanford W. Ascherman Professor of Genetics at Stanford Medicine, summarized why microsampling is an attractive approach to profiling changes in health: “It lets you measure hundreds to thousands of molecules with good accuracy, it’s convenient and the samples are collected in a natural setting – i.e., a person’s home – which should give a better indication of their true biochemical state rather than using samples collected in a clinic.
“Microsampling can help science and medicine become more inclusive, reduce costs for healthcare systems and help people access important insights about their health,” Dr. Jennifer Van Eyk, professor of cardiology, director of the Advanced Clinical Biosystems Institute and the Erika Glazer Endowed Chair in Women’s Heart at Cedars-Sinai Medical Center, said. Van Eyk has been developing MS-based workflows for remote sampling devices, and comparing their viability to existing clinical sampling methods, for several years.16
Snyder and colleagues recently published a strategy for regularly capturing and analyzing thousands of metabolites, lipids, cytokines and proteins from 10 μl of blood.17 “In two proof-of-principle studies, we first demonstrate the profiling of a dynamic response to ingestion of a mixed meal shake and discover high heterogeneity in individual metabolic and immune responses, and second, we perform high-resolution profiling of an individual over one week enabling the identification and quantification of thousands of molecular changes and associations across ‘omes’ at a personal level,” the authors said.
Snyder said that his team runs most of his studies using a microsampling approach now. “I think many, if not most, studies will be run this way in the future. It will also be used for home health monitoring,” Snyder said.
A recent review by Protti et al. emphasized how microsampling technologies hold the potential to democratize access to high-quality analytical testing.15 “However, for the very same reasons, to obtain a transformative effect on a wide range of analytical applications for both clinical and non-clinical purposes, these innovative microsampling techniques must first achieve wide acknowledgment and adoption, which in turn can only be obtained through consistently effected commercial availability, as well as regulatory approval for the intended use, in most countries and regions,” they said.
“It’s going to be a long road, but I think we have a responsibility to help people access their data and have some level of control over their health,” Van Eyk said.
The next era of biomarker innovation
Biomarker research is evolving rapidly. While the trends and innovations highlighted in this article are by no means an exhaustive list, they represent increasingly precise, scalable and accessible approaches to understanding human health and disease. As multiomics integration and decentralized data collection continue to advance, the future of biomarker discovery holds great potential for transforming diagnostics, monitoring wellness and personalizing treatments across a wide range of conditions.
*Interview completed in March 2025.
** This article is based on research findings that are yet to be peer-reviewed. Results are therefore regarded as preliminary and should be interpreted as such. Find out about the role of the peer review process in research here. For further information, please contact the cited source.
References:
1. Gajula SNR, Khairnar AS, Jock P, et al. LC-MS/MS: A sensitive and selective analytical technique to detect COVID-19 protein biomarkers in the early disease stage. Expert Rev Proteomics. 2023;20(1-3):5-18. doi: 10.1080/14789450.2023.2191845
2. Jang HN, Moon SJ, Jung KC, et al. Mass spectrometry-based proteomic discovery of prognostic biomarkers in adrenal cortical carcinoma. Cancers. 2021;13(15). doi: 10.3390/cancers13153890
3. Tao QQ, Cai X, Xue YY, et al. Alzheimer’s disease early diagnostic and staging biomarkers revealed by large-scale cerebrospinal fluid and serum proteomic profiling. The Innovation. 2024;5(1):100544. doi: 10.1016/j.xinn.2023.100544
4. Ye X, Cui X, Zhang L, et al. Combination of automated sample preparation and micro-flow LC–MS for high-throughput plasma proteomics. Clin Proteom. 2023;20(1):3. doi: 10.1186/s12014-022-09390-w
5. Kverneland AH, Harking F, Vej-Nielsen JM, et al. Fully automated workflow for integrated sample digestion and evotip loading enabling high-throughput clinical proteomics. Mol Cell Proteom. 2024;23(7). doi: 10.1016/j.mcpro.2024.100790
6. Pedersen AL, Ernest M, Affolter M, Dayon L. Analyzing various biological fluids with a single automated proteomic workflow for biomarker discovery. In: Islam Williams T, ed. Tissue Proteomics: Methods and Protocols. Springer US; 2025:157-178. doi: 10.1007/978-1-0716-4298-6_11
7. Boschetti E, Righetti PG. Low-abundance protein enrichment for medical applications: the involvement of combinatorial peptide library technique. Int J Mol Sci. 2023;24(12):10329. 2023. doi: 10.3390/ijms241210329
8. Palstrøm NB, Nielsen KB, Campbell AJ, et al. Affinity-enriched plasma proteomics for biomarker discovery in abdominal aortic aneurysms. Proteomes. 2024;12(4). doi: 10.3390/proteomes12040037
9. Kirsher DY, Chand S, Phong A, Nguyen B, Szoke BG, Ahadi S. The current landscape of plasma proteomics: technical advances, biological insights, and biomarker discovery. bioRxiv. 2025. doi: 10.1101/2025.02.14.638375
10. Gonçalves E, Poulos RC, Cai Z, et al. Pan-cancer proteomic map of 949 human cell lines. Cancer Cell. 2022;40(8):835-849.e8. doi: 10.1016/j.ccell.2022.06.010
11. Shulman ED, Campagnolo EM, Lodha R, et al. Path2space: An ai approach for cancer biomarker discovery via histopathology inferred spatial transcriptomics. bioRxiv. 2024. doi: 10.1101/2024.10.16.618609
12. Winchester LM, Harshfield EL, Shi L, et al. Artificial intelligence for biomarker discovery in Alzheimer's disease and dementia. Alzheimers Dement. 2023;19(12):5860-5871. doi: 10.1002/alz.13390
13. Khan S, Mohsen F, Shah Z. Genetic biomarkers and machine learning techniques for predicting diabetes: systematic review. Artif Intell Rev. 2024;58(2):41. doi: 10.1007/s10462-024-11020-w
14. DeGroat W, Abdelhalim H, Patel K, Mendhe D, Zeeshan S, Ahmed Z. Discovering biomarkers associated and predicting cardiovascular disease with high accuracy using a novel nexus of machine learning techniques for precision medicine. Sci Rep. 2024;14(1):1. doi: 10.1038/s41598-023-50600-8
15. Protti M, Milandri E, Di Lecce R, Mercolini L, Mandrioli R. New trends in bioanalysis sampling and pretreatment: How modern microsampling is revolutionising the field. Adv Sample Prep. 2025;13:100161. doi: 10.1016/j.sampre.2025.100161
16. Whelan SA, Hendricks N, Dwight ZL, et al. Assessment of a 60-biomarker health surveillance panel (hsp) on whole blood from remote sampling devices by targeted LC/MRM-MS and discovery DIA-MS analysis. Anal Chem. 2023;95(29):11007-11018. doi: 10.1021/acs.analchem.3c01189
17. Shen X, Kellogg R, Panyard DJ, et al. Multi-omics microsampling for the profiling of lifestyle-associated changes in health. Nat Biomed Eng. 2024;8(1):11-29. doi: 10.1038/s41551-022-00999-8
