Yan Zhang, President of proteomic sciences at Thermo Fisher Scientific, writes about developments in precision medicine and how biobanks are key to realising this therapy area’s full impact. 

While a one-size-fits-all approach to patient care is still common in many areas, there have been meaningful strides toward more personalised healthcare in recent years. Targeted therapies have reshaped outcomes for many cancers. Companion diagnostics have moved from concept to standard of care in oncology with usage is increasing in other disease areas, too. The broader goal of making precision medicine work on a global scale for a full spectrum of disease, however, has remained stubbornly out of reach.  

Today, we are on the precipice of change. The convergence of technological innovation and national population-scale biobanking initiatives have made real-time disease biology for earlier detection, prevention and personalised healthcare a reality.

High-plex proteomics is having its genomics moment yet with a key differentiator. Its impact is expected to be much broader in the number of diseases influenced, and the testing to be real-time and recurring since the bio-fluid based proteomic profiles are dynamic and longitudinal monitoring now becomes feasible.

There are many parallels between proteomic and genomic discovery. For years, sequencing a single genome was expensive, slow and technically demanding. Advances in technology increased throughput and the field crossed a critical threshold when population-scale studies became feasible. The science evolved because the technology did. Proteomics is crossing that same threshold now. 

Proteins are the functional layer of human biology. Unlike the largely static genome, the proteome is dynamic. It responds to disease, treatment, environmental conditions and time and age, reflecting what’s happening in the body in real time and not just predisposition. For years, studying proteins at scale was impractical due to small panels, low throughput and overall investment. However, advanced technologies, such as mass spectrometry, and multiplexed affinity-based platforms have changed that. 

Researchers can now measure thousands of proteins from a single blood sample and process those samples at a scale that makes biobank-linked discovery studies genuinely tractable. What used to take a decade — such as mapping protein-disease associations or identifying genetically supported drug targets — can happen in months today. It’s a massive compression that’s changing timelines, economics and the ambition of what’s possible in precision medicine. 

Biobanks as strategic infrastructure  

Discovery at scale requires the availability of samples at scale, all linked to longitudinal health records, disease progression data, treatment outcomes and genomic context collected over years. This model reflects what we often find in a modern biobank, with more robust clinical insights driving a shift in what precision medicine can accomplish. Nations are increasingly investing in population-scale studies that expand capacity for discovery, validation and translation, often leveraging biobanks to deliver the clinical insight needed to accelerate this research. 

There are several highly visible examples of biobanks in action. The UK Biobank Pharma Proteomics Project (UKB-PPP) is tackling the largest human proteomics study of its kind, analysing more than 5,400 proteins across 600,000 samples. Results from the program’s pilot phase has already been cited in more than 400 publications (1) and even enabled identification of protein risk factors for cancer a median of twelve years before diagnosis.  

Precision Health Research Singapore (PRECISE) is undertaking the region’s most ambitious and diverse population-scale biobank initiative, profiling 100,000 volunteers and integrating multiomic data with long-term health records, to identify which biomarkers are consistent across large cohorts. The diversity of this biobank is intentional: for new diagnostics and therapies to translate across real-world populations, the underlying research needs to reflect them. 

Biobanks also help remove traditional barriers to scientific resources. In the largest multi-cohort proteogenomics study to date (2), an international collaboration of 118 investigators across 89 institutions integrated genetic and blood protein data from more than 78,000 individuals across 38 cohorts, identifying more than 24,000 genetic links to over 1,100 proteins. The Systematic and Combined AnaLysis of Olink Proteins (SCALLOP) study moved beyond statistical association toward biological mechanism, opening new pathways for drug target discovery, drug repurposing and biomarker identification, because scientists were able to combine and compare data across samples worldwide.  

Biobanks are the research infrastructure for the next decade of precision medicine, underpinning biomarker discovery, accelerating therapeutic target identification and enabling disease risk modeling at a scale that was simply not possible before. For example, biomarker panels identified through population-scale proteomics are accelerating candidate selection in oncology and cardiovascular disease. Further along in development pipelines, biobank data are shrinking the distance between discovery and Investigational New Drug (IND)-enabling studies, changing what’s commercially viable to pursue.  Pharmaceutical companies are already using this data to revive and reposition drug programs. The value of biobank infrastructure compounds as science evolves. 

The multiomics layer and the role of AI  

Proteomics generates the most powerful insights when integrated with other molecular data. The most informative studies now combine protein measurements with genomics, transcriptomics and clinical phenotyping to build a systems-level understanding of disease. With longitudinal clinical records, these data layers produce a picture of disease that is both mechanistically grounded and clinically relevant and enables the move from population-level association to patient decision-making. 

When it comes to data interpretation, AI is consequential. Population-scale datasets exceed what traditional analytical approaches can resolve. AI enables researchers to identify patterns in protein structures with disease subtypes, predict therapeutic response, and stratify patient populations much faster than manual analysis. When applied responsibly, AI can extract biologically meaningful interpretations from the data that translate to clinical impact. 

Closing the precision medicine gap  

Precision medicine has long been hindered by the field’s ability to generate insights faster than they can be validated and implemented. Biomarker discovery, for example, has accelerated considerably with advances in instrument sensitivity, but clinical deployment hasn’t been able to keep up.  

For precision medicine to be a reality – at scale – it needs to function as a continuous cycle that connects discovery, validation, regulatory alignment and clinical use. Closing that gap requires treating biobanking as essential health infrastructure, with the same sustained investment, governance and interoperability standards that are applied to clinical systems. It also requires harmonised protocols that allow data to move responsibly across institutions and geographies and coordinated collaboration across disciplines and sectors that the scale of the problem demands. 

In the next decade, biobanks will become standard infrastructure in translational and clinical workflows, much like how genome sequencing transitioned from research instrument to clinical tool. Proteomics will play a growing role in understanding the dynamic biology of disease. The convergence of multiomic integration and AI-driven interpretation will reshape how drug targets are identified, how patient populations are stratified and how we think about healthcare from prevention to treatment. 

The promise of precision medicine still holds true. What’s changing is our capacity to scale novel insights from the research bench to the clinic. The next decade will be defined not by whether precision medicine is possible, but by how effectively we build the infrastructure, data ecosystems and collaborative frameworks required to deliver it at population scale. 

About the author: 

Dr Zhang has served as President of the proteomic sciences business since August 2025. Formerly known as Olink, this business was acquired by Thermo Fisher Scientific in 2024. She is a global veteran of the life sciences industry, with extensive leadership experience spanning business management and commercial strategy. 

References 

1:https://www.nature.com/articles/s41586-023-06592-6 

2: https://www.cell.com/cell/fulltext/S0092-8674(26)00385-5  

From DDW Volume 27 – Issue 3, Summer 2026 – Read the digital issue here