Biobanks, Proteomics, and AI Converge To Advance Precision Medicine

Precision medicine increasingly depends on the ability to connect molecular data with real-world patient outcomes at scale. As research moves beyond isolated datasets, biobanks have become central to this shift—enabling longitudinal analysis, multiomic integration, and population-level discovery that can inform earlier diagnosis and more targeted therapies.

At the same time, advances in proteomics and AI are unlocking new layers of biological insight, particularly in understanding how disease develops before symptoms emerge. Together, these capabilities are reshaping how researchers generate, validate, and translate evidence into clinical practice.

In this interview, Dr. Yan Zhang, president of the proteomic sciences business at Thermo Fisher Scientific, discusses how these trends are converging and what is needed to fully realize the potential of population-scale biology.

Biobanks enable longitudinal, real-world disease insights at scale

“Biobanks are foundational to precision medicine at scale,” Zhang explained, emphasizing that “by housing deeply annotated samples alongside longitudinal clinical records, biobanks allow researchers to track disease progression over time.” She noted that this depth of information enables “a shift beyond single snapshots to generate the kind of robust, reproducible evidence that translates into real clinical impact.”

Zhang highlighted how “the sheer magnitude of scale and depth of data that modern biobanks can support” is transforming research capabilities.

Initiatives such as the UK Biobank Pharma Proteomics Project (UKB-PPP) reflect this shift, analyzing more than 5,400 proteins across 600,000 samples. At this level, “the connection between molecular biology and population health becomes actionable,” Zhang said.

Why biobanks matter:

• Biobanks provide longitudinal datasets that improve understanding of disease progression
• Scale allows researchers to generate reproducible, clinically relevant evidence
• Multiomic integration connects molecular biology with population health outcomes

National biobanks are becoming digitally integrated research ecosystems

How do you see the role of national biobanks evolving over the next few years?

“National biobanks are already becoming core research infrastructure—they are automated, digitally connected, and directly integrated into clinical systems,” Zhang noted. She added that this integration enables “continued participant engagement and data enrichment over time,” reflecting a fundamental shift in how data is collected and used.

Large-scale programs such as Singapore’s PRECISE-SG100K initiative highlight this evolution. The study is profiling 100,000 volunteers across a uniquely diverse cohort, using complementary proteomics technologies to strengthen outcomes.

Zhang also highlighted the growing importance of large collaborative datasets, noting that recent research integrating genetic and proteomic data from more than 78,000 individuals has identified over 24,000 protein quantitative trait loci across more than 1,100 proteins, providing new insight into biological mechanisms.

Looking ahead, Zhang emphasized convergence across technologies. “Much like genome sequencing moved from research tool to clinical standard, biobanks are on the same trajectory, with proteomics playing a growing role in uncovering disease biology at a speed and scale that wasn't possible before and enabling more precise patient stratification, clinical trial recruitment, and translational research.” 

Key trends shaping the future of national biobanks:

• National biobanks are evolving into automated, interoperable research ecosystems
• Integration with clinical systems enables continuous data growth and relevance
• Multiomic and AI convergence will define the next phase of biobank impact

Proteomics unlocks dynamic biology and enables earlier disease detection

What can proteins tell us about disease biology that genes simply can't?

Tech Spotlight: High-Throughput Screening 2026

Explore reliable high-throughput screening workflows for vaccine and antibody discovery, covering assay design, validation and data analysis to improve reproducibility, data quality and candidate prioritization.

View Webinar

Advertisement

Proteomics provides a dynamic view of biology that complements the static insights offered by genomics. While genes indicate predisposition to disease, proteins reflect real-time physiological changes. This dynamic capability makes proteomics particularly valuable for early detection.

“When you can characterize thousands of proteins simultaneously, you can see systemic patterns before they surface clinically,” Zhang explained.

She highlighted how research from the UKB-PPP pilot program, which identified protein risk factors for cancer up to seven years before diagnosis, represents “a shift toward identifying biological signatures of disease before symptoms even appear.”

Population-scale analysis is critical to making these insights actionable. As Zhang explained, “being able to study proteins at the population-scale gives researchers the sample volume needed to find real disease signals, identify rare but meaningful protein patterns, and build models that can be translated across diverse real-world populations.”

Why this matters for early detection:

• Proteomics captures real-time biological changes rather than static risk
• Protein biomarkers can signal disease years before clinical diagnosis
• Large datasets improve the detection of rare signals and strengthen predictive models

Population diversity determines clinical relevance of proteomic models

Why is population diversity critical in proteomics research?

Zhang stressed that “population diversity is critical because biological variation is not uniform across populations.” She explained that “protein expression patterns and disease prevalence can be vastly different.”

Circulating Immune Proteins in Liver Cancer Research

This webinar explores how circulating immune-related proteins can improve our understanding of hepatocellular and gallbladder cancer risk across diverse populations with viral hepatitis.
View Webinar

Advertisement

“A proteomic model based predominantly on one population may not translate to another,” she cautioned, limiting its utility for broad clinical implementation. This limitation extends to multiple downstream applications, including diagnostics, drug targets, and patient stratification.

Conversely, capturing diversity at scale enables more meaningful insights. Zhang noted that “molecular heterogeneity across populations, when captured at sufficient scale like in the PRECISE biobank, can help reveal aspects of disease biology that are more broadly meaningful for therapeutic development and support the development of diagnostics, therapeutics, and risk models that are more generalizable across real-world populations.”

Why diversity shapes outcomes:

• Biological variation across populations impacts model performance
• Lack of diversity can limit diagnostic and therapeutic applicability
• Diverse cohorts improve generalizability and clinical relevance

Standardization and reproducibility are essential for regulatory adoption

Why is reproducibility so important in population-scale biology?

For discoveries to influence clinical care, they must be reproducible across datasets, institutions, and populations. Variability in methods and data quality can undermine confidence in findings.

“Reproducibility is critical because scientists need to be able to turn compelling research results into evidence strong enough to guide clinical and regulatory decisions,” Zhang said. “Findings must hold up when tested in different settings, on different populations, and across platforms.”

She highlighted that findings may fail if “sample handling protocols differ, data systems are incompatible, or governance frameworks don't allow meaningful data sharing.”

Unified frameworks can help address these challenges. Zhang explained that harmonized protocols can help make sample collection, processing, and annotation more consistent.

How Agentic AI Is Transforming the Modern Lab

This whitepaper examines how agentic AI is reshaping laboratory operations and what this shift means for the future of scientific discovery.
View Whitepaper

Advertisement

She also emphasized interoperability, noting that “interoperable data systems can also help researchers across geographies combine datasets without losing integrity,” enabling broader validation efforts. These elements, she concluded, “must be foundational for precision medicine to become the standard of care.”

Core lessons for translation and regulation:

• Reproducibility underpins clinical and regulatory confidence
• Standardized protocols ensure consistency across studies
• Interoperable systems enable large-scale validation

Bridging the gap between discovery and clinical implementation

What challenges exist in realizing the potential of population-scale proteomics research?

Zhang described the biggest challenge as “the gap between insight generation and translation.” While technologies have advanced rapidly, researchers must still build systems that support validation and implementation at scale.

She noted that infrastructure must include “the ability to recontact participants, validate findings across diverse populations, and support clinical trial recruitment.” At the same time, sustained investment is essential.

Governance is another key factor. Zhang emphasized the need for “frameworks that allow for responsible and streamlined data sharing and AI governance,” alongside regulatory guidance to safely translate insights into patient care.

Long-term data quality also matters. Samples and records must be collected with standardized metadata, quality metrics, and longitudinal clinical context, ensuring that they remain useful as technologies evolve.

Barriers and priorities moving forward:

• Translation from discovery to clinic remains a key bottleneck
• Infrastructure for validation and trials is essential
• Long-term investment and governance frameworks are critical

Episode 4: Acceleration of Process Development & Characterization With Hybrid Modeling & Transfer Learning

Discover how hybrid modeling and transfer learning are transforming therapeutic antibody process development. Learn how machine learning enables knowledge reuse to improve process understanding, reduce experimental effort and accelerate development.
View Webinar

Advertisement

AI and machine learning extract meaning from complex biobank data

What role can AI and machine learning play in biobanking?

AI is increasingly central to unlocking the value of large-scale, complex datasets generated by modern biobanks. Traditional analytical approaches struggle with the volume and dimensionality of these data.

According to Zhang, “AI's most important role in biobanking is making sense of the data that already exists.”

“Applying AI to biobank data can help accelerate drug target discovery , improve biomarker identification, enable more accurate patient stratification, and identify participants most likely to benefit from specific therapies or clinical trials,” she added.

How AI is reshaping analysis:

• AI enables pattern recognition in complex, high-dimensional datasets
• Machine learning accelerates drug discovery and biomarker development
• Advanced analytics improve patient stratification and trial design

This content includes text that has been created with the assistance of generative AI and has undergone editorial review before publishing. Technology Networks' AI policy can be found here.