Multi-omics

Dalia Daujotyte at Illumina

As scientists continue to explore the vast web of life’s molecular setups, it has become increasingly clear that a singular approach, focusing narrowly on genomics, transcriptomics or proteomics, can no longer encompass the full picture of cellular biology. The integration of multi-omics has emerged as a transformative methodology, combining the study of DNA, RNA and protein layers along with metabolites to illuminate the complexities of life at an unprecedented depth.

Proteins are the cornerstone of biological systems – from functional players in countless physiological processes to life-saving drug targets. The applications of protein analysis stretch across diagnostics, disease monitoring, biomarker discovery, therapeutic development and more, making it crucial to explore the proteome to advance our understanding of human health overall.

Proteomics has entered a function-first era. Rather than treating proteins as static structures of single drug targets, leading research now frames disease as the result of dynamic protein interaction networks, transient complexes and context-dependent functional states.

Breakthroughs in protein-protein interaction profiling, structural biology advances (eg, cryo-electron microscopy [cryo-EM], artificial intelligence [AI]-assisted modelling), systems biology and, recently, multi-omic approaches, are influencing how the pharma industry discovers targets, validates mechanisms, and designs both small molecule and biological therapeutics.

The genome defines what is possible, but the proteome – through alternative isoforms, posttranslational modifications, structure and dynamics – defines what is actually true in a cell at any moment. The genome is information-dense but largely static, whereas proteome is where complexity explodes: one gene may produce many proteoforms due to alternative splicing and post-translational modifications.1 Typically, DNA and RNA provide a framework for understanding biology, acting as surrogates when it comes to researching a concerted (functional) protein role. However, across organisms and tissues, mRNA abundance and protein abundance are correlated only modestly, and in a contextdependent manner. Transcript levels typically explain a minority of the variance in steadystate protein levels, and changes in mRNA often do not reliably predict changes in proteins, especially across conditions, time points or perturbations. This finding has been replicated for over two decades and is now considered a feature of biological regulation – not a measurement artifact.2 RNA tells us what the cell is trying to do; proteins tell us what the cell is actually doing. While RNA can be used for hypothesis generation, understanding protein levels is essential for biomarker validation and go/no-go decisions, which underscores the need to directly analyse the protein profile for a deeper insight into cellular function.3

Although individual interrogations of DNA, RNA or proteins can yield significant findings, it is the synthesis of these approaches that unlocks their full potential. The hallmark of multi-omics is its capacity to bridge multiple molecular data sets – including genomics, transcriptomics, proteomics and beyond – and provide a more holistic perspective of biological systems and driving function-first discovery.

Multi-omics is becoming increasingly relevant in medicine, where it can provide a more comprehensive understanding of complex diseases – cancer, as well as metabolic, cardiovascular, neurological and autoimmune disorders – by uncovering underlying molecular pathways and identifying potential therapeutic targets.

Proteomics-anchored multi-omics studies meaningfully advance our understanding of Alzheimer’s disease by linking genetic and transcriptomic signals to functional protein changes, as presented in a recent review of 27 studies.4

Several studies position non-alcoholic steatohepatitis (NASH) as a systemslevel disease and show that only integrated multi-omics, anchored by proteomics, can capture the complexity and functional disease biology needed for reliable diagnosis, prognosis and biomarker discovery.5 Moreover, highplex serum proteomics can noninvasively reproduce liver biopsy features of non-alcoholic fatty liver disease, and sensitively track treatmentinduced histological changes, supporting a future ‘liquid biopsy’ approach to NASH.6,7

In cancer, multi-omics integration increases clinical relevance and specificity. For example, this study establishes a broad proteogenomic atlas of cancer vulnerabilities, highlighting many new potential drug and peptide targets that could inform both small molecule therapy and immunotherapy, significantly expanding on the relatively limited set of targets addressed by existing cancer treatments.8 Another example demonstrates that inherited genetic variation significantly influences the protein landscape of tumours across cancers, shaping tumour behaviour and clinical features. By bridging genetics and proteomics, the study opens new avenues for personalised oncology that consider both germline and somatic tumour biology.9

While integrated approaches help uncover insights relevant to disease progression and clinical translation, there is still a need for improved methodological frameworks and better interpretation of large heterogeneous data sets to fully realise the potential of multi-omics.

The advancements that propel proteomics and multi-omics forward are rooted in cutting-edge technologies and analytical tools. Historically, drug discovery focused on individual proteins with enzymatic active sites. However, research shows that many diseases are driven by aberrant protein complexes and interaction networks, not isolated proteins. A comprehensive review in Cell summarises decades of protein-protein interactions research and shows how modern techniques – affinitypurification mass spectrometry (AP-MS), crosslinking MS, cryoEM and AIbased modelling – now allow systematic identification of diseaserelevant protein complexes rather than anecdotal interactions.10 Protein function is dynamic, structural biology no longer treats proteins as static entities, and proteins lacking stable 3D structure – once dismissed as ‘undruggable’ – are now recognised and important regulators of cellular organisation.11

Modern drug discovery is being reshaped by proteincentric discovery technologies that move beyond static target lists. Affinitybased proteomics, high-resolution MS, chemoproteomics and integrated multi-omics using next-generation sequencing now enable direct, systemwide measurement of protein function, interactions and mechanisms of action in biologically relevant contexts.

These tools, combined with groundbreaking machine learning algorithms like AlphaFold, which predicts protein folding structures with remarkable precision, are revolutionising the field.

However, with these advancements comes an immense challenge: managing and analysing the vast amounts of data generated. This is where computational platforms are crucial for deriving actionable insights, as they translate complex data sets into meaningful relationships between biomolecules. Integrating omics-derived data into predictive models is becoming increasingly feasible, offering the possibility of tailoring therapeutic interventions based on individual molecular profiles.

To understand function, we need integrated insights across -omes and modalities. Large studies pave the way to deeper insights. Accessibility, robustness, scalability and reproducibility across cohorts are needed for assays to be used in population studies. Subsequently, responsible deployment of the data and computational integration will lead into clinical readiness and decision-making.

We are still on the cusp of unlocking proteomics’ full potential, with unanswered questions about proteoforms, post-translational modifications, the ‘dark proteome’ roles and specifics, RNA-protein discordance, spatial context and beyond. Despite the progress made, significant challenges remain. These unanswered questions underscore the complexity of cellular systems and the need for further innovation in both experimental methodologies and data analysis algorithms. Increasingly, the primary bottleneck lies less in data generation but in effective multi-omic integration, warranting greater emphasis on analytical approaches. Proteomics has advanced the ability to measure, but not yet the problems of completeness, causality and clinical relevance. The sheer scale of multi-omic data sets presents logistical and analytical hurdles. Current computational infrastructures must evolve to handle the volume, velocity and variety of -omics data. Without robust, scalable solutions capable of integrating and interpreting this information, researchers risk missing critical insights embedded in the data.

“Progress in science depends on new techniques, new discoveries and new ideas, probably in that order…” Dr Sydney Brenner, a distinguished scientist and the 2002 Nobel Prize Laureate in Physiology or Medicine.

The journey to unlock the proteome through multi-omics is not without its challenges, but the opportunities ahead far outweigh them. Future breakthroughs in proteomics promise to transform biology and medicine, offering deeper insights into human health and enabling the development of personalised therapies. Collaboration between biologists, data scientists and technology developers will be pivotal in overcoming the field’s remaining obstacles. As the tools and techniques of multi-omics continue to evolve, we stand poised to uncover the intricate symphony of life’s molecular players. By integrating genomic, transcriptomic and proteomic perspectives, researchers can unlock the very nature of biological discovery.

1. Walsh C (2006), ‘Posttranslational modification of proteins: Expanding nature's inventory’, Englewood (CO): Roberts and Co. Publishers, xxi, 490

2. Buccitelli et al (2020), ‘mRNAs, proteins and the emerging principles of gene expression control’, Nat Rev Genet, 21(10), 630-644

3. Payne S (2014), ‘The utility of protein and mRNA correlation’, Trends Biochem Sci, 40(1), 1-3

4. Visit: mdpi.com/2035-8377/17/12/197

5. Niu L et al (2021), ‘Defining NASH from a Multi-Omics Systems Biology Perspective’, J Clin Med, 10(20), 4673

6. Sanyal A J et al (2023), ‘Defining the serum proteomic signature of hepatic steatosis, inflammation, ballooning and fibrosis in non-alcoholic fatty liver disease’, J Hepatol, 78(4), 693-703

7. Sivakumar P et al (2024), ‘SomaLogic proteomics reveals new biomarkers and provides mechanistic, clinical insights into Acetyl coA Carboxylase (ACC) inhibition in Nonalcoholic Steatohepatitis (NASH)’, Scientific Reports, 14, 17072

8. Savage S R et al (2024), ‘Pan-cancer proteogenomics expands the landscape of therapeutic targets’, Cell, 187(16), 4389-4407

9. Rodrigues F M et al (2025), ‘Precision proteogenomics reveals pan-cancer impact of germline variants’, Cell, 188(9), 2312-2335

10. Greenblatt J F et al (2024), ‘Discovery and significance of protein-protein interactions in health and disease’, Cell, 187(23), 6501-6517

11. Lin Y H et al (2026), ‘Evolution of intrinsically disordered regions in vertebrate galectins for phase separation’, EMBO Rep, 27(5), 1254-1269

Dr Dalia Daujotyte is senior director of Proteomics Product Management at Illumina, where she leads portfolio strategy and partnerships across proteomics, next-generation sequencing applications and multi-omic solutions. Dalia brings a strong scientific foundation with a PhD in biochemistry from Vilnius University, Lithuania, and extensive academic research experience in structural and molecular biology from the Medical Research Council Laboratory of Molecular Biology, Cambridge, UK, studying epigenetics and nucleic acid-protein interactions in RNA transport and localised translation in neurons. At Illumina, Dr Daujotyte combines her scientific expertise with product leadership to expand multi-omics capabilities through innovations and strategic partnerships – including recent efforts to integrate broader proteomics technologies into the multiomics ecosystem.