Introduction 

We are now a good few years into the Biobank Era, and the scale and richness of molecular data available from large biobanks and longitudinal studies like the UK, Estonian, and Finnish biobanks have grown massively. As the field continues to embrace multiomics, researchers increasingly have an opportunity to move beyond analyzing a single ‘omics data type and instead capture multiple layers of biology to gain deeper insights into human health and disease.

We believe that the integration of metabolomics and proteomics makes for particularly fruitful multiomics analyses, as we discussed in our recent webinars on the subject.

Expanding Biological Coverage 

Proteins regulate biological processes, while metabolites reflect the downstream consequences of those processes and provide a direct readout of physiological state. Analysing both together offers a more comprehensive understanding of disease mechanisms than either approach alone. Proteomics captures lower-level biology: what enzymes are cells making, what specific immune pathways are firing, what extracellular matrix components are cells using? Metabolomics captures the integrated effect of these enzymatic pathways – amino acid metabolism, lipid composition and, at the top, overall systemic metabolic health. No single ‘omics platform fully captures both facets.

As well as more fully describing the underlying biology, the complementary nature of metabolomics and proteomics can be important for capturing risk of common diseases. A recent cardiovascular disease prediction study by Luo et al. (2026) used UK Biobank data, to test metabolomics-based (MetScore) and proteomics-based (ProScore) molecular risk scores. Both scores independently predicted cardiovascular outcomes, and adding MetScore on top of ProScore significantly improved prediction of coronary artery disease and stroke. A similar study by Du et al. (2026) demonstrated this on a larger set of diseases, showing that integrating metabolomics and proteomics improved prediction across 17 incident diseases compared with clinical predictors alone. The combined two-omics model showed consistent improvement over single-omics models for renal disease, COPD, peripheral artery disease, and fractures.

By measuring both proteins and metabolites from the same sample, researchers can build a more complete picture of disease biology and identify molecular signals that may be missed when using a single platform.

Improving Biological Interpretation 

We have also found that multiomics can be key to improving biological interpretation.

Big proteomic studies often identify large numbers of disease-associated proteins, and sorting through them to determine which are causal drivers of pathology, which are downstream consequences of disease, and which are markers of broader confounding biological processes is extremely difficult. You can use metabolomics to provide the missing mechanistic context.

We are particularly fond of a paper from van der Graaf et al. (2025) that clearly demonstrated how ´integrating metabolomics, proteomics and genetics can reveal the biological mechanisms linking molecular traits to disease. By tracing the impact of genetics on both enzymes and metabolites, they discovered possible new metabolic pathways and mapped their effects on disease and drug targets. The scale of these sorts of genotype-metabolite-other omics analyses stepped up earlier this year with the publication of a huge UK/Estonian GWAS meta-analysis by Tambets et al. (2026), which utilized the very high power of metabolite + genetics analyses to understand the relationship between metabolomics, transcriptomics, proteomics and disease. This allowed the authors to triangulate disease risk and included an interesting story of how a cluster of proteins impacts red blood cell metabolism and thus modifies the risk of deep vein thrombosis.

These findings show how combining metabolomic, proteomic and genetic data can help researchers connect molecular biomarkers to the biological pathways that drive disease in a way that would be impossible to do using any single omics platform alone.

Refining Disease Stratification 

Predicting future disease is important, but understanding heterogeneity of people who have already developed chronic diseases is equally valuable. Patients sharing the same clinical diagnosis often have distinct underlying biology.

Combining metabolomic and proteomic data enables researchers to identify these biologically distinct subgroups. For instance, Cerdo et al (2026) used proteomics and metabolomics to split up lupus patients into two distinct categories, with different inflammatory and metabolic pathways, and differing profiles of disease complications. This mirrors similar work in tissue samples, for example Sharma et al (2024) demonstrated that a combination of proteomics and metabolomics on cancer biopsies can cluster breast cancer patients into groups with different disease prognoses.

These multiomic clustering approaches can reveal patient populations with different disease trajectories, treatment responses or underlying mechanisms, supporting more precise study design and, ultimately, more personalised approaches to prevention and treatment.

Using Metabolomics to Guide Proteomic Discovery 

Even before you start your project, you can use metabolomics to help design your multiomic study, by helping you pick which samples to profile. Nightingale NMR metabolomics is highly scalable and cost-effective, so investigators can profile entire cohorts and use this data to prioritise samples for deeper (and more expensive) proteomic characterization. This could include targeting individuals with unusual metabolic signatures, elevated cardiometabolic risk or specific exposure profiles for proteomic analysis.

The strategy was used effectively by Bizzarri et al (2025) to select 100 individuals with usually high or low metabolomic profiles of middle age frailty. They profiled these individuals using targeted and untargeted proteomics, uncovering a range of inflammatory and lipid-related proteins that predicted metabolic health.

The Future of Population-Scale Biology 

Evidence from large population resources including the UK Biobank, Estonian Biobank and other large-scale cohorts demonstrates the value of combining metabolomics, proteomics and genetics to improve disease prediction, reveal biological pathways and refine our understanding of disease mechanisms.

While evidence supporting integrated metabolomic and proteomic profiling continues to grow, relatively few studies have directly evaluated its benefits at scale. As increasingly rich biobank datasets become available, substantial opportunities remain for researchers to generate first-mover discoveries, uncover novel disease biology and improve our understanding of health, disease and prevention.

Interested in Integrating Multiomics into Your Study?

Nightingale Health offers integrated NMR metabolomics and proteomics through Olink and Alamar platforms, enabling multiomics data generation from a single blood sample through a streamlined workflow.

Researchers can benefit from support in study design, platform selection, data interpretation and replication planning, as well as opportunities for enhanced quality control through complementary metabolomic and proteomic measurements.

Contact the Nightingale Health research team at research@nightingalehealth.com to discuss your study.

 

References

Bizzarri, D., van den Akker, E. B., Reinders, M. J. T., Pool, R., Beekman, M., Lakenberg, N., Drouin, N., Stecker, K. E., Heck, A. J. R., Knol, E. F., Vergeer, J. M., Ikram, M. A., Ghanbari, M., van Gool, A. J., Deelen, J., BBMRI-NL, Boomsma, D. I., & Slagboom, P. E. (2025). Extreme MetaboHealth scores in three cohort studies associate with plasma protein markers for inflammation and cholesterol transport. Immunity & Ageing, 22, Article 34. https://doi.org/10.1186/s12979-025-00527-7 

Cerdó, T., Woodridge, L., Corrales, S., Muñoz-Castañeda, J. R., Torralbo, A. I., Rahman, A., Farinha, F., Ortega Castro, R., Segui, P., Sanchez-Pareja, I., Muñoz-Barrera, L., Merlo, C., Ruiz-Vilchez, D., Ábalos-Aguilera, M. C., Font, P., Barbarroja Puerto, N., PRECISESADS Clinical Consortium, Alarcón-Riquelme, M., Escudero-Contreras, A., … Lopez-Pedrera, C. (2026). Unveiling endotypes in systemic lupus erythematosus through multiomic analysis: Insights into cardiovascular and renal complications. Arthritis & Rheumatology. https://doi.org/10.1002/art.70127 

Du, J., Zhou, M., Wang, H., Wang, J., Raffield, L. M., Zhou, R., Li, Y., Chen, C., & Sun, Q. (2026). Multi-omics integration predicts the incidence of 17 diseases in the UK Biobank. Nature Communications, 17, Article 6271. https://doi.org/10.1038/s41467-026-73017-z 

Luo, Y., Zhang, N., Yang, J., Cui, M., Tsoi, K. K. F., Lip, G. Y. H., Liu, T., & Zhang, Q. (2026). AI-based multiomics profiling reveals complementary omics contributions to personalized prediction of cardiovascular disease. Nature Communications, 17, Article 2269. https://doi.org/10.1038/s41467-026-68956-6 

Sharma, A., Debik, J., Naume, B., Ohnstad, H. O., Oslo Breast Cancer Consortium (OSBREAC), Bathen, T. F., & Giskeødegård, G. F. (2024). Comprehensive multi-omics analysis of breast cancer reveals distinct long-term prognostic subtypes. Oncogenesis, 13, Article 22. https://doi.org/10.1038/s41389-024-00521-6 

Tambets, R., Jesse, M., Kronberg, J., van der Graaf, A., Abner, E., Võsa, U., Rahu, I., Taba, N., Kolde, A., Yarish, D., Abdullayeva, S., Alekseienko, A., Veidenberg, A., Estonian Biobank Research Team, Fischer, K., Kutalik, Z., Esko, T., Alasoo, K., & Palta, P. (2026). Genetic analysis of circulating metabolic traits in 619,372 individuals. Nature, 655, 971–978. https://doi.org/10.1038/s41586-026-10532-5 

van der Graaf, A., Rizi, S., Auwerx, C., & Kutalik, Z. (2026). Mendelian randomization linking metabolites with enzymes reveals pathway regulation and therapeutic avenues. American Journal of Human Genetics, 113(2), 309–323. https://doi.org/10.1016/j.ajhg.2025.12.013