"Mapping the plasma metabolome to human health and disease in 274,241 adults"
"Metabolites, signifying a complex interplay between genotype, behaviour and environment1,2,3, provide a unique readout of human health and disease4,5. Compared to other blood-based biomarkers, metabolites are more closely tied to phenotypes due to their key roles in physiological function control6. Therefore, accurately evaluating metabolic disturbance offers a comprehensive, precise and dynamic view of the disease status6."
"This study introduced the largest human metabolome–phenome atlas, encompassing a comprehensive collection of health and disease phenotypes in 274,241 individuals, with a follow-up duration of 14.9 years. By evaluating over 1.4 million potential associations, the atlas uncovered 52,836 metabolite–disease associations and 73,639 metabolite–trait associations. The atlas mapped metabolite variations assessed at different times before diagnosis, revealing key time points of metabolic changes over a decade preceding onset. Machine-learning-based MetRS demonstrated favourable performance (AUC > 0.8) for discriminating 81 incident and 94 prevalent diseases. In search of promising therapeutic targets, we identified 454 potentially causal metabolite–disease associations, among which 402 share common genetic determinants. Taken together, the open-access atlas provides a comprehensive panorama of the metabolome–phenome, advancing the research community by improving the understanding of disease pathophysiology, and guiding the development of biomarkers for enhanced diagnostic, predictive and treatment strategies."
https://www.nature.com/articles/s42255-025-01371-1
Principal Scientist, CEO and Founder of Eminent Biosciences, Indore, India. Director and Co-Founder of LeGene Biosciences Pvt Ltd. Indore, India. Instagram: dr.anuraj_nayarisseri Facebook anuraj_embs
1dMapping the plasma metabolome in 274K individuals across 15 years has yielded an atlas that goes far beyond a catalog of 50K+ metabolite–disease links , it reveals temporal shifts emerging years before clinical diagnosis. By combining robust machine-learning models (AUC > 0.8) with hundreds of causal signals anchored in shared genetic determinants, this open-access resource advances the field from mere associations to actionable insights in early detection, risk stratification, and therapeutic discovery. Its accessibility empowers researchers worldwide to validate findings, uncover novel biomarkers, and accelerate the transition from reactive medicine to proactive health. Given that the atlas highlights metabolic disturbances up to a decade pre-diagnosis, how might these signals be best integrated with multi-omics layers (genomics, proteomics, epigenomics) and longitudinal lifestyle/exposure data to separate causation from correlation? More importantly, could such integrative models clarify whether early metabolite changes are true disease drivers, compensatory adaptations, or simply risk indicators, and in turn, shape the design of preventive interventions and precision therapeutics?
Science & Tech Communicator | AI & Digital | Life Sciences | Chemistry
1dI think there is clear momentum with this kind of ststems