“Mapping the spectrum from well-being to disease using metabolomics”

Metabolomics has become a powerful approach for probing physiological state, enabled by advances in LC–MS and GC–MS that allow large-scale profiling of small molecules. We report two cross-sectional studies demonstrating metabolomics-based disease staging, risk prediction, and molecular assessment of the protective effects of marathon running. Untargeted metabolomics was applied to chronic metabolic disease in the Asian Indian population, with a focus on type 2 diabetes and its complications. Distinct circulating metabolite panels associated with kidney and cardiovascular complications were identified, enabling disease staging and risk stratification. In parallel, analysis of marathon runners revealed coordinated changes in energy metabolism and amino acid turnover, providing insights into metabolic flexibility and long-term adaptation. Together, these studies present a systems-level view of metabolic health across the continuum from well-being to disease, discussed in the context of a recent American Heart Association advisory emphasizing integrated metabolic, cardiovascular, and kidney health.

To address challenges in processing large, noisy LC–MS datasets, we developed MSOne, an AI/ML-based platform for automated and rigorous data processing, and MetaMine, a curated repository of re-analyzed public-domain metabolomics datasets. These tools enable robust analysis and facilitate translation of metabolomics findings into precision medicine.