In silico metabolic profiling of an Indian cohort refutes the "One-Diet-Fits-All" paradigm in colorectal cancer

The metabolic impact of diet on colorectal cancer (CRC) is governed by complex, personalized microbial networks that elude static nutritional guidelines. Therefore, we integrated shotgun metagenomics with personalized genome-scale metabolic modeling (GSMM) to simulate metabolic fluxes in an Indian cohort of 30 CRC patients and 110 healthy controls. Under six simulated dietary conditions, the CRC microbiome- characterized by Fusobacterium nucleatum enrichment, exhibited highly variable metabolic responses. While high-fiber interventions restored protective butyrate in 61% of patients, they paradoxically elevated the oncometabolite succinate in 79% of models. Furthermore, high-fat diets resulted in net hydrogen sulfide consumption defects, potentially exacerbating local toxicity. Conversely, Mediterranean and Vegan diets successfully restored microbial diversity and suppressed pathogenic species. These findings indicate that metabolic outcomes are strictly conditional on baseline microbial composition, refuting "one-diet-fits-all" guidelines. Our study highlights the capacity of computational modeling to predict non-intuitive metabolic side effects, providing a framework for precision nutritional oncology.