Computational analysis of the gut microbiota-mediated drug metabolism.

The gut microbiota, an extensive ecosystem harboring trillions of bacteria, plays a pivotal role in human health and disease, influencing diverse conditions from obesity to cancer. Among the microbiota's myriad functions, the capacity to metabolize drugs remains relatively unexplored despite its potential implications for drug efficacy and toxicity. Experimental methods are resource-intensive, prompting the need for innovative computational approaches. We present a computational analysis, termed MDM, aimed at predicting gut microbiota-mediated drug metabolism. This computational analysis incorporates data from diverse sources, e.g., UHGG, MagMD, MASI, KEGG, and RetroRules. An existing tool, PROXIMAL2, is used iteratively over all drug candidates from experimental databases queried against biotransformation rules from RetroRules to predict potential drug metabolites along with the enzyme commission number responsible for that biotransformation. These potential metabolites are then categorized into gut MDM metabolites by cross referencing UHGG. The analysis' efficacy is validated by its coverage on each of the experimental databases in the gut microbial context, being able to recall up to 74 % of experimental data and producing a list of potential metabolites, of which an average of about 65 % are relevant to the gut microbial context. Moreover, explorations into ranking metabolites, iterative applications to account for multi-step metabolic pathways, and potential applications in experimental studies showcase its versatility and potential impact beyond raw predictions. Overall, this study presents a promising computational framework for further research and applications gut MDM, drug development and human health.