Metabolism has long been recognized as a significant factor in the success of generally time- and cost-expensive drug development projects, as it impacts the pharmacology, bioavailability, and toxicology of drugs. In fact, the metabolism of a drug can produce metabolites that have significantly different pharmacological and/or toxicological profiles, compared to their parent. Understanding how chemicals are metabolized within humans is therefore crucial to filter out bad candidates in early drug development phases, and reduce attrition rates. Over the last two decades, in silico prediction has raised increasing interest in the drug industry, and thus, sparked the development of several software tools and web servers for rapid, accurate, and free or low-cost prediction of small molecule metabolism. Unfortunately, even with the growing abundance of in silico metabolism prediction tools, there continue to be a number of significant limitations, especially with regard to their performance, their utility and their accessibility. In particular: 1) very few tools predict more than the sites of metabolism; 2) only a small number of tools provide predicted structures, and those that do generally place restrictions on their distribution; 3) very few of the tools make their databases or training sets available; 4) none of the comprehensive prediction tools are freely available; 6) many tools seriously over-predict metabolites and have remarkably high false positive rates; and 7) none of the tools combine phase I, phase II, gut microbial metabolism, and promiscuous metabolism together. To address these issues, we have developed BioTransformer, a freely available software tool for the prediction of small molecule metabolism. BioTransformer combines a knowledge-based approach and machine learning to predict the metabolism of small endogenous and exogenous molecules in humans (phase I and phase II) as well as their gut microbiota. Moreover, it integrates a module for predicting the microbial degradation of small molecules in soil and aquatic environments. In this presentation, I will provide a detailed description of BioTransformer, and some of its most recent applications, including the expansion of the DrugBank and HMDB databases, as well as a collaborative study of polyphenol and terpenoid metabolism by humans and their gut microbiota. BioTransformer is freely available at https://bitbucket.org/djoumbou/.

BioTransformer: An accurate, freely available tool for predicting small molecule metabolism– applications in drug discovery

255th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.