In silico metabolomics is an emerging branch of metabolite profiling that combines cheminformatics and machine learning to help generate novel structures and identify novel metabolites. In this presentation I will describe recent developments that our lab has undertaken to advance the field of in silico metabolomics. I will first describe two new software tools that we developed (CypReact and CypBoM) that use machine learning to identify cytochrome P450 substrates and reaction sites. I will then show how these tools have been integrated into the latest version of BioTransformer, an open access, web-based biotransformation predictor. I will also describe some of the latest developments with BioTransformer and will introduce BioTransformerDB, a database of more than 5 million predicted biotransformation products generated from known starting compounds (drugs, pesticides, food chemicals and human metabolites). Finally I will show how machine learning is being used by our group to develop software that can accurately predict the retention indices (RI), collisional cross section (CCS), NMR chemical shifts and MS/MS spectra of these predicted compounds. Together, these resources should make the identification of novel metabolites much easier and much faster in both agrochemical and drug discovery research.<br/>

In silico tools for the generation and identification of novel metabolites

ACS Fall 2020 Virtual Meeting

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