You are here:
Comparing performance of in silico metabolism tools using data derived from literature and non-targeted analysis
Citation:
Boyce, M., B. Meyer, Chris Grulke, AND G. Patlewicz. Comparing performance of in silico metabolism tools using data derived from literature and non-targeted analysis. QSAR 2021 International Workshop on QSAR in Environmental and Health Sciences, Virtual, NC, June 07 - 10, 2021. https://doi.org/10.23645/epacomptox.15074556
Impact/Purpose:
Presentation to the QSAR 2021 International Workshop on QSAR in Environmental and Health Sciences June 2021. Exposure models used for risk-based chemical evaluation require measurement data to enable accurate and precise predictions. Specifically, measurement data are used as direct exposure model inputs (parameterization) and to evaluate and calibrate exposure model outputs. New approach methodologies (NAMs) for exposure science, including non-targeted analysis (NTA) methods, will further provide occurrence data and, in some instances, quantitative measurement data across thousands of poorly studied compounds. Emphasis will be given to characterizing metabolic and environmental transformation pathways for select compounds. Data from these efforts will prioritize metabolites and transformation products for toxicity testing and inform read-across methods that support provisional peer-reviewed toxicity values (PPRTV). Individual products/data will be used by other projects within CSS, and directly by program, regional and state partners.
Description:
Understanding the metabolic fate of a chemical substance is important for evaluating its toxicity. Changes in the regulatory landscape of chemical safety assessment provide opportunities to use in silico tools for metabolism prediction. In this study, a set of 37 structurally diverse chemicals were compiled from the EPA ExpoCast inventory to compare and contrast a selection of in silico tools, in terms of their coverage and performance. The tools were Systematic Generation of Metabolites (SyGMa), Meteor Nexus, BioTransformer, Tissue Metabolism Simulator (TIMES), OECD Toolbox, and Chemical Transformation Simulator (CTS). Performance as characterized by sensitivity and precision were determined by comparing predictions against metabolites reported in literature. Reported metabolites (438 in total) were extracted from 49 papers. Coverage was calculated to provide a relative comparison between tools. Meteor, TIMES, Toolbox, and CTS predictions were run in batches, using default settings. SyGMa and BioTransfomer were run with user-defined settings, (two passes of phase I and one pass of phase II). Hierarchical clustering revealed high similarity between TIMES and Toolbox. SyGMa had the highest coverage, matching an average of 41.2% of predictions generated by the other tools. SyGMa was also prone to significant overpredicting, generating a total of 5,125 predictions or 67% of total predictions. Precision and sensitivity values ranged from 4.7-23.7% and 15-27.5% respectively. TIMES had the highest performance overall. Current efforts are focused on evaluating the concordance of in vitro data, newly generated, relative to the literature data and in silico predictions. This abstract does not reflect EPA policy.
URLs/Downloads:
DOI: Comparing performance of in silico metabolism tools using data derived from literature and non-targeted analysis
BOYCE_QSAR2021_POSTER_FINAL.PDF (PDF, NA pp, 570.757 KB, about PDF)