You are here:
Evaluating Machine Learning Models for Describing Chemical Distribution within Animals
Citation:
Wambaugh, J. Evaluating Machine Learning Models for Describing Chemical Distribution within Animals. FDA, NA, DC, September 06 - 08, 2022. https://doi.org/10.23645/epacomptox.21320538
Impact/Purpose:
This is a presentation to the U.S. FDA 10th Annual Scientific Computing Days in a session titled "A Collaborative and Multi-Disciplinary Approach to Advance Computing Methods." The meeting runs from September 6-8, and this session is scheduled on September 8.
Description:
Toxicokinetics (TK) is key for interpreting high-throughput screening (HTS) data in a public health risk context (Coecke et al., 2012). High Throughput TK (HTTK) is the combination of chemical-independent (generic) PBTK models and in vitro measurements of key TK determinants (plasma binding, metabolism). HTTK is applicable to thousands of chemicals and based on open source, free, and evaluated software. New machine learning models are allowing predictions of TK for chemicals lacking in vitro measurements. Confidence is determined by comparing with more traditional data sources – requires careful understanding of variability in data.
URLs/Downloads:
DOI: Evaluating Machine Learning Models for Describing Chemical Distribution within Animals
PRESENTATION.PDF (PDF, NA pp, 2993.596 KB, about PDF)