In Silico Guidance for In Vitro Androgen and Glucocorticoid Receptor ToxCast Assays

• DOI: 10.1021/acs.est.0c01105
• Publication Date: 2020-05-20T17:13:21Z
• Authors: Timothy E. H. Allen; Mark D. Nelms; Stephen W. Edwards; Jonathan M. Goodman; Steve Gutsell; Paul J. Russell
• Keywords: Adverse Outcome Pathway; Computational model

Molecular initiating events (MIEs) are key in adverse outcome pathways that link molecular chemistry to target biology. As they based on chemistry, these interactions excellent targets for computational approaches silico modeling. In this work, we aim ligand chemical structures MIEs androgen receptor (AR) and glucocorticoid (GR) binding using ToxCast data. This has been done an automated algorithm perform maximal common substructure searches binders each from the dataset. The models developed show a high level of accuracy, correctly assigning 87.20% AR 96.81% GR 25% test set holdout cross-validation. 2D structural alerts can be used as predict guidance vitro assays confirm hits. These such experimental reducing number performed gain required toxicological insight. Development also allowed some identified predictors agonist or antagonist behavior at target. work represents first step methods guide approaches.