Development, Validation, and Application of a Human Reproductive Toxicity Prediction Model Based on Adverse Outcome Pathway

• DOI: 10.1021/acs.est.2c02242
• Publication Date: 2022-08-12T13:10:28Z
• Authors: Haoyue Tan; Jinqiu Wu; Rong Zhang; Chi Zhang; Wei Li; Qinchang Chen; Xiaowei Zhang; Hongxia Yu; Wei Shi
• Keywords: Adverse Outcome Pathway; Reproductive toxicity; Applicability domain

A growing number of environmental contaminants have been proved to reproductive toxicity males and females. However, the unclear toxicological mechanism toxicants limits development virtual screening methods. By consolidating androgen (AR)-/estrogen receptors (ERs)-mediated adverse outcome pathways (AOPs) with more than 8000 chemical substances, we uncovered relationships between features, a series pathway-related effects, apical outcomes─changes in sex organ weights. An AOP-based computational model named RepTox was developed evaluated predict characterize chemicals' for Results showed that has three outstanding advantages. (I) Compared traditional models (37 81% accuracy, respectively), AOP significantly improved predictive robustness (96.3% accuracy). (II) application domain (AD) based on small vivo datasets, expanded ADs by 1.65-fold male 3.77-fold female, respectively. (III) implied hydrophobicity, cyclopentanol substructure, several topological indices (e.g., hydrogen-bond acceptors) were important, unbiased features associated toxicants. Finally, applied inventory existing substances China identified 2100 7281 potential female systems,