You are here:
Development, validation and integration of in silico models to identify androgen active chemicals
Citation:
Manganelli, S., A. Roncaglioni, K. Mansouri, R. Judson, E. Benfenati, A. Manganaro, AND P. Ruiz. Development, validation and integration of in silico models to identify androgen active chemicals. CHEMOSPHERE. Elsevier Science Ltd, New York, NY, 220:204-215, (2019). https://doi.org/10.1016/j.chemosphere.2018.12.131
Impact/Purpose:
This paper describes integration of in silico approaches as a virtual screening tool for use in risk assessment of endocrine disrupting chemicals. Five classification models for predicting Androgen receptor-binding activity potential were developed and evaluated. ANNs, SVM and DT modeling approaches are presented and validated for transparency and reliability. These results enhance our understanding of the Androgen receptor-binding activity of environmental chemicals.
Description:
Humans are exposed to large numbers of environmental chemicals, some of which potentially interfere with the endocrine system. Nevertheless, most of these chemicals have not yet been tested for their endocrine disrupting potential, in particular, for their ability to interact with the androgen receptor. In this framework, the identification of potential endocrine disrupting chemicals (EDCs) has gained increasing priority in the assessment of environmental hazard. The U.S. Environmental Protection Agency (U.S. EPA) has developed the Endocrine Disruptor Screening Program (EDSP) which aims to prioritize and screen potential EDCs. The Toxicity Forecaster (ToxCast) program has generated data using in vitro high-throughput screening (HTS) assays measuring activity of chemicals at multiple points along the androgen receptor (AR) pathway. In the present study, five classification (quantitative) structure–activity relationship ((Q)SAR) models for predicting AR-activity potential were developed and validated using a large and diverse data set of 1,667 chemicals provided by the U.S. EPA from the combined ToxCast AR assays in the framework of the Collaborative Modeling Project for Androgen Receptor Activity (CoMPARA). Two models were built using ADMET predictorPredictor™, one is and are based on Artificial Neural Networks (ANNs) technology and the other use a and Support Vector Machine (SVM) algorithms; one model is a Decision Tree (DT) developed in R; and two models make use of differently combined sets of structural alerts (SAs) automatically extracted by SARpy. We used two strategies to integrate predictions from single models; one is based on a majority vote approach and the other on predictions convergence. These strategies led to enhanced statistical performance in most cases. Moreover, the majority vote approaches improved prediction coverage, when one or more single models are not able to provide any estimations. This study integrates multiple QSAR approaches as a virtual screening tool for use in risk assessment of endocrine disrupting chemicals.
URLs/Downloads:
DOI: Development, validation and integration of in silico models to identify androgen active chemicals