CoMPARA: Collaborative Modeling Project for Androgen Receptor Activity

Abstract

Background: Endocrine disrupting chemicals (EDCs) are xenobiotics that mimic the interaction of natural hormones and alter synthesis, transport, or metabolic pathways. The prospect of EDCs causing adverse health effects in humans and wildlife has led to the development of scientific and regulatory approaches for evaluating bioactivity. This need is being addressed using high-throughput screening (HTS) in vitro approaches and computational modeling.

Objectives: In support of the Endocrine Disruptor Screening Program, the U.S. Environmental Protection Agency (EPA) led two worldwide consortiums to virtually screen chemicals for their potential estrogenic and androgenic activities. Here, we describe the Collaborative Modeling Project for Androgen Receptor Activity (CoMPARA) efforts, which follows the steps of the Collaborative Estrogen Receptor Activity Prediction Project (CERAPP).

Methods: The CoMPARA list of screened chemicals built on CERAPP's list of 32,464 chemicals to include additional chemicals of interest, as well as simulated ToxCast™ metabolites, totaling 55,450 chemical structures. Computational toxicology scientists from 25 international groups contributed 91 predictive models for binding, agonist, and antagonist activity predictions. Models were underpinned by a common training set of 1,746 chemicals compiled from a combined data set of 11 ToxCast™/Tox21 HTS in vitro assays.

Results: The resulting models were evaluated using curated literature data extracted from different sources. To overcome the limitations of single-model approaches, CoMPARA predictions were combined into consensus models that provided averaged predictive accuracy of approximately 80% for the evaluation set.

Discussion: The strengths and limitations of the consensus predictions were discussed with example chemicals; then, the models were implemented into the free and open-source OPERA application to enable screening of new chemicals with a defined applicability domain and accuracy assessment. This implementation was used to screen the entire EPA DSSTox database of $\sim \mathrm{875,000}$ chemicals, and their predicted AR activities have been made available on the EPA CompTox Chemicals dashboard and National Toxicology Program's Integrated Chemical Environment. https://doi.org/10.1289/EHP5580.

Figure 1.

Workflow of the project defining the major steps and the different data sets used for training, evaluation, and prediction.

Figure 2.

Scores of the categorical binding (black), agonist (white) and antagonist (gray) models based on the evaluation set and the scoring Equation 1.

Figure 3.

Scores of the continuous binding (black), agonist (white) and antagonist models based on the evaluation set and the scoring Equation 1 (See Supplemental Material 1 for groups’ abbreviations).

Figure 4.

Histogram showing the distribution of the number of binding (black), agonist (white) and antagonist (gray) models covering the prediction set (minimum of 11 models for agonist and antagonist and 20 for binding).

Figure 5.

Histogram showing the distribution of the concordance of the binding (black), agonist (white) and antagonist (gray) single models.

Figure 6.

Histogram showing the distribution of the concordance between the binding models over the active predictions.

Figure 7.

Histogram showing the coverage and S-score of the single binding models in comparison with the consensus binding predictions for the full CoMPARA set.

Figure 8.

Box plot showing the correlation between concordance and accuracy of prediction for the evaluation set chemicals. The box represents the interquartile range. The lower and upper box boundaries represent the 25th and 75th percentiles, respectively. The horizontal line splitting the box represents the median value. The upper and lower whiskers represent the minimum and maximum values, respectively. Outliers are represented by the $+$ symbol.

Figure 9.

Box plot showing the correlation between concordance and potency for the active binders of the evaluation set chemicals. The box represents the interquartile range. The lower and upper box boundaries represent the 25th and 75th percentiles, respectively. The horizontal line splitting the box represents the median value. The upper and lower whiskers represent the minimum and maximum values, respectively. Outliers are represented by the $+$ symbol.

Figure 10.

Selected descriptors for the binding (. symbol), agonist (* symbol), and antagonist (x symbol) models and corresponding balanced accuracy (BA) calculated in five-fold cross-validation in forward selection based on the genetic algorithm (GA) ranking. The ranked descriptors are not overlapping for the three modalities.