A Toxicity Prediction Tool for Potential Agonist/Antagonist Activities in Molecular Initiating Events Based on Chemical Structures

Abstract

Because the health effects of many compounds are unknown, regulatory toxicology must often rely on the development of quantitative structure–activity relationship (QSAR) models to efficiently discover molecular initiating events (MIEs) in the adverse-outcome pathway (AOP) framework. However, the QSAR models used in numerous toxicity prediction studies are publicly unavailable, and thus, they are challenging to use in practical applications. Approaches that simultaneously identify the various toxic responses induced by a compound are also scarce. The present study develops Toxicity Predictor, a web application tool that comprehensively identifies potential MIEs. Using various chemicals in the Toxicology in the 21st Century (Tox21) 10K library, we identified potential endocrine-disrupting chemicals (EDCs) using a machine-learning approach. Based on the optimized three-dimensional (3D) molecular structures and XGBoost algorithm, we established molecular descriptors for QSAR models. Their predictive performances and applicability domain were evaluated and applied to Toxicity Predictor. The prediction performance of the constructed models matched that of the top model in the Tox21 Data Challenge 2014. These advanced prediction results for MIEs are freely available on the Internet.

Predictive Capability of QSAR Models Based on the CompTox Zebrafish Embryo Assays: An Imbalanced Classification Problem

“…A total of four endpoints we selected for this part of modeling in order to verify the correctness of modelling methodology and the usefulness of sets of structural features (DS and FP) applied to CompTox data are from the Tox21 set of data. We compare our results with four participants [ 31 , 33 , 34 , 37 ] on that Tox21 challenge ( Table S3 ), and among them are the results of the winning solutions [ 37 ], as well as the second-ranked group [ 33 ]. We were able to reproduce two metrics calculated and used in the display of results on the CompTox set, i.e., MCC and BA calculated for the test set.…”

“…(1200–1500 compounds). Thus, our external set for validation of predictive abilities of models is twice as large as in the models with which they are compared [ 31 , 33 , 34 , 37 ]. This means that, with a similar value of individual evaluation metrics, the reliability of the parameter related to our method is higher than the method with which we compare here.…”

“…In the modeling, 10 data sets of molecular descriptors are calculated and used together with other modeling options used in optimization of associative neural networks (ASNN) which were used as the algorithm in ref. [ 31 ].…”

“…To achieve this, we have tested three different classifiers and chemical representations in a large experiment matrix of 209 model runs (11 models × 19 targets) to obtain results independent of machine learning, chemical representation (i.e., descriptors and fingerprints) and model hyperparameters. The prediction results are reported on train and test sets by means of multiple evaluation metrics such as the Matthew’s correlation coefficient (MCC) [ 28 ], Cohen’s Kappa [ 29 ] and Real-Accuracy (RA) (previously named [ 30 ]), which were selected due to their ability to capture performance in imbalanced datasets [ 31 , 32 , 33 ]. Additionally, the same methodology was applied on a 6–7 times larger set of toxicities from the Tox21 US EPA database.…”

Molecular Determinants of the Kinetic Binding Properties of Antihistamines at the Histamine H1 Receptors

“…The 3D-structure of each chemical structure was drawn using Marvin Sketch 18.10.0 (ChemAxon, Budapest, Hungary, (accessed on 1 February 2021)), and optimized using Toxicity Predictor [ 21 ]. Energy minimization calculations were performed using the Merck Molecular Force Field.…”