Abstract:Traditional toxicity testing reliant on animal models is costly and low throughput, posing a significant challenge with the increasing numbers of chemicals that humans are exposed to in the environment. The purpose of this investigation was to build optimal prediction models for various human in vivo/organ-level toxicity end points (extracted from ChemIDPlus) using chemical structure and Tox21 in vitro quantitative high-throughput screening (qHTS) bioactivity assay data. Several supervised machine learning algorithms were applied to model 14 human toxicity end points pertaining to vascular, kidney, ureter and bladder, and liver organ systems. Three metrics were used to evaluate model performance: area under the receiver operating characteristic curve (AUC-ROC), balanced accuracy (BA), and Matthews correlation coefficient (MCC). The top four models, with AUC-ROC values >0.8, were derived for endocrine (0.90 ± 0.00), musculoskeletal (0.88 ± 0.02), peripheral nerve and sensation (0.85 ± 0.01), and brain and coverings (0.83 ± 0.02) toxicities, whereas the best model AUC-ROC values were >0.7 for the remaining 10 toxicities. Model performance was found to be dependent on the specific data set, model type, and feature selection method used. In addition, chemical structure and assay data showed different levels of contribution to the prediction of different toxicity end points. Although in vitro assay data, when combined with chemical structure, slightly improved the predictive accuracy for most end points (11 out of 14), a noteworthy finding was the near equal success of the structure-only models, which do not require Tox21 qHTS screening data, and the relatively poor performance of assay-only models. Thus, the top-performing structure-only models from this study could be applied for hazard screening of large sets of chemicals for potential human toxicity, whereas the largest assay contributions to models (i.e., cellular targets) could be used, along with the top-contributing structural features, to provide insight into toxicity mechanisms.

Machine Learning-Based QSAR Models and Structural Alerts for Prediction of Mitochondrial Dysfunction

Development of Classification Models for Predicting Inhibition of Mitochondrial Fusion and Fission Using Machine Learning Methods

Machine learning models on chemical inhibitors of mitochondrial electron transport chain

In Silico Prediction of Mitochondrial Toxicity of Chemicals Using Machine Learning Methods

Discriminant Models on Mitochondrial Toxicity Improved by Consensus Modeling and Resolving Imbalance in Training.

QSAR Study on Toxicity of Chemical Components of Chinese Materia Medica and Acute Toxicity of Rats

Modeling and Insights into the Structural Characteristics of Chemical Mitochondrial Toxicity

In Silico Prediction Of Chemical Toxicity For Drug Design Using Machine Learning Methods And Structural Alerts

Machine learning–based QSAR for safety evaluation of environmental chemicals

Linking electron ionization mass spectra of organic chemicals to toxicity endpoints through machine learning and experimentation

Hybrid non-animal modeling: A mechanistic approach to predict chemical hepatotoxicity

Predicting the Mitochondrial Toxicity of Small Molecules: Insights from Mechanistic Assays and Cell Painting Data

Predictive Models for Human Organ Toxicity Based on in Vitro Bioactivity Data and Chemical Structure

Centralized data sources and QSAR methods for the prediction of idiosyncratic adverse drug reaction

XML-CIMT: Explainable Machine Learning (XML) Model for Predicting Chemical-Induced Mitochondrial Toxicity

Computational Approaches to Identify Structural Alerts and Their Applications in Environmental Toxicology and Drug Discovery.

In silico prediction of mitochondrial toxicity by using GA-CG-SVM approach.

In Silico Prediction of Chemical Genotoxicity Using Machine Learning Methods and Structural Alerts.

Prediction Models for Agonists and Antagonists of Molecular Initiation Events for Toxicity Pathways Using an Improved Deep-Learning-Based Quantitative Structure–Activity Relationship System

Predicting Chemical Immunotoxicity through Data-Driven QSAR Modeling of Aryl Hydrocarbon Receptor Agonism and Related Toxicity Mechanisms

Computational Modeling for the Prediction of Hepatotoxicity Caused by Drugs and Chemicals