Abstract:In this paper we report a successful application of machine learning approaches to the prediction of chemical carcinogenicity. Two different approaches, namely a support vector machine (SVM) and artificial neural network (ANN), were evaluated for predicting chemical carcinogenicity from molecular structure descriptors. A diverse set of 844 compounds, including 600 carcinogenic (CG+) and 244 noncarcinogenic (CG−) molecules, was used to estimate the accuracies of these approaches. The database was divided into two sets: the model construction set and the independent test set. Relevant molecular descriptors were selected by a hybrid feature selection method combining Fischer's score and Monte Carlo simulated annealing from a wide set of molecular descriptors, including physiochemical properties, constitutional, topological, and geometrical descriptors. The first model validation method was based a five-fold cross-validation method, in which the model construction set is split into five subsets. The five-fold cross-validation was used to select descriptors and optimise the model parameters by maximising the averaged overall accuracy. The final SVM model gave an averaged prediction accuracy of 90.7% for CG+ compounds, 81.6% for CG− compounds and 88.1% for the overall accuracy, while the corresponding ANN model provided an averaged prediction accuracy of 86.1% for CG+ compounds, 79.3% for CG− compounds and 84.2% for the overall accuracy. These results indicate that the hybrid feature selection method is very efficient and the selected descriptors are truly relevant to the carcinogenicity of compounds. Another model validation method, i.e. a hold-out method, was used to build the classification model using the selected descriptors and the optimised model parameters, in which the whole model construction set was used to build the classification model and the independent test set was used to test the predictive ability of the model. The SVM model gave a prediction accuracy of 87.6% for CG+ compounds, 79.1% for CG− compounds and 85.0% for the overall accuracy. The ANN model gave a prediction accuracy of 85.6% for CG+ compounds, 79.1% for CG− compounds and 83.6% for the overall accuracy. The results indicate that the built models are potentially useful for facilitating the prediction of chemical carcinogenicity of untested compounds.

Prediction of skin sensitization potential using D-optimal design and GA-kNN classification methods

Prediction of the Skin Sensitising Potential and Potency of Compounds Via Mechanism-Based Binary and Ternary Classification Models.

Prediction of skin sensitization using machine learning

Hybrid optimal descriptors as a tool to predict skin sensitization in accordance to OECD principles

In Silico Prediction of Skin Sensitization for Compounds Via Flexible Evidence Combination Based on Machine Learning and Dempster-Shafer Theory

Quantitative structure activity relationship model for predicting the depletion percentage of skin allergic chemical substances of glutathione.

Classification of Skin Sensitizers on the Basis of Their Effective Concentration 3 Values by Using Adaptive Boosting Method.

Integration of the Natural Language Processing of Structural Information Simplified Molecular-Input Line-Entry System Can Improve the In Vitro Prediction of Human Skin Sensitizers

Machine-learning based prediction models for assessing skin irritation and corrosion potential of liquid chemicals using physicochemical properties by XGBoost

Prediction of Chemical Carcinogenicity by Machine Learning Approaches.

Increasing Accessibility of Bayesian Network-Based Defined Approaches for Skin Sensitisation Potency Assessment

Iskin: Integrated Application of Machine Learning and Mondrian Conformal Prediction to Detect Skin Sensitizers in Cosmetic Raw Materials

Classification study of skin sensitizers based on support vector machine and linear discriminant analysis.

Improved Defined Approaches for Predicting Skin Sensitization Hazard and Potency in Humans

A Novel Machine Learning Model and a Web Portal for Predicting the Human Skin Sensitization Effects of Chemical Agents

In Vitro Prediction of Skin-Sensitizing Potency Using the GARDskin Dose–Response Assay: A Simple Regression Approach

How to resolve inconclusive predictions from defined approaches for skin sensitisation in OECD Guideline No. 497

Deriving a Continuous Point of Departure for Skin Sensitization Risk Assessment Using a Bayesian Network Model

Updating Reaction Mechanistic Domains for Skin Sensitization: 1. Nucleophilic Skin Sensitizers

Estimation of the Skin Sensitization Potential of Chemicals of the Acyl Domain Using DFT-Based Calculations

Quantum-Mechanics Calculations Elucidate Skin-Sensitizing Pharmaceutical Compounds