hERG-toxicity prediction using traditional machine learning and advanced deep learning techniques

Erik Ylipää; Swapnil Chavan; Maria Bånkestad; Johan Broberg; Björn Glinghammar; Ulf Norinder; Ian Cotgreave

doi:10.1016/j.crtox.2023.100121

hERG-toxicity prediction using traditional machine learning and advanced deep learning techniques

Curr Res Toxicol. 2023 Sep 1:5:100121. doi: 10.1016/j.crtox.2023.100121. eCollection 2023.

Authors

Erik Ylipää¹, Swapnil Chavan², Maria Bånkestad¹, Johan Broberg¹, Björn Glinghammar³, Ulf Norinder⁴, Ian Cotgreave²

Affiliations

¹ Computer Systems Unit, Research Institutes of Sweden RISE, Kista 164 40, Sweden.
² Unit of Chemical and Pharmaceutical Toxicology, Research Institutes of Sweden RISE, Södertalje 151 36, Sweden.
³ Preclinical Development & Translational Medicine, Swedish Orphan Biovitrum AB, Solna 171 65, Sweden.
⁴ Department of Computer and Systems Sciences, Stockholm University, Kista 164 07, Sweden.

Abstract

The rise of artificial intelligence (AI) based algorithms has gained a lot of interest in the pharmaceutical development field. Our study demonstrates utilization of traditional machine learning techniques such as random forest (RF), support-vector machine (SVM), extreme gradient boosting (XGBoost), deep neural network (DNN) as well as advanced deep learning techniques like gated recurrent unit-based DNN (GRU-DNN) and graph neural network (GNN), towards predicting human ether-á-go-go related gene (hERG) derived toxicity. Using the largest hERG dataset derived to date, we have utilized 203,853 and 87,366 compounds for training and testing the models, respectively. The results show that GNN, SVM, XGBoost, DNN, RF, and GRU-DNN all performed well, with validation set AUC ROC scores equals 0.96, 0.95, 0.95, 0.94, 0.94 and 0.94, respectively. The GNN was found to be the top performing model based on predictive power and generalizability. The GNN technique is free of any feature engineering steps while having a minimal human intervention. The GNN approach may serve as a basis for comprehensive automation in predictive toxicology. We believe that the models presented here may serve as a promising tool, both for academic institutes as well as pharmaceutical industries, in predicting hERG-liability in new molecular structures.

Keywords: Deep Learning; Graph-neural Network; Random Forest; Recurrent-neural Network; Support-vector Machines; hERG Channel.