hERG-Att: Self-attention-based deep neural network for predicting hERG blockers

Hyunho Kim, Hojung Nam

Abstract: Highlights•The first self-attention-based, interpretable deep neural network model for predicting hERG blockers is proposed.•The various datasets ranging from public databases to publicly available private datasets are collected to train and test the model.•The proposed model not only shows better performance compared to other related studies but also provides potential hERG-related substructures.