BCM-DTI: A fragment-oriented method for drug–target interaction prediction using deep learning

The identification of drug-target interaction (DTI) is significant in drug discovery and development, which is usually of high cost in time and money due to large amount of molecule and protein space. The application of deep learning in predicting DTI pairs can overcome these limitations through feature engineering. However, most works do the features extraction using the whole drug and target, which do not take the theoretical basis of pharmacological reaction that the interaction is closely related to some substructure of molecule and protein into consideration, thus poor in performance. On the other hand, some substructure-oriented studies only consider a single type of fragment, e.g., functional group. To address these issues, we propose an end-to-end predicting framework for drug-target interaction named BCM-DTI that takes diverse fragment types into account, including branch chain, common substructure and motif/fragments, and applies a feature learning module based on CNN to learn the synergistic effect between these fragments. We implement BCM-DTI on four public datasets, and the results show that BCM-DTI outperforms state-of-the-art approaches and requires lower training cost.