GraMDTA: Multimodal Graph Neural Networks for Predicting Drug-Target Associations

Identifying novel drug-target associations is vital for drug discovery. However, the screening of millions of small molecules for a selected target protein is challenging. Several computational approaches have been developed in the past using machine learning methods for computational drug-target association (DTA) prediction, which predominantly uses the structural data of drugs and proteins. Some of these approaches use knowledge graph networks and link prediction. To the best of our knowledge, there have been no approaches that use both structural learning, which offers molecular-based representations, and knowledge graph-based learning, which offers interaction-based representations, for DTA discovery. Based on the premise that multimodal sources of information acting complimentarily could improve the robustness of DTA predictions, we developed GraMDTA, a multimodal graph neural network that learns both structural and knowledge graph representations by utilizing multihead attention to fuse multimodal representations. We compared GraMDTA with other computational approaches for DTA prediction to demonstrate the power of multimodal fusion in the discovery of DTA.