MFFDTA: A Multimodal Feature Fusion Framework for Drug-Target Affinity Prediction

Accurate prediction of drug-target binding affinity (DTA) is a crucial issue in drug discovery and drug repurposing. In recent years, with the advancement of computer applications, many deep learning methods have been proposed for drug-target binding affinity prediction. However, existing deep learning prediction methods mostly rely solely on sequence information or structural information of drugs and targets, which leads to valuable shared information in sequences and structures being underexplored, this affects the accuracy of drug-target binding affinity prediction and the interpretability of the model. Therefore, we propose a novel method named MFFDTA, which utilizes both sequence and structural information to comprehensively mine important information in the drug and target. Furthermore, we introduce a feature cooperative fusion device to ensure that information between different pattern features can complementarily integrate. Finally, DTA prediction is employed an adaptive attention mechanism with the fusing drug and target features. Experimental results on two benchmark datasets demonstrate that MFFDTA outperforms state-of-the-art models, indicating the effectiveness and feasibility of this method.