多模态跨尺度特征融合的药物靶标亲和力预测OA

To address the current challenges in relying solely on single-modal features of target proteins and neglec-ting network-scale features of biological networks,a drug-target affinity prediction(DTA)model based on multimo-dal cross-scale feature fusion was proposed.Target proteins both as sequences and graphs for feature extraction were presented,with semantic and topological features respectively,to enhance the target proteins' feature representa-tion.The strong affinity relationships between drugs and target proteins were analyzed to construct a heterogeneous graph network of drug-target interaction.A cross-scale feature fusion method was then used to effectively integrate the scale features of the heterogeneous graph network,then to enrich the feature representations of both target pro-teins and drug molecules.Experimental results on the DAVIS and KIBA datasets demonstrated that,compared with the more advanced model SISDTA,the proposed model achieved reductions in MSE by 0.015 and 0.003,respec-tively,and increases in CI by 0.005 and 0.004,respectively,improving the accuracy and stability of affinity pre-diction.It demonstrated the effectiveness of multimodal and cross-scale feature fusion in DTA prediction tasks.