一种基于多模态神经网络模型的DTA预测方法
By using a multimodal neural network model, combining GCN and CNN to extract drug and protein features, and fusing multi-scale features through an attention mechanism, the data dependence and accuracy problems in drug-target affinity prediction are solved, achieving more efficient prediction results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHENGZHOU UNIV
- Filing Date
- 2025-01-16
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for drug-target affinity prediction suffer from problems such as strong data dependence, limited prediction accuracy, and insufficient model generalization ability, and lack effective protein sequence map conversion methods.
A multimodal neural network model is adopted, including a drug feature extraction module, a protein overall feature extraction module, a protein feature fusion module, and an affinity prediction module. GCN and CNN are used to extract features of drugs and proteins respectively, and multi-scale features are fused through an attention mechanism to construct an MDNN-DTA prediction model.
It improves the accuracy and efficiency of drug-target affinity prediction, effectively uncovers potential interactions between drug and target data, and enhances the model's generalization ability and prediction accuracy.
Smart Images
Figure CN119993272B_ABST