Attending on Multilevel Structure of Proteins enables Accurate Prediction of Cold-Start Drug-Target Interactions
This work addresses cold-start drug-target interaction prediction for drug discovery, offering a novel approach but is incremental in improving accuracy through multi-level structure modeling.
The paper tackled the problem of predicting interactions between novel drugs and proteins (cold-start drug-target interactions) by attending to multi-level protein structures, resulting in ColdDTI consistently outperforming previous methods on benchmark datasets.
Cold-start drug-target interaction (DTI) prediction focuses on interaction between novel drugs and proteins. Previous methods typically learn transferable interaction patterns between structures of drug and proteins to tackle it. However, insight from proteomics suggest that protein have multi-level structures and they all influence the DTI. Existing works usually represent protein with only primary structures, limiting their ability to capture interactions involving higher-level structures. Inspired by this insight, we propose ColdDTI, a framework attending on protein multi-level structure for cold-start DTI prediction. We employ hierarchical attention mechanism to mine interaction between multi-level protein structures (from primary to quaternary) and drug structures at both local and global granularities. Then, we leverage mined interactions to fuse structure representations of different levels for final prediction. Our design captures biologically transferable priors, avoiding the risk of overfitting caused by excessive reliance on representation learning. Experiments on benchmark datasets demonstrate that ColdDTI consistently outperforms previous methods in cold-start settings.