Relational similarity-based graph contrastive learning for DTI prediction.

Area of Science:

• Computational Biology
• Bioinformatics
• Drug Discovery

Background:

• Accurate drug-target interaction (DTI) prediction is crucial for efficient drug repurposing.
• Existing deep learning methods for DTI prediction often focus solely on structural or relational features, limiting performance.
• Integrating diverse feature types can significantly enhance DTI prediction accuracy.

Purpose of the Study:

• To develop an advanced DTI prediction model that leverages both structural and relational features of drugs and proteins.
• To improve the accuracy and efficiency of drug repurposing through enhanced DTI prediction.
• To introduce a novel graph contrastive learning approach for feature extraction in DTI prediction.

Main Methods:

• Proposed Relational Similarity-based Graph Contrastive Learning for DTI prediction (RSGCL-DTI).
• Extracted inter-protein and inter-drug relational features using graph contrastive learning on a heterogeneous drug-protein interaction network.
• Combined extracted relational features with structural features from D-MPNN and CNN for comprehensive feature representation.

Main Results:

• The RSGCL-DTI model demonstrated superior performance compared to eight state-of-the-art baseline models across four benchmark datasets.
• The proposed method showed robust performance on imbalanced datasets, a common challenge in DTI prediction.
• RSGCL-DTI exhibited excellent generalization capabilities, effectively predicting interactions for unseen drug-protein pairs.

Conclusions:

• Combining graph contrastive learning-derived relational features with structural features significantly enhances DTI prediction.
• RSGCL-DTI offers a more accurate and reliable approach for drug repurposing.
• The model's strong performance and generalization ability highlight its potential for real-world drug discovery applications.