DCGCN: Dual-Channel Graph Convolutional Network-Based Drug-Target Interaction Prediction Method with 3D Molecular...

Area of Science:

• Computational chemistry
• Drug discovery
• Bioinformatics

Background:

• Drug-target interactions (DTIs) are vital for developing new therapeutics.
• Current DTI prediction methods often rely on simplified molecular representations like SMILES or 2D graphs, which overlook crucial 3D structural information.
• Limitations in existing methods hinder accurate prediction of drug efficacy and potential side effects.

Purpose of the Study:

• To develop an advanced computational method for predicting drug-target interactions (DTIs) by incorporating three-dimensional (3D) molecular structures.
• To address the limitations of existing DTI prediction techniques that primarily use 1D or 2D molecular data.
• To enhance the accuracy and reliability of DTI prediction for accelerating drug discovery.

Main Methods:

• Proposed DCGCN, a DTI prediction model leveraging 3D molecular structures.
• Decomposed 3D point cloud data into atomic sequence, connectivity, and distance map.
• Employed a dual-channel graph convolutional network (GCN) for atomic relationship analysis and 1D convolutional layers for sequence information extraction.

Main Results:

• DCGCN demonstrated superior performance compared to several state-of-the-art DTI prediction methods on two public datasets.
• The model effectively captured complex atomic relationships using 3D structural information.
• Results indicate a significant improvement in DTI identification accuracy by incorporating 3D molecular data.

Conclusions:

• Incorporating 3D molecular structures significantly enhances drug-target interaction prediction.
• DCGCN offers a promising approach for more accurate and efficient DTI identification in drug discovery.
• The study highlights the importance of leveraging comprehensive molecular structural data for computational drug development.