MM-DRPNet: A multimodal dynamic radial partitioning network for enhanced protein-ligand binding affinity prediction.

Area of Science:

• Computational chemistry
• Drug discovery
• Bioinformatics

Background:

• Accurate drug-target binding affinity prediction is crucial for drug discovery.
• Current computational methods have limitations in accuracy and often neglect 3D structural information.
• This hinders their application in computer-aided drug design (CADD).

Purpose of the Study:

• To present MM-DRPNet, a novel multimodal deep learning framework.
• To enhance drug-target binding affinity prediction by integrating structural, interaction, and physicochemical data.
• To overcome limitations of existing methods by incorporating 3D structural information.

Main Methods:

• Developed MM-DRPNet, a multimodal deep learning framework.
• Introduced a dynamic radial partitioning (DRP) algorithm for adaptive 3D space segmentation.
• Integrated protein-ligand structural information, interaction features, and physicochemical properties.
• Incorporated molecular topological features to model structural and spatial relationships.

Main Results:

• MM-DRPNet significantly outperforms state-of-the-art methods on benchmark datasets.
• Ablation studies confirmed the significant contribution of each component of the MM-DRPNet architecture.
• The dynamic radial partitioning (DRP) algorithm demonstrated superior spatial interaction capture compared to fixed methods.

Conclusions:

• MM-DRPNet offers a significant advancement in drug-target binding affinity prediction.
• The framework's multimodal approach and novel DRP algorithm enhance accuracy and utility in CADD.
• MM-DRPNet provides a powerful tool for accelerating drug discovery research.