Group Ligands Docking to Protein Pockets

2025-01-25

AI Summary

Existing molecular docking methods typically treat protein–ligand interactions as isolated pairs, neglecting structural commonalities and cooperative information arising when multiple ligands bind to the same target protein. To address this limitation, we propose GroupBind—the first multi-ligand cooperative docking framework designed for ligands targeting the same protein binding pocket. Our approach introduces three key innovations: (1) a ligand-group interaction graph neural network that explicitly models 3D geometric relationships among protein–ligand and inter-ligand pairs; (2) a geometry-aware triangular attention mechanism encoding local conformational constraints; and (3) an end-to-end pose co-optimization strategy integrated with diffusion-based generative modeling. Evaluated on the PDBBind blind test set, GroupBind achieves significant improvements over state-of-the-art methods, demonstrating breakthrough advances in both binding pose prediction accuracy and cross-ligand generalizability.

Technology Category

Natural Language Processing: Language Grounding & Multi-modal NLPMachine Learning: Multimodal LearningGame Theory and Economic Paradigms: Cooperative Game Theory

Application Category

Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphsSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsSearch and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Molecular docking is a key task in computational biology that has attracted increasing interest from the machine learning community. While existing methods have achieved success, they generally treat each protein-ligand pair in isolation. Inspired by the biochemical observation that ligands binding to the same target protein tend to adopt similar poses, we propose extsc{GroupBind}, a novel molecular docking framework that simultaneously considers multiple ligands docking to a protein. This is achieved by introducing an interaction layer for the group of ligands and a triangle attention module for embedding protein-ligand and group-ligand pairs. By integrating our approach with diffusion-based docking model, we set a new S performance on the PDBBind blind docking benchmark, demonstrating the effectiveness of our proposed molecular docking paradigm.

Problem

• Molecular Docking
• Protein-Ligand Interaction
• Similarity Utilization

Innovation

• GroupBind
• Triangular Attention Module
• Diffusion-based Docking Model