GDEGAN: Gaussian Dynamic Equivariant Graph Attention Network for Ligand Binding Site Prediction

2026-03-20

AI Summary

Accurately predicting protein–ligand binding sites remains a central challenge in structure-based drug design. This work proposes a novel equivariant graph neural network that introduces a Gaussian dynamic attention mechanism: replacing conventional dot-product attention with a Gaussian kernel, it adaptively modulates attention bandwidth using the mean and variance of local residue features and incorporates a learnable temperature parameter to enable context-aware importance modeling. Evaluated on the COACH420, HOLO4k, and PDBBind2020 benchmarks, the method substantially outperforms existing approaches, achieving a 37–66% improvement in distance-to-closest-contact (DCC) metrics and a 7–19% increase in distance-to-closest-atom (DCA) success rates.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Game Theory and Economic Paradigms: Mechanism Design

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Accurate prediction of binding sites of a given protein, to which ligands can bind, is a critical step in structure-based computational drug discovery. Recently, Equivariant Graph Neural Networks (GNNs) have emerged as a powerful paradigm for binding site identification methods due to the large-scale availability of 3D structures of proteins via protein databases and AlphaFold predictions. The state-of-the-art equivariant GNN methods implement dot product attention, disregarding the variation in the chemical and geometric properties of the neighboring residues. To capture this variation, we propose GDEGAN (Gaussian Dynamic Equivariant Graph Attention Network), which replaces dot-product attention with adaptive kernels that recognize binding sites. The proposed attention mechanism captures variation in neighboring residues using statistics of their characteristic local feature distributions. Our mechanism dynamically computes neighborhood statistics at each layer, using local variance as an adaptive bandwidth parameter with learnable per-head temperatures, enabling each protein region to determine its own context-specific importance. GDEGAN outperforms existing methods with relative improvements of 37-66% in DCC and 7-19% DCA success rates across COACH420, HOLO4k, and PDBBind2020 datasets. These advances have direct application in accelerating protein-ligand docking by identifying potential binding sites for therapeutic target identification.

Problem

• ligand binding site prediction
• protein-ligand interaction
• computational drug discovery
• binding site identification
• 3D protein structure

Innovation

• Equivariant Graph Neural Networks
• Gaussian Attention
• Dynamic Bandwidth
• Binding Site Prediction
• Local Variance