Apo2Mol: 3D Molecule Generation via Dynamic Pocket-Aware Diffusion Models

2025-11-18

AI Summary

Traditional structure-based drug design often treats protein binding pockets as rigid, neglecting their intrinsic flexibility and ligand-induced conformational rearrangements. To address this limitation, we propose the first diffusion-based generative framework that jointly models 3D ligand molecules and their dynamically adapting protein pocket conformations, explicitly capturing the apo-to-holo co-variation process. Trained on a high-quality dataset of 24,000 apo–holo structural pairs, our method integrates all-atom graph neural networks with a dual-path diffusion mechanism and introduces a dynamic pocket-aware module to enable synergistic ligand generation and protein flexibility optimization. Experiments demonstrate state-of-the-art performance in predicted ligand binding affinity, holo-conformation accuracy, and recapitulation of induced-fit effects—significantly enhancing the feasibility of de novo drug design under realistic, dynamic pocket conditions.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Search and Optimization: Sampling/Simulation-based Search
• Humans and AI: Game Design — Procedural Content Generation & Storytelling

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Deep generative models are rapidly advancing structure-based drug design, offering substantial promise for generating small molecule ligands that bind to specific protein targets. However, most current approaches assume a rigid protein binding pocket, neglecting the intrinsic flexibility of proteins and the conformational rearrangements induced by ligand binding, limiting their applicability in practical drug discovery. Here, we propose Apo2Mol, a diffusion-based generative framework for 3D molecule design that explicitly accounts for conformational flexibility in protein binding pockets. To support this, we curate a dataset of over 24,000 experimentally resolved apo-holo structure pairs from the Protein Data Bank, enabling the characterization of protein structure changes associated with ligand binding. Apo2Mol employs a full-atom hierarchical graph-based diffusion model that simultaneously generates 3D ligand molecules and their corresponding holo pocket conformations from input apo states. Empirical studies demonstrate that Apo2Mol can achieve state-of-the-art performance in generating high-affinity ligands and accurately capture realistic protein pocket conformational changes.

Problem

• Generating 3D molecules for flexible protein binding pockets
• Overcoming limitations of rigid pocket assumptions in drug design
• Simultaneously creating ligands and their induced pocket conformations

Innovation

• Dynamic pocket-aware diffusion model for 3D molecule generation
• Hierarchical graph-based diffusion generating ligands and holo pockets
• Utilizes apo-holo structure pairs to model protein flexibility