3DMolFormer: A Dual-channel Framework for Structure-based Drug Discovery

2025-02-07

AI Summary

Current structure-based drug discovery methods struggle to jointly optimize protein–ligand docking and pocket-aware 3D molecular generation, hindered by limited 3D geometric modeling capacity and scarcity of labeled training data. To address this, we propose the first unified dual-channel Transformer framework that explicitly captures the intrinsic duality between docking and generative tasks. Our approach introduces a novel token-numeric parallel serialization scheme for joint representation of proteins, binding pockets, and ligands, and employs large-scale mixed-data pretraining followed by task-adaptive fine-tuning—integrating supervised learning for docking accuracy and reinforcement learning for molecular validity and drug-likeness. On standard benchmarks including PDBbind, our method achieves state-of-the-art performance: docking precision (RMSD < 2 Å) surpasses prior work, and generated molecules exhibit exceptional quality (QED > 0.92, SA score > 0.92, validity > 92%). The implementation is publicly available.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Search and Optimization: Sampling/Simulation-based Search
• Natural Language Processing: Generation

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Structure-based drug discovery, encompassing the tasks of protein-ligand docking and pocket-aware 3D drug design, represents a core challenge in drug discovery. However, no existing work can deal with both tasks to effectively leverage the duality between them, and current methods for each task are hindered by challenges in modeling 3D information and the limitations of available data. To address these issues, we propose 3DMolFormer, a unified dual-channel transformer-based framework applicable to both docking and 3D drug design tasks, which exploits their duality by utilizing docking functionalities within the drug design process. Specifically, we represent 3D pocket-ligand complexes using parallel sequences of discrete tokens and continuous numbers, and we design a corresponding dual-channel transformer model to handle this format, thereby overcoming the challenges of 3D information modeling. Additionally, we alleviate data limitations through large-scale pre-training on a mixed dataset, followed by supervised and reinforcement learning fine-tuning techniques respectively tailored for the two tasks. Experimental results demonstrate that 3DMolFormer outperforms previous approaches in both protein-ligand docking and pocket-aware 3D drug design, highlighting its promising application in structure-based drug discovery. The code is available at: https://github.com/HXYfighter/3DMolFormer .

Problem

• Dual-channel transformer for drug discovery
• Overcomes 3D modeling challenges in drug design
• Addresses protein-ligand docking and 3D drug design

Innovation

• Dual-channel transformer model
• Large-scale pre-training strategy
• Unified framework for drug discovery