Fast and Accurate Blind Flexible Docking

2025-02-20

AI Summary

Existing molecular docking methods are constrained by the rigid-protein assumption and generative sampling, limiting their ability to simultaneously address binding-pocket identification, ligand conformation prediction, and protein flexibility modeling in blind docking. This work introduces the first end-to-end regression framework that unifies these three tasks—binding-pocket localization, ligand conformation generation, and protein pocket structural refinement—via a tri-module architecture with iterative coordinate updates. The method integrates geometry-aware graph representations, SE(3)-equivariant feature learning, and differentiable coordinate refinement to enable efficient blind flexible docking. On standard benchmarks, it achieves state-of-the-art accuracy while accelerating inference by 208× over the best generative approach, enabling millisecond-scale response times. This breakthrough significantly enhances computational feasibility for real-world drug discovery applications.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Multi-modal Vision

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Molecular docking that predicts the bound structures of small molecules (ligands) to their protein targets, plays a vital role in drug discovery. However, existing docking methods often face limitations: they either overlook crucial structural changes by assuming protein rigidity or suffer from low computational efficiency due to their reliance on generative models for structure sampling. To address these challenges, we propose FABFlex, a fast and accurate regression-based multi-task learning model designed for realistic blind flexible docking scenarios, where proteins exhibit flexibility and binding pocket sites are unknown (blind). Specifically, FABFlex's architecture comprises three specialized modules working in concert: (1) A pocket prediction module that identifies potential binding sites, addressing the challenges inherent in blind docking scenarios. (2) A ligand docking module that predicts the bound (holo) structures of ligands from their unbound (apo) states. (3) A pocket docking module that forecasts the holo structures of protein pockets from their apo conformations. Notably, FABFlex incorporates an iterative update mechanism that serves as a conduit between the ligand and pocket docking modules, enabling continuous structural refinements. This approach effectively integrates the three subtasks of blind flexible docking-pocket identification, ligand conformation prediction, and protein flexibility modeling-into a unified, coherent framework. Extensive experiments on public benchmark datasets demonstrate that FABFlex not only achieves superior effectiveness in predicting accurate binding modes but also exhibits a significant speed advantage (208 $ imes$) compared to existing state-of-the-art methods. Our code is released at https://github.com/tmlr-group/FABFlex.

Problem

• Improves molecular docking accuracy
• Addresses protein flexibility challenges
• Enhances computational efficiency in docking

Innovation

• Regression-based multi-task learning
• Iterative update mechanism
• Specialized module architecture