Re-Dock: Towards Flexible and Realistic Molecular Docking with Diffusion Bridge

TL;DR

Re-Dock is a novel diffusion-based model that predicts protein-ligand binding poses and pocket sidechain conformations simultaneously, addressing limitations of existing methods and improving accuracy and practicality in molecular docking.

Contribution

The paper introduces Re-Dock, a diffusion bridge generative model for flexible molecular docking that models both binding energy and conformations on geometric manifolds.

Findings

Outperforms existing methods in accuracy and efficiency

Effectively models sidechain flexibility during docking

Demonstrates superior results on benchmark datasets

Abstract

Accurate prediction of protein-ligand binding structures, a task known as molecular docking is crucial for drug design but remains challenging. While deep learning has shown promise, existing methods often depend on holo-protein structures (docked, and not accessible in realistic tasks) or neglect pocket sidechain conformations, leading to limited practical utility and unrealistic conformation predictions. To fill these gaps, we introduce an under-explored task, named flexible docking to predict poses of ligand and pocket sidechains simultaneously and introduce Re-Dock, a novel diffusion bridge generative model extended to geometric manifolds. Specifically, we propose energy-to-geometry mapping inspired by the Newton-Euler equation to co-model the binding energy and conformations for reflecting the energy-constrained docking generative process. Comprehensive experiments on designed benchmark datasets including apo-dock and cross-dock demonstrate our model's superior effectiveness and efficiency over current methods.