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

Yufei Huang,Odin Zhang,Lirong Wu,Cheng Tan,Haitao Lin,Zhangyang Gao,Siyuan Li,Stan.Z. Li

2024-02-21

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.

Biomolecules,Artificial Intelligence,Machine Learning,Chemical Physics

What problem does this paper attempt to address?

This paper proposes a new method called Re-Dock, aiming to address the flexibility and realism issues in molecular docking. Molecular docking is a critical task in drug design to predict protein-ligand binding structures. However, current methods often rely on idealized protein structures (known ligand-bound state) or neglect pocket side-chain conformations, resulting in inaccurate and unrealistic predictions. Re-Dock extends to geometric manifolds by introducing a diffusion bridge generation model to simultaneously predict ligand and pocket side-chain conformations, simulating the induced fit mechanism to achieve more accurate and realistic binding conformation predictions.

Specifically, the paper introduces the concept of energy-to-geometry mapping, inspired by the Newton-Euler equation, to jointly model binding energy and conformation, reflecting the energy-constrained docking generation process. Experiments on benchmark datasets, including apo-dock and cross-dock, demonstrate the superior performance and efficiency of the Re-Dock model compared to current methods.

The main contributions of the paper include:
1. Proposing the task of flexible docking and designing a new benchmark dataset.
2. Introducing Re-Dock, a diffusion bridge-based generation framework that can handle flexibility of pocket side-chains and integrate interaction priors to guide the generation process.
3. Demonstrating the potential advantages in realistic world applications like cross-docking.

Re-Dock solves the problem of neglecting side-chain flexibility and physical realism in existing methods through a diffusion bridge model on non-Euclidean manifolds, thus improving the accuracy of binding conformation prediction.