The Docking Game: Loop Self-Play for Fast, Dynamic, and Accurate Prediction of Flexible Protein-Ligand Binding

TL;DR

This paper introduces a game-theoretic framework and LoopSelf-Play algorithm for more accurate and dynamic prediction of protein-ligand binding modes, significantly improving docking performance.

Contribution

It proposes a novel game-based modeling approach and LoopSelf-Play algorithm that enables mutual adaptation and refinement in protein-ligand docking predictions.

Findings

Achieves approximately 10% improvement in binding mode prediction accuracy

Demonstrates stable convergence of the LoopSelf-Play algorithm

Outperforms previous state-of-the-art docking methods on benchmark datasets

Abstract

Molecular docking is a crucial aspect of drug discovery, as it predicts the binding interactions between small-molecule ligands and protein pockets. However, current multi-task learning models for docking often show inferior performance in ligand docking compared to protein pocket docking. This disparity arises largely due to the distinct structural complexities of ligands and proteins. To address this issue, we propose a novel game-theoretic framework that models the protein-ligand interaction as a two-player game called the Docking Game, with the ligand docking module acting as the ligand player and the protein pocket docking module as the protein player. To solve this game, we develop a novel Loop Self-Play (LoopPlay) algorithm, which alternately trains these players through a two-level loop. In the outer loop, the players exchange predicted poses, allowing each to incorporate the other's structural predictions, which fosters mutual adaptation over multiple iterations. In the inner loop, each player dynamically refines its predictions by incorporating its own predicted ligand or pocket poses back into its model. We theoretically show the convergence of LoopPlay, ensuring stable optimization. Extensive experiments conducted on public benchmark datasets demonstrate that LoopPlay achieves approximately a 10\% improvement in predicting accurate binding modes compared to previous state-of-the-art methods. This highlights its potential to enhance the accuracy of molecular docking in drug discovery.