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

2025-08-06

AI Summary

Existing molecular docking methods exhibit significantly weaker performance in ligand conformation prediction compared to protein binding-pocket prediction, primarily due to disparities in structural complexity and dynamic adaptability between ligands and proteins. Method: We propose Docking Game, a game-theoretic framework that formulates flexible protein–ligand docking as a two-player zero-sum game between ligand and protein. We design Loop Self-Play—a nested self-play algorithm—enabling alternating optimization of dual modules via coordinated inner- and outer-loop training. Theoretical convergence is proven. Structural prediction feedback and multi-task self-supervised iterative refinement are integrated to enhance dynamic conformational adaptation. Results: On mainstream benchmarks, our method achieves ~10% improvement over SOTA in the percentage of predictions with RMSD < 2 Å, markedly enhancing docking consistency and accuracy. This establishes a more reliable paradigm for flexible binding prediction in drug discovery.

Technology Category

• Game Theory and Economic Paradigms: Adversarial Learning
• Search and Optimization: Learning to Search
• Machine Learning: Online Learning & Bandits

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
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

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.

Problem

• Improving protein-ligand binding prediction accuracy
• Addressing performance gap in ligand docking
• Developing game-theoretic framework for docking

Innovation

• Game-theoretic framework for protein-ligand docking
• Loop Self-Play algorithm for mutual adaptation
• Dynamic pose refinement in two-level loop