PoseX: AI Defeats Physics Approaches on Protein-Ligand Cross Docking

2025-05-03

AI Summary

Existing protein–ligand docking benchmarks suffer from limited practicality (e.g., supporting only self-docking) or reliance on training data and complex pipelines. This work introduces PoseX—the first lightweight, open-source benchmark unifying evaluation for both self-docking and cross-docking—comprising 2,030 protein–ligand complexes and systematically benchmarking 22 docking methods. We propose a novel, training-free cross-docking evaluation framework and, for the first time, introduce energy-driven conformational relaxation as a universal post-processing step. Our analysis reveals that AI-based models (e.g., DiffDock, AlphaFold3) consistently outperform traditional force-field methods in RMSD (1.2 Å lower on average in cross-docking) and that relaxation significantly ameliorates their stereochemical inaccuracies. The AI-plus-relaxation pipeline achieves state-of-the-art performance. All datasets, code, and a live leaderboard are publicly released.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Humans and AI: Other Foundations of Human Computation & AI
• Constraint Satisfaction and Optimization: Solvers and Tools

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Recently, significant progress has been made in protein-ligand docking, especially in modern deep learning methods, and some benchmarks were proposed, e.g., PoseBench, Plinder. However, these benchmarks suffer from less practical evaluation setups (e.g., blind docking, self docking), or heavy framework that involves training, raising challenges to assess docking methods efficiently. To fill this gap, we proposed PoseX, an open-source benchmark focusing on self-docking and cross-docking, to evaluate the algorithmic advances practically and comprehensively. Specifically, first, we curate a new evaluation dataset with 718 entries for self docking and 1,312 for cross docking; second, we incorporate 22 docking methods across three methodological categories, including (1) traditional physics-based methods (e.g., Schr""odinger Glide), (2) AI docking methods (e.g., DiffDock), (3) AI co-folding methods (e.g., AlphaFold3); third, we design a relaxation method as post-processing to minimize conformation energy and refine binding pose; fourth, we released a leaderboard to rank submitted models in real time. We draw some key insights via extensive experiments: (1) AI-based approaches have already surpassed traditional physics-based approaches in overall docking accuracy (RMSD). The longstanding generalization issues that have plagued AI molecular docking have been significantly alleviated in the latest models. (2) The stereochemical deficiencies of AI-based approaches can be greatly alleviated with post-processing relaxation. Combining AI docking methods with the enhanced relaxation method achieves the best performance to date. (3) AI co-folding methods commonly face ligand chirality issues, which cannot be resolved by relaxation. The code, curated dataset and leaderboard are released at https://github.com/CataAI/PoseX.

Problem

• Existing benchmarks lack practical evaluation setups for protein-ligand docking.
• Current frameworks are inefficient for assessing docking methods comprehensively.
• AI docking methods face stereochemical and generalization challenges.

Innovation

• Open-source benchmark for self and cross-docking evaluation
• Incorporates 22 docking methods across three categories
• Post-processing relaxation method refines binding pose