Guidelines for evaluating structure-based methods for protein–ligand binding prediction nature.com

Viet-Khoa Tran-Nguyen and Pedro J. Ballester propose guidelines for evaluating deep-learning scoring methods used to generate protein–ligand poses and predict binding affinity. They call for transparent reporting of data, code and procedures, safeguards against data leakage and sampling bias, and fair comparisons using shared data splits, repeated runs and baseline methods.