ProtLID2.0: A Residue-Specific Pharmacophore Framework for Protein–Protein Docking

Abstract

Predicting biologically relevant protein–protein interactions remains a major challenge in structural biology. Here, we present ProtLID2.0, a residue-specific pharmacophore framework that generates coordinate-based interaction preferences from molecular dynamics simulations of single-residue probes at receptor interfaces and uses these preferences to score docked complexes and identify cognate ligands. Benchmarking across 30 receptor systems and multiple docking methods demonstrated that ProtLID2.0 performs comparably to or better than HADDOCK for identifying near-native poses and ranking cognate ligands from decoy sets, while combining both scoring functions consistently improved performance. These results establish ProtLID2.0 as a robust and complementary scoring framework for protein–protein docking and ligand identification.