CLUB-MARTINI: Selecting Favourable Interactions amongst Available Candidates, a Coarse-Grained Simulation Approach to Scoring Docking Decoys

摘要: Large-scale identification of native binding orientations is crucial for understanding the role protein-protein interactions in their biological context. Measuring free energy method choice to estimate strength and reveal relevance particular conformations which proteins interact. In a recent study, we successfully applied coarse-grained molecular dynamics simulations measure two protein complexes with similar accuracy full-atomistic simulation, but 500-fold less time consuming. Here, investigate efficacy this approach as scoring identify stable from thousands docking decoys produced by programs. To test our method, first it calculate energies all CAPRI (Critical Assessment PRedicted Interactions) benchmark dataset, included over 19000 solutions 15 targets. Based on energies, ranked select near-native modes under assumption that native-solutions have lowest energies. top 100 structures, 'easy' targets many conformations, obtain strong enrichment acceptable or better quality structures; 'hard' without decoys, still able retain structures contacts. Moreover, 10 selections, CLUB-MARTINI shows comparable performance when compared other state-of-the-art functions. As proof concept, performs remarkably well pinpoint selections. best knowledge, interaction calculated MD been used rank at large scale.