Achieving reliability and high accuracy in automated protein docking: Cluspro, PIPER, SDU, and stability analysis in CAPRI rounds 13–19

• DOI: 10.1002/prot.22835
• Publication Date: 2010-07-23T15:09:33Z
• Authors: Dima Kozakov; David R. Hall; Dmitri Beglov; Ryan Brenke; Stephen R. Comeau; Yang Shen; Keyong Li; Jiefu Zheng; Pirooz Vakili; Ioannis Ch. Pasch...; Sandor Vajda
• Keywords: 0303 health sciences; Protein Conformation; Computational Biology; Proteins; Molecular Dynamics Simulation; 03 medical and health sciences; Models, Chemical; Cluster Analysis; Protein Multimerization; Monte Carlo Method; Algorithms; Software; Protein Binding

Abstract Our approach to protein—protein docking includes three main steps. First, we run PIPER, a rigid body program based on the Fast Fourier Transform (FFT) correlation approach, extended use pairwise interactions potentials. Second, 1000 best energy conformations are clustered, and 30 largest clusters retained for refinement. Third, stability of is analyzed by short Monte Carlo simulations, structures refined medium‐range optimization method SDU. The first two steps this implemented in ClusPro 2.0 protein–protein server. Despite being fully automated, last step computationally too expensive be included When comparing models obtained CAPRI rounds 13–19 ClusPro, refinement predictions all predictor groups, arrived at conclusions. time history, our automated server was able compete with human groups. selecting top ranked models, current protocol reliably generates high‐quality complexes from separately crystallized proteins, even absence biological information, provided that there limited backbone conformational change. despite occasional successes, homology modeling requires further improvement achieve reliable results. Proteins 2010. © 2010 Wiley‐Liss, Inc.