Development Of HawkRank:a New Scoring Function For Protein-protein Docking Based On Weighted Energe Terms

Protein-protein interactions(PPIs)play important roles in a variety of biological processes.Deciphering the structural determinants of PPIs is essential to understand the function of proteins at the atomic level,reveal the nature of many problems in life processes,and even discover new drug targets.Extensive efforts have been made to predict and explore PPIs by using theoretical methods because it is quite difficult or even impossible to resolve the three-dimensional structures of all protein-protein complexes by using experimental techniques.Protein-protein docking is the most important theoretical approach to predict the near-native structures for protein-protein complexes by simulating the interactions between protein monomers.The reliability of a protein-protein docking method is dependent on the ability of the scoring function to distinguish the near-native binding structures from a huge number of decoys.In this study,we developed HawkRank,a novel scoring function designed for the sampling stage of protein-protein docking by summing the contributions from several energy terms,including van der Waals potentials,electrostatic potentials and desolvation potentials.First,based on the solvation free energies predicted by the GB(Generalized Born)model for?800 proteins,a SASA(solvent accessible surface area)-based solvation model was developed,which can give the aqueous solvation free energies for proteins by summing the contributions of 21 atom types.Then,the van der Waals potentials and electrostatic potentials based on the Amber ff14SB force field were computed.Finally,the HawkRank scoring function was derived by determining the most optimal weights for five energy terms based on the training set including 105 protein-protein complexes from Benchmark4.0.We compared HawkRank with other three popular protein-protein scoring functions,including ZRANK,FireDock and dDFIRE in terms of Success Rate(SR),Modified Success Rate(MSR)and the correlation between the docking scores and RMSDs.The performance of HawkRank is slightly worse than ZRANK but better than FireDock and dDFIRE according to SR.However,according to MSR,HawkRank performs better than the other three scoring functions,suggesting that HawkRank can find more top-ranked hits from the decoys than the other three scoring functions.Besides,the scores predicted by HawkRank has higher correlation with RMSDs.Therefore,HawkRank is very suitable for the sampling stage of protein-protein docking due to its high computational speed,simple usage and good performance.The software package for the HawkRank scoring function was developed in C++.Users can download the executable file or source code of HawkRank,and use HawkRank on the local machine.Moreover,users also can write a script to call HawkRank to score multiple structures in bulk.Besides,we developed an online computing platform for HawkRank.Users can submit computational tasks one by one or in bulk on the website.The website adopts the B/S model and MVC design model.On the front end,we use HTML + CSS + JavaScript;and on the backend,we use Python + MySQL.