Flexible Peptide-protein Docking Based On Ensemble Of Peptide Conformer

From genomics to proteomics,the understanding of life continues to deepen.Many life activities are done by protein-protein interactions.Peptide-mediated protein-protein interactions are an important part of them,which are ubiquitous in cells and affect physiological health.The structure of the complex determines its function.In order to study the mechanism of protein-peptide interaction,it is necessary to understand its structure.However,due to the properties of peptide-mediated interaction and experimental technical limitations,there are few protein-peptide experimental structures.Therefore,many computational methods have been developed in this field to predict protein-peptide complex structures,and molecular docking is one of the most efficient computational tools in structure-based methods.Molecular docking also play a key role in the design and development of peptide drugs.One of the challenges of the protein-peptide docking is the flexibility of the peptide,and the existing docking methods are time consuming to deal with this problem.In this paper,we consider the flexibility of peptides by ensemble docking,and developed a hierarchical flexible docking algorithm HPepDock,which rigidly docking the 1000 peptide structures generated by the rapid peptide conformer generation program MODPEP to the binding site.We modified the protein-ligand docking method MDock to search for the binding conformation of the peptide and evaluated the sampled peptide conformation using the ITScore-PP,which is an iterative knowledge-based scoring function for protein-protein recognition.The results of the 53 complexes on the LEADS-PEP dataset show that the HPepDock algorithm performs well on protein-peptide local docking and is superior to the five small molecule docking programs in docking accuracy(DOCK,AutoDock,AutoDock Vina,Surflex and GOLD).More importantly,the HPepDock algorithm is very fast,with an average docking time of around 15 minutes.Based on the HPepDock local docking,this paper has developed an algorithm that can perform global docking of protein-peptide and built the corresponding web server HPEPDOCK,which is available at http://huanglab.phys.hust.edu.cn/hpepdock/.We compared the HPEPDOCK and pepATTRACT web servers in 57 unbound structures.The results show that HPEPDOCK has a better success rate than pepATTRACT in predicting the peptide binding conformations.Our HPEPDOCK web server is also very efficient,and the average time to complete a global docking is 29.8 minutes.