Research On Protein-Peptide Docking Algorithm Based On Swarm Intelligence

Protein-peptide interaction are prevailing in computational biology, the accurate binding site prediction of protein-peptide is of great importance for understanding the complicated signaling pathway and the function of protein, and rational drug design. Since computational protein-peptide docking methods can provide sufficient modeling, the performance of them is crucial for experimental success and docking accuracy. Thus, protein-peptide docking algorithm has been an important research topic.Protein-peptide docking can be interpreted as a parameter optimization problem, the goal of its optimization is to find the low-energy compound structure of receptor and ligand, and the docking performance mainly depends on rapidly and valid conformation searching algorithm and accurate energy evalution function. To further improve the success rate and accuracy of protein-peptide docking, this paper mainly starts with conformation searching algorithm, the representative of swarm intelligence-artificial bee colony algorithm is chosen and improved, and further introduced into protein-peptide docking.In order to deal with the basic ABC algorithm for its slow convergence, tending to get stagnation on local optima, this paper proposed an improved artificial bee colony algorithm called enhanced mutual learning ABC algorithm(EMLABC) basing on multi comparatively prior neighbors for mutual learning, further to improve its searching efficency in exploration and exploitation, and it demonstrates significant improvements as compared to ABC together with its variants and several well-received PSO variants in terms of convergence, accuracy and stability. EMLABC is further extended to protein-peptide docking as its confromation search method, meanwhile, utilizing the AutoDock4.2 score function to direct search, and proposed the EMLABCDock docking algorithm, it shows comparatively better performance as regard to docking success rate and accuracy when compared to other current prevailing protein-peptide docking algorithm, including HADDOCK, DynaDock, PepATTRACT, GalaxyPepDock, pepCrawler and Rosetta FlexPepDock.