Molecular Docking Based On Particle Swarm Optimization

Molecular docking play an increasing important role in drug research and development. An efficient Molecular docking method is desirable in order to finish the screen process in proper time. In this paper, Particle Swarm Optimization was introduced to incorporate with Molecular docking, and comprehensive experiments were conducted on this platform. An efficient Molecular docking method named PetDock was implemented. PetDock was not only efficient in protein-ligand docking, but also efficient in protein-peptide docking.The aim of docking is to find the conformation with the lowest binding free energy. There are two important factors in the docking problem. One is an efficient search algorithm, which can explore the relevant conformational space, and the other is a good scoring function, which should reflect the affinity of the two proteins and guide the process of search.In this paper, the classical variants of PSO were introduced into the classical molecular docking platform (AutoDock), Comprehensive experiments were conducted on all of those optimization methods. The results demonstrated that PSO suppress LGA, especially FIPS. Using FIPS as the search method, a new molecular docking method named PetDock was implemented. This investigation also compared PetDock with the four state-of-the-art docking methods. PetDock suppress AutoDock in term of search ability and suppress all the four docking methods in term of accuracy. PetDock was also evaluated on Molecular-Peptide redocking experiments. The results demonstrate that PetDock is also efficient in complex Molecular-Peptide docking.