GAsDock: a new approach for rapid flexible docking based on an improved multi-population genetic algorithm

Abstract

Based on an improved multi-population genetic algorithm, a new fast flexible docking program, GAsDock, was developed. The docking accuracy, screening efficiency, and docking speed of GAsDock were evaluated by the docking results of thymidine kinase (TK) and HIV-1 reverse transcriptase (RT) enzyme with 10 available inhibitors of each protein and 990 randomly selected ligands. Nine of the ten known inhibitors of TK were accurately docked into the protein active site, the root-mean-square deviation (RMSD) values between the docking and X-ray crystal structures are less than 1.7 Å; binding poses (conformation and orientation) of 9 of the 10 known inhibitors of RT were reproduced by GAsDock with RMSD values less than 2.0 Å. The docking time is approximately in proportion to the number of rotatable bonds of ligands; GAsDock can finish a docking simulation within 60 s for a ligand with no more than 20 rotatable bonds. Results indicate that GAsDock is an accurate and remarkably faster docking program in comparison with other docking programs, which is applausive in the application of virtual screening.