Protein-ligand Docking Application and Comparison using Discovery Studio and AutoDock

Abstract

Protein-ligand docking is a structure-based computational method, which is used to predict the small molecule binding modes and binding affinities with protein receptors. The goals of this study are to compare the docking performances of different software and apply the docking method to predict how protein fatty acid desaturase 1 (FADS1) interact with ligands. Two docking software, Discovery Studio and AutoDock, are used for docking comparison of 195 protein-ligand complexes from PDBind dataset. AutoDock performs a little bit better than Discovery Studio on the docking percentage, which is the percent of the docked complexes out of 195. On the other hand, Discovery Studio has a higher accuracy (successfully docked complexes, within 5 RMSD of the native complex structures) than AutoDock. The interaction between FADS1 and Sesamin shows a similar pattern comparing to the interaction between a homolog of FADS1 and a ligand shown in a PDB structure (PDB id 1EUE).