Comparative analysis of blind docking reproducibility

Abstract

Virtual screening methods of computer aided drug discovery (CADD) has revolutionisedthe drug discovery process. It has not only paced the discovery of new drugs but also reduced the cost needed for it. Cost effectiveness and time saving attributes has made CADD a popular choice in lead identification. But these methods are also prone to certain errors. Blind docking (BD) is a popular choice for analysis binding of ligand with the protein. But can blind docking be relied upon blindly. The present work is aimed at giving insight about the effectiveness of BD approach and its reproducibility. AutodockVina was used for blind docking of three natural compounds viz. poncirin, quercitrin and squalene with factor Xa. Each compound was docked five times to assess the reproducibility of protocol in terms of interacting residues of proteins and binding affinity. BD being influenced by exhaustiveness of ligand was also analysed by using two values of exhaustiveness i.e. 8 and 32 in an otherwise same protocol. Overall, this study indicates that it is easy to generate in silico data but it is required tovalidate and integrate data in the manner that generate reproducible information.