Development of an efficient molecular docking strategy for the prediction of interaction of COX-2 inhibitors with binding site of enzyme

Abstract

Introduction: Long-term use of Non-steroidal anti-inflammatory drugs (NSAIDs) for the treatment of chronic inflammatory disorders is severely restricted due to their dangerous gastrointestinal side effects. Therefore, the search for new, safe and effective anti-inflammatory drugs is always considered an essential need. The application of computer-aided drug design (CADD) is an essential approach for developing effective drugs. Protein-ligand docking is a typical method for structure-based drug discovery and design that aims to find the best conformation of a ligand against a protein receptor target with the lowest energy. Several docking programs are available nowadays, therefore it is of great importance to evaluate their performance. Purpose: Development of an efficient molecular docking strategy for the prediction of interaction of COX-2 inhibitors with binding site of enzyme.Materials and methods: In this study 51 protein-ligand complexes were used to evaluate the performance of 5 docking softwares including GOLD, Autodock, LeadIT, MVD and Glide. Protein and ligand preparation steps were done before docking. Finally, RMSD values between docked pose and crystallographic pose were calculated.Results: For 51 experimental structures of two enzymes COX-1 and COX-2 with inhibitory ligands, GOLD (ChemPLP), GOLD (GoldScore), GOLD (ChemScore), GOLD (ASP), Autodock with 10 runs, Autodock with 50 runs, LeadIT, MVD and Glide achieved success rate of 76.46, 59.01, 74.7, 76.27, 80.39, 82.35, 60.78, 70.58 and 100 percent respectively.Conclusion: Glide software performed better than the other docking softwares because it was able to dock 100 percent of the cases with RMSD less than 2 angstrom.