QSAR modeling and pharmacoinformatics of SARS coronavirus 3C-like protease inhibitors

Abstract

The search for effective treatment against novel coronavirus (COVID-19) remains a global challenge due to controversies on available vaccines. In this study, data of SARS coronavirus 3C-like protease (3CLpro) inhibitors; a key drug target in the coronavirus genome was retrieved from CHEMBL database. Quantitative Structure-Activity Relationship (QSAR) studies, Molecular docking, Absorption-Distribution-Metabolism-Excretion-Toxicity (ADMET) and molecular dynamics simulation (MDS) were carried out using these 3CLpro inhibitors. QSAR model constructed using the data had correlation coefficient R2 value of 0.907; cross-validated correlation coefficient Q2 value of 0.866 and test set predicted correlation coefficient R2pred value of 0.517. Variance inflation factor (VIF) values for descriptors contained in the model ranged from 1.352 to 1.68, hence, these descriptors were orthogonal to one another. Therefore, the model was statistically significant and can be used to screen and design new molecules for their inhibitory activity against 3CLpro. Molecular docking showed that seven of the compounds (inhibitors) used in the study had a remarkable binding affinity (−9.2 to −10.3 kcal/mol) for 3CLpro. ADMET study revealed that five (CHEMBL Accession IDs 19438, 196635, 377150, 208763, and 210097) of the seven compounds with good binding ability obeyed Lipinski's rule of five. Hence, they were compounds with drug-like properties. MDS analysis revealed that 3CLpro-compound 21, 3CLpro-compound 22, 3CLpro-compound 40 complexes are very stable as compared to the reference 3CLpro-X77 complex. Therefore, this study identified three potent inhibitors of 3CLpro viz. CHEMBL194398, CHEMBL196635, and CHEMBL210097 that can be further explored for the treatment of COVID-19.