Large-scale structure-activity relationship study of hepatitis C virus NS5B polymerase inhibition using SMILES-based descriptors

Abstract

Hepatitis C virus (HCV) is composed of structural and non-structural proteins involved in viral transcription and propagation. In particular, NS5B is an RNA-dependent RNA polymerase for viral transcription and genome replication and is a target for designing anti-viral agents. In this study, classification and quantitative structure-activity relationship (QSAR) models of HCV NS5B inhibitors were constructed using the Correlation and Logic software. Molecular descriptors for a set of 970 HCV NS5B inhibitors were encoded using the simplified molecular input line entry system notation, and predictive models were built via the Monte Carlo method. The QSAR models provided acceptable correlation coefficients of [Formula: see text] and [Formula: see text] in the ranges of 0.6038-0.7344 and 0.6171-0.7294, respectively, while the classification models displayed sensitivity, specificity, and accuracy in ranges of 88.24-98.84, 83.87-93.94, and 86.50-94.41 %, respectively. Furthermore, molecular fragments as substructures involved in increased and decreased inhibitory activities were explored. The results provide information on QSAR and classification models for high-throughput screening and mechanistic insights into the inhibitory activity of HCV NS5B polymerase.

Keywords: Data mining; HCV NS5B polymerase inhibitors; Hepatitis C virus; Monte Carlo method; Structure-activity relationship.