Abstract

Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis (MTB). In this study, various molecular modeling tools were used to search for potent TB therapeutics. Approximately 792 anti-TB inhibitors were collected from the literature, along with their experimentally determined biological activities. These compounds were then classified into three groups based on their biological activities. The geometries of all the collected inhibitors were optimized, followed by docking them into a few anti-TB receptors obtained from the Protein Data Bank. The docking results express the nature of interactions between the ligands and their respective receptors, which are prevalently noncovalent, dominated by hydrogen bonds and van der Waals’ interactions. Later, we generated structure-based 14 quantitative structure-activity relationship models using appropriate descriptors to predict the bioactivity values of the selected anti-TB inhibitors against a few cell lines. The best models were selected on the basis of statistical parameters and were validated by training and test set division. The predicted bioactivities of the selected inhibitors are comparable to the experimentally determined bioactivity values.

Key words: Drug designing, Density Functional Theory, analogue based modeling, Molecular Docking, 2D and 3D QSAR, Mycobacterium tuberculosis