Identification of InhA inhibitors: A combination of virtual screening, molecular dynamics simulations and quantum chemical studies.

In the present work, multiple pharmacophore-based virtual screening of the SPECS natural product database was carried out to identify novel inhibitors of the validated biological target, InhA. The pharmacophore models were built from the five different groups of the co-crystallized ligands present within the active site. The generated models with the same features from each group were pooled and subjected to the test set validation, receiver-operator characteristic analysis and Güner-Henry studies. A set of five hypotheses with sensitivity > 0.5, specificity > 0.5, area under curve (AUC) > 0.7, and goodness of hit score > 0.7 were retrieved and exploited for the virtual screening. The common hits (87 molecules) obtained from these hypotheses were processed via drug-likeness filters. The filtered molecules (27 molecules) were compared for the binding modes and the top scored molecules (12 molecules) along with the reference (triclosan (TCL), docking score = -11.65 kcal/mol) were rescored and reprioritized via molecular mechanics-generalized Born surface area approach. Eventually, the stability of reprioritized (10 molecules) docked complexes was scrutinized via molecular dynamics simulations. Moreover, the quantum chemical studies of the dynamically stable compounds (9 molecules) were performed to understand structural features essential for the activity. Overall, the protocol resulted in the recognition of nine lead compounds that can be targeted against InhA.