Development and Estimation of an in Silico Model for Anti-HIV-1 Integrase Inhibitor Using Genetic Function Approximation

Aim: Integrase inhibitors are an essential enzyme required for replication of the acquired immunodeficiency syndrome virus. It is a potent target for anti- HIV therapy. A QSAR study is performed on the series indole -diketo, diketo acid and carboxamide derivatives in order to analyze the physicochemical requirements of integrase inhibitors and to provide structural insight into the binding mode of the molecules to the enzyme. This will help in the design of these molecules as integrase inhibitors and predicting the inhibitory activity of the newly designed analogues.

Materials and Methods: All the derivatives in the series were sketched using ChemDraw ultra v12.0.2 module of ChemOffice 2010 and the sketched structures were consequently used for the calculation of molecular descriptors available in QSAR software Spartan’14 and PaDEL-Descriptors software. Quantum, constitutional, topological and functional group descriptors for all molecules were calculated using Spartan’14 v1.1.2, 2013 and PaDEL-Descriptors software v2.18, 2011 and correlation between the biological activity and molecular descriptors was found through genetic function approximation adopted by statistical program material studio v7.0.

Results: The generated QSAR models revealed that SsF, minHBint3, minHdsCH, FPSA-1 and RHSA descriptors have good correlation to the integrase inhibitors activity.

Conclusion: The results obtained by regression analysis indicated that minHBint3, minHdsCH and RHSA is negatively contributing to inhibitory activity thus; enhancement of inhibitory activity can be achieved by decreasing the respective descriptors. Positive contribution of SsF and FPSA-1 specifies that increase of sum of atom-type E-state: -F and PPSA-1/total molecular surface area, will impart positive influence on activity.