QSAR-based drug designing studies on HIV-1 integrase inhibitors

In this study, QSAR modeling was performed for predicting the IC 50 value for a set of HIV-1 integrase inhibitors using multiple regression and partial least square method obtaining an optimized model for each method. These models were used to predict a set of test compounds obtained by performing a chemical similarity search of the training set from the NCBI PubChem database subjected to the Lipinski rule of five filters. The predicted IC 50 value for the test set compounds was further analyzed for molecular docking simulation against HIV-1 integrase revealing that the test set compounds have a more binding affinity than the training set compounds and the market approved drug raltegravir. The stability of the docked compounds (protein–ligand complexes) was further validated by performing molecular dynamics simulations for 20 ns using Gromacs 5.0, and the RMSD backbone was analyzed. Last, the ADME–toxicity analysis was carried out for the top docking hit compounds and the market approved drug raltegravir revealing that the docked compounds have enhanced pharmacological parameters than raltegravir.