A novel application of Kier–Hall electrotopological descriptors in datamining new HIV-1 non-nucleoside reverse transcriptase inhibitors (NNRTIs) on the basis of chemical similarity

The present study deals with the application of Electrotopological state (E-State) indices in identifying structural attributes, responsible for exhibiting better antiviral activity for the NNRTI class of the HEPT derivatives. Special thrust is given on the potential of using these simple 2D QSAR models in prefiltering a large chemical database. The relevance of substituents at 1-N, 5-C and thio-aryl ring in enhancing the antiviral activity has been discussed in detail and compared with 3D structural studies. The effects of nearby atomic descriptors on 1-N and 3-N (crucial for the antiviral activity) position have been analyzed with a new insight. Application of similarity-based filtering for data mining of an external database using the best model derived was successfully applied to PubChem dataset, wherein the hits retrieved, not only showed better antiviral activity (predicted), but also complimented with the docking results.