In silico molecular modeling and prediction of activity of substituted tetrahydropyrans as COX-2 inhibitor

Abstract

Series of mono-, di-, and triaryl substituted tetrahydropyrans were taken for quantitative structure–activity relationship (QSAR) study using density functional theory-based (DFT-based) quantum chemical and empirical descriptors. Several QSAR equations were formulated through regression analysis, which were able to explain 93–98 % of the variance in the data. The best equations were selected from the various statistically significant equations by following established statistical procedures. The model equations revealed that the presence of 4-cholorobenzene at R1 position is favorable for the activity and the presence of 4-bromo benzene at R1 position seems to be unfavorable for the activity, however, DFT-based descriptors demonstrate that the high values of HOMO and softness whereas low values of LUMO, hardness, and absence of –CH3 group at R1 position enhances the activity. The predictive ability of equations were cross validated by evaluating of the residual activity and R 2 values and also by leave one out (LOO) technique.