Reduced density gradient as a novel approach for estimating QSAR descriptors, and its application to 1, 4-dihydropyridine derivatives with potential antihypertensive effects

The relationship between the chemical structure and biological activity (log IC50) of 40 derivatives of 1,4-dihydropyridines (DHPs) was studied using density functional theory (DFT) and multiple linear regression analysis methods. With the aim of improving the quantitative structure-activity relationship (QSAR) model, the reduced density gradient s( r) of the optimized equilibrium geometries was used as a descriptor to include weak non-covalent interactions. The QSAR model highlights the correlation between the log IC50 with highest molecular orbital energy (E HOMO), molecular volume (V), partition coefficient (log P), non-covalent interactions NCI(H4-G) and the dual descriptor [Δf(r)]. The model yielded values of R 2=79.57 and Q 2=69.67 that were validated with the next four internal analytical validations DK=0.076, DQ=-0.006, R P =0.056, and R N=0.000, and the external validation Q 2boot=64.26. The QSAR model found can be used to estimate biological activity with high reliability in new compounds based on a DHP series. Graphical abstract The good correlation between the log IC50 with the NCI (H4-G) estimated by the reduced density gradient approach of the DHP derivatives.