Best Multi-linear Regression Research Articles

Molecular modeling studies of pyridopurinone derivatives—Potential phosphodiesterase 5 inhibitors

Two- and three-dimensional quantitative structure-activity relationship (QSAR) and docking studies were carried out on a series of pyridopurinones, to model their phosphodiesterase 5 (PDE5) inhibitory activities. 2D-QSAR was performed using the heuristic and best multi linear regression (BMLR) methods in CODESSA (comprehensive descriptors for structural and statistical analysis), which had given linear models between the inhibitory activity and five descriptors of PDE5 inhibitors, with r 2 = 0.987, 0.987, q 2 = 0.970, 0.970, F = 166.71, 166.71 and s 2 = 0.0004, 0.0176, respectively. Comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) have provided statistically significant models with q 2 values of 0.784, 0.742 and r 2 values of 0.975, 0.972 respectively. The predictive ability of the models was validated using a set of 6 compounds that were not included in the training set and the predictive r 2 obtained for the test set was 0.901 and 0.888 respectively. Docking studies were employed to determine the probable binding conformation of these analogues in the PDE5 active site using the programs GOLD and AutoDock whose results were found complementary with 3D-QSAR maps. Since the potency towards PDE5 and the selectivity over PDE6 is important for the successful development of new PDE5 inhibitors, a PDE6 homology model was built using Insight II and Modeller with Phi-Psi BLAST alignment. The molecules were docked in the active site of PDE6 and analyzed the probable reasons for selectivity of these molecules towards PDE5 over PDE6. Mapping the 3D-QSAR models to the active site of PDE5 provides a new insight into the protein-inhibitor interactions and helpful in designing potent and selective PDE5 inhibitors for the treatment of erectile dysfunction.