Abstract
A QSAR analysis of 25 isosteviol derivatives was carried out to interpret the relationship between structural properties and α-glucosidase inhibitory activity. The selected significant models have good predictive ability, which was validated by LOO cross validation techniques (Q2 > 0.6) and which gives significant Cooks distances ( < 1) and other statistical parameters. The selected significant models suggest that the nitrogen count amongst those that contribute most to α- glucosidase inhibitory activity is positively accounted for activity. Besides nitrogen count, the XlogP (by Kellog method) and the SlogP (by Audry method) mosthydrophobichydrophilic distance descriptors also contribute positively and the Quadrupole2, Avg+P and SAMH parameters are negatively contributed for the activity while a triple bonded atom connected with another triple bonded atom by four or seven bonds distance is important for the activity. In conclusion, the presence of nitrogen atom along with the hydrophobic-hydrophilic distance in the molecule is important for the α- glucosidase inhibitory activity.
Keywords: QSAR, glucosidase inhibitor, Isosteviol, Glycoprotein, Nitrogen count, Hydrophobicity, validation, nitrogen, XlogP, Kellog method, Avg+P, bonds, diabetes, HIV, CD4, antiHIV agents, chiral carbons, enzyme, Stevioside, steviol, Molecular Design Suite, UFF, descriptors, highest correlation coefficient, variable inflation factor, Durbin-Watson (DW) test, LOO, regression coefficients, Quadrupole2, Hydrophobicity XlogP k, Hydrophobicity SlogP A, Electrostatic, Alignment Independent, molecular liphophilicity potentials, hydroxyl group, F value
18