In the present work, 2D- and 3D-quantitative structure-activity relationship (QSAR) analysis has been employed for a diverse set of eighty-nine quinoxalinones to identify the pharmacophoric features with significant correlation with the aldose reductase inhibitory activity. Using genetic algorithm (GA) as a variable selection method, multivariate linear regression (MLR) models were derived using a pool of molecular descriptors. All the six-descriptor based GA-MLR QSAR models are statistically robust with coefficient of determination (R2 )>0.80 and cross-validated R2 >0.77. The derived GA-MLR models were thoroughly validated using internal and external and Y-scrambling techniques. The CoMFA like model, which is based on a combination of steric and electrostatic effects and graphically inferred using contour plots, is highly robust with R2 >0.93 and cross-validated R2 >0.73. The established QSAR and CoMFA like models are proficient in identify key pharmacophoric features that govern the aldose reductase inhibitory activity of quinoxalinones.
Keywords: Aldose Reductase Activity; CoMFA like model; QSAR; Quinoxalinones.
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.