Molecular Modeling Studies of Thiophenyl C-Aryl Glucoside SGLT2 Inhibitors as Potential Antidiabetic Agents

Abstract

A QSAR study on thiophenyl derivatives as SGLT2 inhibitors as potential antidiabetic agents was performed with thirty-three compounds. Comparison of the obtained results indicated the superiority of the genetic algorithm over the simulated annealing and stepwise forward-backward variable method for feature selection. The best 2D QSAR model showed satisfactory statistical parameters for the data set (r2=0.8499, q2=0.8267, and pred_r2=0.7729) with four descriptors describing the nature of substituent groups and the environment of the substitution site. Evaluation of the model implied that electron-rich substitution position improves the inhibitory activity. The good predictive 3D-QSAR models by k-nearest neighbor (kNN) method for molecular field analysis (MFA) have cross-validated coefficient q2 value of 0.7663 and predicted r2 value of 0.7386. The results have showed that thiophenyl groups are necessary for activity and halogen, bulky, and less bulky groups in thiophenyl nucleus enhanced the biological activity. These studies are promising for the development of novel SGLT2 inhibitor, which may have potent antidiabetic activity.