K-NN, quantum mechanical and field similarity based analysis of xanthone derivatives as α-glucosidase inhibitors

Abstract

Herein, we have built up structure-activity relationship (SAR), k-nearest neighbor, quantum mechani- cal, field similarity analysis as well as a sufficient number of quantitative SAR (QSAR) models to explore the pharma- cological activity of Xanthones as a-glucosidase inhibitors. To achieve good results, SAR and QSAR have been inves- tigated apropos of a new descriptor ''similarity score''. These analyses reveal that among the various parameters, the softness and the similarity score have high correlation with activity. In addition, structural features like hydrogen bonding ability and steric factors also play a vital role in deciding the a-glucosidase inhibitory activity. The SAR analysis is vindicated by quantum mechanical studies. The best QSAR model was found to have R 2 = 0.81, Radj = 0.79 and RCV = 0.76.