2D QSAR ANALYSIS ON SUBSTITUTED PYRROLIDINE DERIVATIVES AS DIPEPTIDYL PEPTIDASE –IV INHIBITORS
DOI:
https://doi.org/10.48047/Keywords:
Diabetes, Pyrrolidine, QSAR, Dipeptidyl peptidase IV, Partial least square regression.Abstract
Computational methodology, QSAR was applied in order to achieve best 2D QSAR models. The whole 2D
QSAR study was performed on a series of substituted pyrrolidine derivatives with a sum of total forty
synthesized molecules as dipeptidyl peptidase IV blocking agents. TheQsar studies were done by the aid of
software named as VLifeMDS. All the forty molecules were considered as data set which were further
divided into training set and test set by the help of random selection method associated with stepwise
forward backward method. Partial least square regression analysis was done and the best models were
identified in terms of r2
(squared correlation coefficient) and q2
(cross validatedcorrelation coefficient)
values. Three best models were identified with r2
values of 0.7681, 0.7352, 0.7264and q2
values with
0.5193, 0.6037, 0.5626 for model 1, 2 and 3 respectively. Different descriptors and their contribution in
building two dimensional QSAR models reveals that the generated models were good for predicting
dipeptidyl peptidase IV inhibitory activity.
