3D QSAR analysis on substituted Cyanopyrrolidine derivatives as Dipeptidyl peptidase –IV inhibitors

Abstract

In the current research work twenty five molecules of substituted cyanopyrrolidine derivatives were subjected to
develop best QSAR model. The best model was achieved by means of k nearest neighbor molecular field analysis
associated by stepwise forward backward method using partial least square regression method. Random selection
method was used to differentiate total number of 25 molecules into 20 of training and 5 of test sets. Total ten trials
were run to identify best model in which trial 1 gave significant statistical equation. The best model consist of
predictive internal q2 value of 0.9451 & external predictivity (pred_r2 = 0.9402). The best model bearing three
steric descriptors at different points such as S_667, S_319, S_916 having values in the range of -3.2599 22.9668, -
0.0748665 18.0689, -6.33242 6.35972. The obtained steric descriptors play a key function in determining biological
activity. The grid matrix with different steric points helps to understand the relationship of structural function of
substituted cyanopyrrolidine derivatives and its biotic action.The obtained statistical data would be considered to
design new potent DPP IV inhibitors