THREE-DIMENSIONAL QSAR STUDY OF 2,4 - DISUBSTITUTED-PHENOXY ACETIC ACID DERIVATIVES AS A CRTh2 RECEPTOR ANTAGONIST: USING THE k-NEAREST NEIGHBOR METHOD
Abhishek Kumar Jain, Nimita Manocha, Ravichandran Veerasamy, Vishnu Kant Mourya · 2008
In pursuit of better CRTh2 receptor antagonist agents, 3D- QSAR studies were performed on a series of 2,4-disubstituted-phenoxy acetic acid derivatives. In this paper we report a novel three-dimensional QSAR approach, kNN-MFA, developed based on principles of the k-nearest neighbor method combined with various variable selection procedures. The kNN-MFA approach was used to generate models by all three different methods and predict the activity of test molecules through each of these models. The q 2, pred_r 2, Vn and k value of kNN-MFA with SW, SA & GA were (0.8392, 0.7059, 2/2) (0.6725, 0.6716, 2/4) and (0.6832, 0.6716, 2/4) although there are no common descriptors among these three methods, SW kNN-MFA method have better q2 (0.8392) and pred_r2 (0.7059) than other two methods, model validation correctly predicts activity 83.9 % and 70.5 % for the training and test set respectively. It uses 2 steric descriptors with 2 k nearest neighbor to evaluate activity of new molecule, So model generated by SA kNN-MFA are best model.