A simple, descriptive and interpretable model is developed using Multiple Linear Regression (MLR) approach for predicting the inhibition of CETP enzyme. Five descriptors were selected from a pool of variables using forward selection algorithm to build a pentavariable QSAR model for a series of 2-arylbenzoxazole derivatives as CETP inhibitors useful as antihyperlipidemic agents. The model is validated with N-fold method cross-validation Q2 and conventional r2 values (Q2 is 0.8474 and r2 is 0.9478). The binding conformations of arylbenzoxazole derivatives were determined and predicted by molecular docking. Best docking scores were shown by molecules 22B and 29B
Loading....