Quantitative Structure Activity Relationships Study of Soluble Epoxide Hydrolase Inhibitors Using MLR, ANN, CoMFA and CoMSIA Methods

摘要

Soluble epoxide hydrolase (sEH), a member of the α/β hydro- lase fold family, catalyzes the hydrolysis of epoxy eicosatrienoic acids to vicinal diol which are involved in the regulation of blood pressure and inflammation. In this study, 2D and 3D- QSAR analysis of novel N,N'-disubstituted urea derivatives as sEH inhibitors were performed by stepwise multiple linear regressions (SW-MLR), stepwise artificial neural networks (SW- ANN), Comparative Molecular Field Analysis (CoMFA) and comparative molecular similarity index analysis (CoMSIA). It was found that SW-MLR model with r of 0.954 for the training set and r 2 of 0.876 for the test set were more favorable than model established by CoMFA method with r 2 =0.983, r 2 pred =0.602 and CoMSIA with r 2 =0.987, r 2 pred =0.751 in 3D-QSAR. In addition, obtaining models were validated by cross validation with cut off value of q2 > 0.5. These developed models could be a useful guideline to design and predict the activity of novel com- pounds with enhanced sEH inhibitory activities.