Three-dimensional QSAR analyses of 1,3,4-trisubstituted pyrrolidine-based CCR5 receptor inhibitors

In this study, a series of 1,3,4-trisubstituted pyrrolidine-based CCR5 receptor inhibitors were taken as our target with the method of the three-dimensional quantitative structure-activity relationship (3D-QSAR) analyses in order to investigate the interactions between CCR5 receptor and their inhibitors. For a comparison, Comparative Molecular Field Analysis (CoMFA) and Comparative Molecular Similarity Indices Analysis (CoMSIA) were, respectively, used to build predictive models, which were generated from a training set consisting of 72 selected molecules, derived from literatures. Two alignment rules, including rigid body rms (root mean square) fit and field fit, were performed in the superimposition of inhibitors structures. As a result, a better CoMFA model based on common structure alignment obtains a conventional correlation coefficient r(2) of 0.952 and a leave-one-out cross-validated coefficient q(2) of 0.637, while the desirable CoMSIA model based on the same alignment rule acquires the r(2) of 0.958 and the q(2) of 0.677. To further validate the reliability of the models, we also investigated into the externally test set composed of 39 molecules under the criterions of squared correlation coefficient between experimental and predicted activities with intercept R(2) and without intercept R(0)(2), along with R(m)(2) as the modified R(2) with a penalty function due to difference between R(2) and R(0)(2). At last, the contour map also provides a visual representation of contributions of steric, electrostatic, hydrogen bond and hydrophobic fields, as well as the prospective binding modes. These results may provide meaningful guidance to the further work including the similar lead compounds' structure modification and activity prediction.