QSAR studies on aminothiazole derivatives as aurora a kinase inhibitors

Abstract

Quantitative structure-activity relationship studies on 54 aminothiazole derivatives as Aurora A kinase inhibitors were performed to explore the important factors affecting their biologic activity. For 2D-quantitative structure-activity relationship study, genetic algorithm combined with multiple linear regression was used to select significant molecular descriptors. The MLR model gave squared correlation coefficient of 0.828 and squared cross-validated correlation coefficient of 0.771 for the training set compounds. Comparative molecular field analysis and comparative molecular similarity indices analysis were used to develop 3D-quantitative structure-activity relationship models. The comparative molecular field analysis model gave cross-validated correlation coefficient q2 of 0.695 and non-cross-validated correlation coefficient r2 of 0.977. For comparative molecular similarity indices analysis model, the corresponding q2 and r2 were 0.698 and 0.960, respectively. The proposed 3D-quantitative structure-activity relationship models were validated by the test set compounds not used in the modeling process, with values of 0.788 for comparative molecular field analysis and 0.798 for comparative molecular similarity indices analysis. The 3D contour maps suggested that further modification of the aniline group of compound 22 considering electrostatic, hydrophobic and hydrogen bond properties would influence the inhibitory activity. The results from quantitative structure-activity relationship models would be very useful to understand the structure-activity relationship of these inhibitors and to guide the further structural modification of new potential inhibitors. © 2010 John Wiley & Sons A/S.