文摘
Peripherally acting cannabinoid 1 (CB1) receptor antagonists are considered as potential therapeutics for the treatment of obesity with desired efficacy and reduced central nervous system side effects. A dataset of 72 compounds containing the 1,5-diaryl pyrazole basic skeleton having peripheral CB1 receptor antagonistic activity was utilized for three-dimensional quantitative structure–activity relationship studies. Compounds of the series exhibited high variations in the biological activity and chemical structures. Different types of molecular alignments, such as atom-based, data-based, centroid-based and centroid/atom-based were utilized to develop the best CoMFA model. The best CoMFA model was obtained with a database alignment and the same alignment was further used for the development of a CoMSIA model. The best developed CoMFA model had \(r^{2}_\mathrm{cv} = 0.552\) with six components, \(r^{2}_\mathrm{ncv} = 0.973,\,\mathrm{standard\,error\,of\,estimate} = 0.162,\,F{\text {-value}} = 281.239,\) while the best developed CoMSIA model had \(r^{2}_\mathrm{cv} = 0.571\) with six components, \(r^{2}_\mathrm{ncv} = 0.960,\, \mathrm{standard} \mathrm{error\,of\,estimate} = 0.196\) and \(F{\text {-value}} = 188.701.\) The predictive \(r^{2}\) values of these two models showed test set predictions of 0.528 and 0.679 for the best CoMFA and CoMSIA models, respectively. Based on a higher \(r^{2}_\mathrm{pred}\) value, the CoMSIA model was found to be the best one. The prediction accuracy and reliability of the best developed CoMSIA model have been validated using well-established methods. Using the inputs from the best CoMSIA contour maps, several novel highly selective peripherally acting CB1 receptor antagonists have been designed and reported herein. Keywords 3D-QSAR Endocannabinoids CB1 receptor antagonists 1, 5-Diaryl pyrazole Rimonabant