A simple model for the assessment of the agonistic activity of dibenzazepine derivatives by molecular moieties

Abstract

A simple approach is introduced to assess the agonistic activity of dibenzazepine derivatives including 11H-dibenz[b,e]azepine and the dibenz[b,f][1,4]oxazepine derivatives. Dibenzazepine derivatives show extremely potent activators of the human transient receptor potential ankyrin 1 receptor, which has a key role as a biological sensor. This new approach is based on molecular moieties for desk calculation of these compounds with more reliance on their answers as one could expect from available quantitative structure–activity relationship models. The measured agonistic activities of 31 diverse compounds in terms of pEC50 = −log (EC50), where EC50 is half-maximal effective concentration, are used as training (26 compounds) and test (5 compounds) sets. For the internal cross-validation, the calculated values of leave-one-out (Q2LOO) and the leave-more-out (Q2LMO) cross-validations are the same (0.99550) and very close to the coefficient of determination (R2 = 0.9965), which confirms that the new model is robust, well-behaved, and not over-fitted. The values of mean absolute percent error (MAPE) for training and test sets are 0.5943 and 0.8846, respectively, which are much lower than the outputs of two available complex comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA), i.e., the values of MAPE for training sets of CoMFA and CoMSIA are 0.8799 and 1.0412, while MAPE values for test sets of CoMFA and CoMSIA are 2.1591 and 1.7711, respectively.