Quantitative Structure-Activity Relationship Studies on the Histamin H3 Receptor Inhibitors Using the Genetic Algorithm-Multiple Linear Regressions

A quantitative structure-activity relationship model has been created for forecasting the antagonist potency of benzyl tetrazole derivatives as human histamine receptors. Various kinds of molecular descriptors were used to represent different aspects of the molecular structures. In this method, the whole data set for the compounds were divided into the training and test sets. The model of relationships between molecular descriptors and biological activity of molecules were created by using stepwise multiple linear regressions and a genetic algorithm. Comparison of the results obtained indicated the superiority of the genetic algorithm based multiple linear regression over the stepwise based multiple linear regression. The ultimate quantitative structure-activity relationship model (N =64, R2=0.808, F= 30.806, Q2adj= 0.782, Q2LOO = 0.751, Q2LGO=0.669) was fully approved using the leave-one-out cross-validation method, Fischer statistics (F), external test set and the Y-randomization test. As a result, the produced quantitative structure-activity relationship model could be applied as a valorous instrumentation for sketching analogous groups of new antagonists of histamine receptors.

Keywords:

Quantitative Structure-Activity Relationship , Genetic Algorithms , Multiple Linear Regressions , Histamine Receptor