A novel approach is introduced for the prediction of the psychotomimetic activity of the substituted phenethylamines as psychedelic drugs. In contrast to the available quantitative structure-activity relationship (QSAR) methods where they need complex computer codes, unusual descriptors, and expert users, it requires the molecular structure of the desired phenethylamine derivative. The largest reported values for 116 different substituted phenethylamines are used to derive and test the new correlation by 93 and 23 experimental data, respectively. The new correlation is based on the contributions of specific isomers under certain conditions and halogen atoms. Cross validations of the new correlation corresponding to the coefficients of determination for leave-one-out (Q 2 LOO) and the leave-many-out (Q 2 LMO) are the same and equal to 0.895, which is close to the coefficient of determination (R 2 = 0.900) of the training set. Further statistical parameters also confirm the goodness-of-fit, goodness-of-prediction, accuracy, and precision of the novel model. For eleven compounds where their outputs of external validation data set of psychotomimetic activity from multiple linear regression analysis (MLR) and non-linear multiple regression analysis (MNLR) QSAR models were available, the values of root mean squared error (RMSE) of the new model, MLR and MNLR are 0.172, 0.624 and 0.539, respectively.

No abstract