Artificial neural network prediction of the psychometric activities of phenylalkylamines using DFT-calculated molecular descriptors

In the present work, a quantitative structure-activity relationship (QSAR) method was used to predict the psychometric activity values (as mes- caline unit, log MU) of 48 phenylalkylamine derivatives from their density functional theory (DFT) calculated molecular descriptors and an artificial neu- ral network (ANN). In the first step, the molecular descriptors were obtained by DFT calculation at the 6-311G  level of theory. Then the stepwise multiple linear regression method was employed to screen the descriptor spaces. In the next step, an artificial neural network and multiple linear regressions (MLR) models were developed to construct nonlinear and linear QSAR models, res- pectively. The standard errors in the prediction of log MU by the MLR model were 0.398, 0.443 and 0.427 for training, internal and external test sets, res- pectively, while these values for the ANN model were 0.132, 0.197 and 0.202, respectively. The obtained results show the applicability of QSAR approaches by using ANN techniques in prediction of log MU of phenylalkylamine deri- vatives from their DFT-calculated molecular descriptors.