Comparative study of DFT, AM1, and PM3 optimization methods modeling new psychotropic amphetamines

Abstract

In order to improve the determination of physico-chemical descriptors of organic substances, it is necessary to optimize the computational representations of their molecular structures. The computerized estimation of these descriptors is especially useful in the case of drugs of abuse, as clinical studies are not recommended for new compounds belonging to classes known to have a high toxicity. The optimization of molecular structures allows the improvement of the accuracy in determining the physico-chemical descriptors and consequently of the associated quantitative structure - property or structure - activity relationships (QSAR). We are presenting a comparative study regarding the effect of the AM1, PM3, and DFT optimization methods, which were applied in the case of a series of new psychotropic amphetamines for their physico-chemical characterization based on molecular descriptors.