Abstract
Introduction of artificial neural network (ANN) into the field of GMO detection is the aim of this investigation. The usefulness of ANN to predict transgenic maize (Bt-176) based on chemical composition of the extracted crude oil was evaluated. The training set, comprised of a composition of major and minor lipid components as inputs and outputs. Crude oil extracted from the genetically modified maize (Bt-176) and non-transgenic maize was characterized in terms of its fatty acids, phytosterols and tocopherols distribution as well as of its lipid classes and unsaponifiables amounts. The results obtained from lipid distribution analysis showed that the grains of Bt-176 maize were comparable in their composition to that of the control maize. The analytical data have been elaborated by supervised pattern recognition technique ANN in order to classify genetically modified maize (Bt-176) and conventional maize as well as to authenticate the origin of the samples.