Simultaneous voltammetric determination of cysteine, tyrosine and tryptophan by using principal component-artificial neural networks

Abstract

Voltammetry was used for the simultaneous determination of cysteine, tyrosine and tryptophan. Each analyte has a distinctive response at the glassy carbon electrode in a nearly neutral solution. The main difficulty encountered in their simultaneous determination is the high degree of overlapping. Extraction of individual analyte concentration from voltammetric responses was achieved using artificial neural networks (ANNs), principal component artificial neural networks (PC-ANNs), principal component regression and partial least squares regression methods. The calibration set was orthogonally designed in order to obtain maximum information from the calibration procedure. The calibration set was used as trailing set in artificial neural networks analysis. The different models were used to predict the concentrations of test set. The root mean square error of calibration and root mean square error of prediction were calculated for all models. The results showed that better prediction was achieved with PC-ANN. The number of hidden neurons, learning rate, momentum and the epochs of training were investigated. The combined technique using cyclic voltammetry, PCR and ANN is helpful in the simultaneous detection of mixtures of oxidizable amino acids.