The high content of amino acids of Chinese rice wine (CRW), especially essential amino acids makes it a food increasingly demanded by consumers. Rapid detection technique of amino acid content, which is an important quality and function index of CRW, is highly desirable for consumers, producers as well as administrative authorities. In this study, the potential of Fourier transform infrared spectroscopy (FT-IR) as a novel and rapid analytical technique to determine 17 free amino acids in CRW were investigated. Genetic algorithms (GA) and synergy interval partial least squares (SiPLS) were used to select the most efficient spectral variables to improve the prediction precision of the classic partial least squares (PLS) model constructed on the full-spectrum. The results demonstrated that compared with the PLS model using all wavelengths of FT-IR spectra, the prediction precision of model based on the spectral variables selected by GA and SiPLS was significantly improved, especially for arginine and proline. After systemic comparison and discussion, it was found that GA-SiPLS model achieved the best performance, with the correlation coefficient in calibration (R(2) (cal)) higher than 0.80 and the residual predictive deviation higher than 2.00 for all of the free amino acids analyzed in this study. The overall results confirmed that FT-IR combined with efficient variable selection algorithms is a method that may be useful to replace the traditional methods for routine analysis of free amino acids in CRW.
Keywords: Chinese rice wine; Fourier transform infrared spectroscopy; amino acids; genetic algorithms; synergy interval partial least squares.
© 2015 Institute of Food Technologists®