Using terahertz spectroscopy to identify transgenic cottonseed oil according to physicochemical quality parameters

This study proposed a non-destructive and rapid distinguish methodology for transgenic cottonseed oil according to physicochemical quality parameters applying terahertz spectroscopy and chemometric tools. 50 samples of transgenic cottonseed oils, belong different manufacturers, are used in this paper. The physicochemical quality parameters of cottonseed oils are estimated by GB/T 22223-2008. Terahertz spectra are employed in data preprocessing by of Savitzky-Golay derivative and offset baseline correction after principal component analysis (PCA). The linear discriminant analysis combines with successive projection algorithm (SPA-LDA) and discriminant analysis by partial least squares (PLS-DA) are applied to build identification models. in order to evaluate the performance of these identification models, 50 samples are divided into two groups, including training setsand test sets, by using different selection methods (Kennard-Stone, Duplex and Random). The obtained results show that the terahertz spectra can be apply as an alternative to identify different transgenic cottonseed oil from non-transgenic ones.