Discriminating gamma-irradiated soybean seeds by 1H NMR-based metabonomics

Gamma irradiation has been paramount for enhancing the quality of grain production, mainly in terms of distribution and storage. Gamma irradiation classification models for soybeans, however, are still in development. In this paper, we present a metabonomic model able to distinguish between gamma-irradiated and non-irradiated soybeans. The metabonomic model works on 1H NMR spectra of chloroform extracts and makes use of PCA and PLS-DA formalisms to classify the samples and to investigate which spectral bins are discriminatory. The model has presented an accuracy of 100% in the face of a real dataset involving 49 samples from diverse cultivars. In turn, the most important chemical shifts (δ) for discriminating among the samples were δ 1.57 and 1.62 ppm, which are assigned to β-carboxyl methylene groups of aliphatic chains of fatty acids. Besides, the gamma-irradiated samples showed an increasing in the integration areas at δ 1.57 ppm (assigned to free fatty acid) whilst non-irradiated samples showed an increasing in the integration areas at δ 1.62 ppm (assigned to fatty acids linked to glycerol as esters).