Quinoa variety identification based on fatty acid composition and multivariate chemometrics approaches

Quinoa consumption has increased in worldwide importance due to an extraordinary nutritional value and public acceptance as alternative food. Fatty acid profiles of 10 quinoa varieties grown in the same geographical location were analyzed using different chemometric multivariate approaches [variable in importance partial least square discriminant analysis (VIP-PLS-DA), stepwise linear discriminant analysis (S-LDA), linear discriminant analysis (LDA), random forests (RF) and canonical analysis of principal components (CAP)]. The application of variable selection approaches such as S-LDA and LDA significantly increased the classification accuracy (78% and 74% respectively) of the samples according to their variety. The S-LDA approach allowed to reduce the number of selected fatty acids, representing those fatty acids with higher statistical significance when applying other random and non-random approaches. These fatty acid profiles also allowed the estimation of the nutritional lipid profiles of each variety for suitability in the human diet, providing insights into the various nutritional qualities of each quinoa variety. It is proposed that these results can be used to facilitate the selection of varieties with optimized economic value.