Establishment of near infrared analytic model for major fatty acids content of Brassica napus L.

In order to improve the prediction performance of NIR model for high oleic acid rapeseed and provide a rapid selection method for high oleic acid breeding, we established NIR model of fatty acids content based on a group of high oleic acid rapeseed, which the erucic acid content was 0 approximately, while the oleic acid content ranged from 60% to 80%. The near infrared model was generated based on the near infrared spectrum data from a set of 169 rapeseed samples, the analysis of the fatty acid content determined by gas chromatography and the combination of calibration technology. The results showed that the oleic acid, linoleic acid and linolenic acid presented a better modeling effect, the RSQ (calibration correlation coefficient) were 0.957,0.922,0.891. The cross-validation and external validation exhibited higher correlation coefficient (1-VR: 0.8926, 0.8542, 0.8008; RSQ: 0.926, 0.867, 0.855) and lower standard deviation (SECV: 1.8638, 1.5288, 0.8672; SEP: 1.540, 1.506, 0.725) respectively. It concludes that the model can be used for rapidly screening early material of high oleic acid breeding, and can be applied in the practice of breeding.