Huazhong Agricultural University discovered a new method for non-destructive and rapid identification of A1 milk and A2 milk

Recently, the research results of Professor Wang Qiaohua's team from the School of Engineering of Huazhong Agricultural University are called "Rapid identification of A1 and A2 milk ba".

Milk containing only A2 beta-casein (A2 milk) is gaining worldwide popularity due to its unique health benefits

CARS algorithm to filter characteristic variables

This study analyzes the difference in mid-infrared spectral absorbance between A1 and A2 milk, finds the combination of sensitive bands as the full spectrum, and uses standard normal variable transformation, multivariate scattering correction, normalization, first derivative, second derivative, first order The spectrum is preprocessed by seven methods including difference and second-order difference, and the feature variables that can represent the difference between A1 and A2 milk are screened out by using the uninformative variable elimination method and the competitive adaptive weighting algorithm, and then the partial least squares discriminant analysis is constructed.

UVE algorithm to filter feature variables

This study selects the PLS-DA model as the best model and uses a set of independent samples to externally validate the model

Xiao Shijie, a master student from the School of Engineering of Huazhong Agricultural University, is the first author of the paper, and Professor Wang Qiaohua from the School of Engineering and Professor Zhang Shujun from the School of Animal Science and Technology are the co-corresponding authors

【English summary】

The milk containing only A2 β-casein (called A2 milk) is globally popular because of its unique health benefits.

Paper link: https://doi.