基于多元统计分析的牛排掺假定量判别及差异分析

Quantitative Discrimination and Differential Analysis of Steak Adulteration Based on Multivariate Statistical Analysis

结合市场对食品中不同成分的定量分析的需求,模拟不同比例混合样品,利用高分辨质谱进行数据采集及分析,分别考察主成分分析、偏最小二乘判别分析和正交偏最小二乘判别分析(orthogonal partial least squares-discrimination analysis,OPLS-DA)统计方法对不同混合比例的样品数据进行分析。相比之下,OPLS-DA模型具有更好的判别效果。通过进行S-Plot分析,分别筛选出25条猪源多肽及牛源多肽,经线性拟合分析,其中牛源的10条多肽和猪源的6条多肽的线性相关系数R2大于0.99,并开展了真实性样品的验证实验,证明筛选的多肽可用于预计样品中的成分含量。这项研究建立了一种量化牛排的方法,并为筛选差异显著性的定量多肽提供新思路。

In this study,considering the current market demand for quantitative analysis of adulterated samples,beef samples adulterated with different proportions of pork were analyzed by high-resolution mass spectrometry(HRMS),and the acquired data were evaluated by principal component analysis(PCA),partial least squares-discrimination analysis(PLS-DA),and orthogonal partial least squares-discrimination analysis(OPLS-DA).The results showed that the OPLS-DA model exhibited a better discriminative ability compared with the other methods.By sorting the S-plot values,25 porcine-derived peptides and 25 bovine-derived peptides were selected.The linear correlation coefficients(R2)for 10 bovine-derived peptides and 6 porcine-derived peptides were above 0.99.Confirmatory experiments showed that the selected peptides could be used to predict the contents of adulterants in real samples.This study has established a method for quantifying the adulteration of beef steak and provided a new idea for screening of significantly differential peptides useful for the quantification of adulteration of meat products.