采用主成分分析不同来源蜂蜜理化品质和抗氧化品质特性

Analysis on the Physicochemical and Antioxidant Properties of Honey from Different Sources by Principal Component Analysis

对天然蜂蜜和商品蜂蜜的理化品质与抗氧化性能进行分析,采用相关性分析和主成分分析对样品进行数据分析和分类。结果显示,不同蜂蜜样品中15个品质指标间具有显著差异及相关性,脯氨酸与淀粉酶值(r=0.834)及总黄酮含量(r=0.914)呈现显著正相关关系;脯氨酸与羟甲基糠醛(hydroxymethylfurfural,HMF)(r=0.924)及果糖含量(r=0.937)呈现显著负相关关系。主成分分析(principal component analysis,PCA)将蜂蜜品质特性的15个特征指标降维得到4个相对独立主成分(principal component,PC),这4个主成分累计贡献率达88.83%。主成分1(principal component1,PC1)结果显示天然蜂蜜样品中淀粉酶值、总黄酮含量、脯氨酸含量较高,商品蜂蜜样品中的酸度、果糖和HMF含量较高。主成分综合得分排在前两位的为天然荆条蜂蜜P3和P5,排在后三位的为商品蜂蜜的C1(荆条蜜)、C3(槐花蜜)和C2(槐花蜜)。通过PC1和PC2可以100%将天然蜂蜜和商品蜂蜜分为2类。

The physicochemical quality and antioxidant properties of natural honey and commercial honey were analyzed.Correlation analysis and principal component analysis were used to analyze and classify the samples.The results showed that there was a significant correlation between the 15 quality indexes in different honey samples,and there was a significant positive correlation between proline and amylase value(r=0.834)and total flavonoids content(r=0.914).There was a significant negative correlation between proline and hydroxymethylfurfural(HMF)(r=0.924)and fructose content(r=0.937).Principal component analysis(PCA)reduced the dimension of 15 characteristic indexes of honey quality characteristics to obtain four relatively independent principal components(PC),and the cumulative contribution rate of these four principal components reached 88.83%.The results of principal component 1(PC1)showed that the amylase value,total flavonoids content and proline content of natural honey samples were higher,and the acidity,fructose and HMF contents of commercial honey samples were higher.The comprehensive scores of principal components ranked first and second for natural Jinghua honey P3 and P5,and the last three samples for commercial honey C1(Jinghua honey),C3(pagoda flower honey)and C2(pagoda flower honey).Natural honey and commercial honey can be 100%classified into 2 categories by PC1 and PC2.