基于主成分分析和判别分析法对不同品种及规格鹿茸的差异性研究

目的：通过主成分分析（PCA）和判别分析（DA），探讨马鹿茸及梅花鹿茸的相似性和差异性。方法：采用高效液相色谱法测定17个氨基酸、核苷组分、磷脂组分、胆固醇及多胺的含量；采用紫外分光光度法测定水溶性蛋白质及总磷脂含量，以测定的营养成分含量作为主成分分析和判别分析法的分析数据源进行统计分析，化学指标数据采用Excel 2007处理，数据含量差异较大，分析前先将原始数据进行标准化处理。结果：以氨基酸含量为变量，进行主成分分析和判别分析，第1~2主成分得分图显示马鹿茸和梅花鹿茸交叉分布，而判别分析区分马鹿茸和梅花鹿茸正确率为94.8%；以总氨基酸、必需氨基酸、水溶性蛋白等八大营养成分为变量，进行主成分分析和判别分析，第1~2主成分得分图显示梅花鹿茸和马鹿茸明显分为2组，但不同规格的鹿茸交叉分布，而判别分析对梅花鹿茸和马鹿茸正确率仅为81.6%。结论：根据变量指标的不同，应用主成分分析和判别分析统计分析法能够分析不同品种鹿茸的相似性和差异性，为鹿茸的开发和应用提供了重要的理论依据。

Objective: To employ principal component analysis(PCA) and discriminant analysis(DA) were employed to explore the similarities and differences between Cervus elaphus Linnaeus and Cervus nippon Temminck. Methods: The contents of 17 amino acids,nucleoside,phospholipids cholesterol and polyamine were determined by high performance liquid chromatography(HPLC).Ultraviolet spectrophotometry(UV) was employed for the determination of the contents of water soluble protein and total phospholipid.The principal component analysis was applied to study characteristic,of velvet antler based on the data of analysis of the content of nutrient contents.The chemical index data was processed by excel 2007.The raw data varied greatly and were standardized before analysis.Results: PCA of amino acid contents cannot distinguish Cervus elaphus Linnaeus from Cervus nippon Temminck.DA can distinguish them and the accuracy was 94.8%.Cervus elaphus Linnaeus and Cervus nippon Temminck were obviously divided into two groups by PCA of other nutrients,but different specifications of velvet antler were crossed.The accuracy of other nutrients DA was only 84.6% for Cervus elaphus Linnaeus and Cervus nippon Temminck.Conclusions: Using PCA and DA statistical analysis,the similarity and difference of different species of antler could be distinguished,which provide,the important theoretical basis for the development and application of velvet antler.