化学
色谱法
西洋参
人参
线性判别分析
气相色谱法
高效液相色谱法
质谱法
气相色谱-质谱法
数学
医学
统计
病理
替代医学
作者
Mengying Shuai,Yang Yang,Fuqing Bai,Lijun Cao,Ruyan Hou,Chuanyi Peng,Huimei Cai
标识
DOI:10.1016/j.chroma.2022.463284
摘要
American ginseng (Panax quinquefolius L.) is protected by Geographical Indications as harvested roots have different regional characteristics. The aim of this study was to distinguish 95 American ginseng samples from four origins (including America, Canada, Shandong province and the Northeast provinces in China), and to further discriminate samples harvested within the protected designation of origin (PDO) from those harvested in non-PDO regions. Two metabolomic methods were used to acquire qualitative data on the metabolites of American ginseng samples from different origins, namely high-performance liquid chromatography (HPLC) and headspace-gas chromatography-mass spectrometry (HS-GC-MS). There were significant differences in numerous metabolites, including volatile compounds and ginsenosides. American ginsengs from four different regions were discriminated based on 25 volatile compounds and 8 ginsenosides using linear discriminant analysis (LDA), which had a 96.8% accuracy and a 74.7% cross validation rate, and random forest (RF) modeling, which reached 100% accuracy using the training set and 92.9% accuracy using the testing set. The same 33 analytes combined with LDA and RF were compared for discrimination of PDO and non-PDO samples. The 100% accuracy was again obtained using the RF model, but only when using data from both HS-GC-MS and HPLC. The result showed that chemical composition combined with chemometric is effectively and accurately to study the origins of American ginseng.
科研通智能强力驱动
Strongly Powered by AbleSci AI