基于聚类分析和主成分分析的红参高效液相色谱特征图谱研究

Study on the specific chromatogram of red ginseng by high performance liquid chromatography based on cluster analysis and principal component analysis

中文关键词:  红参  高效液相色谱法  特征图谱  聚类分析  主成分分析

英文关键词:red ginseng  high performance liquid chromatography  specific chromatogram  cluster analysis  principal component analysis

中文摘要:

      目的 建立高效液相色谱法(high performance liquid chromatography, HPLC)研究红参的特征图谱, 并对红参进行聚类分析和主成分分析, 为红参的鉴别和质量控制提供借鉴。方法 采用HPLC法, 选择Thermo BDS Hypersil C18色谱柱, 以乙腈?0.1%磷酸水溶液为流动相进行梯度洗脱, 流速1.0 mL/min, 柱温40 ℃, 进样10 μL, 在203 nm波长处检测。以人参皂苷Rb1峰为参照, 绘制26批红参样品的HPLC图谱, 采用《中药色谱指纹图谱相似度评价系统》(2004A版)评价其相似度, 采用SPSS 23.0软件进行聚类分析和主成分分析。 结果 26批红参样品的HPLC图谱有12个共有特征峰, 相似度均大于0.900。聚类分析结果表明, 当平方欧式距离为25时, 26批红参可聚为2类, S1、S2、S3、S5、S6、S25、S26聚为一类, 其余聚为一类; 当平方欧式距离为10时, 26批红参可聚为5类, S5为一类, S2、S3和S26聚为一类, S1、S6和S25聚为一类, S11、S13、S14和S20聚为一类, 其余可聚为一类。红参的主成分分析结果显示, 4个主成分因子的累计方差贡献率为90.365%, 其中, S8样品的主成分因子综合得分最高。结论 该研究方法及分析结果可为红参的鉴别和质量控制提供参考。

英文摘要:

      Objective To establish the specific chromatogram of red ginseng by high performance liquid chromatography (HPLC), and make cluster analysis and principal component analysis of red ginseng, so as to provide reference for the identification and quality control of red ginseng. Methods A Thermo BDS Hypersil C18 column was selected for HPLC, and gradient elution was performed using acetonitrile-0.1% phosphoric acid aqueous solution as the mobile phase at the flow rate of 1.0 mL/min, the column temperature of 40 ℃, and injection 10 μL. The detection wavelength was 203 nm. Taking ginsenoside Rb1 peak as the reference, the HPLC chromatograms of 26 batches of red ginseng samples were drawn, and the similarity was evaluated by the TCM chromatographic fingerprint similarity evaluation system (2004A edition), and clustering analysis and principal component analysis were performed using SPSS 23.0 software. Results There were 12 common characteristic peaks in the HPLC chromatograms of 26 batches of red ginseng samples, and the similarity was greater than 0.900. The results of clustering analysis showed that when the square euclidean distance was 25, the 26 batches of red ginseng could be clustered into 2 classes, S1, S2, S3, S5, S6, S25, S26 into one class, and the rest into one class. When the square euclidean distance was 10, the 26 batches of red ginseng could be clustered into 5 classes, with S5 as one class, S2, S3 and S26 as one class, S1, S6 and S25 as one class, S11, S13, S14 and S20 as one class, and the rest as one class. The results of principal component analysis of red ginseng showed that the cumulative contribution rate of variance of the 4 principal component factors was 90.365%, and the comprehensive score of principal component factor in S8 sample was the highest. Conclusion The research method and analysis results can provide reference for the identification and quality control of red ginseng.