目的 建立草红藤的UPLC指纹图谱，结合化学模式识别分析手段评价不同产地的草红藤药材质量差异。方法 采用UPLC法建立草红藤药材指纹图谱，进行相似度评价、系统聚类分析、主成分分析（PCA）和正交偏最小二乘法判别分析（OPLS-DA）。结果 20批草红藤药材UPLC指纹图谱共确定9个共有峰，指认出3个色谱峰，分别是二氢杨梅素、杨梅素和花旗松素，相似度均大于0.9。聚类分析和主成分分析将20批草红藤药材分为3类，经OPLS-DA筛选出3个差异成分。结论 建立的草红藤药材UPLC指纹图谱方法简便，重现性良好，同时结合化学模式识别分析，可为草红藤的质量控制和品质提升提供参考。

Objective To establish an UPLC fingerprint spectrum of Shuteria involucrata (Wall.) Wight et Arn var. villosa (Pampan.) Ohashi, and evaluate the quality differences of S. involucrata var. villosa from different origins using chemical pattern recognition analysis methods. Methods UPLC method was used to establish the fingerprint of S. involucrata var. villosa and the fingerprint data was analyzed by similarity evaluation, systematic clustering analysis, PCA, and OPLS-DA. Results A total of 9 common peaks were marked in the UPLC fingerprint spectra of 20 batches of S. involucrata var. villosa, and 3 chromatographic peaks were identified, namely dihydromyricetin, myricetin, and taxifolin. The similarity was greater than 0.9. According to cluster analysis and principal component analysis, the 20 batches of samples were divided into three groups. And three differentially expressed components were screened using OPLS-DA. Conclusion The established UPLC fingerprint method is simple and has good reproducibility, and combined with chemical pattern recognition analysis, which can provide reference for the quality control of S. involucrata var. villosa.

R286.02