目的 利用化学模式识别技术对绿萼梅药材指纹图谱数据进行分析，筛选特征性指标成分从而建立绿萼梅Q-Marker的定量分析策略，为绿萼梅质量评价提供科学依据。方法 采用超高液相色谱法建立3个产地绿萼梅的指纹图谱，运用相似度分析、聚类分析、主成分分析和偏最小二乘判别分析等化学模式识别技术筛选出不同产地绿萼梅化学成分的特征成分作为绿萼梅Q-Marker，并定量分析。结果 30批绿萼梅UPLC指纹图谱标定8个共有峰，相似度在0.816～0.969；通过聚类分析、主成分分析和偏最小二乘判别分析较好的区分各产地绿萼梅，综合分析筛选绿原酸、芦丁、金丝桃苷、异槲皮苷作为绿萼梅Q-Marker，质量分数分别在3.08%～4.71%、0.41%～0.71%、0.13%～0.25%、0.25%～0.38%。结论 通过指纹图谱结合化学模式识别技术的分析策略可快速有效的筛选绿萼梅Q-Marker，为绿萼梅质量评价提供参考。

Objective To analyze the UPLC fingerprints data of Mume Flos using chemical pattern recognition technology, screen the characteristic components and perform quantitative analysis, so as to provide scientific basis for the quality evaluation of Mume Flos. Methods A total of 30 batches of Mume Flos samples from three production areas were collected, and the UPLC fingerprint of Mume Flos was established. The main characteristic peaks were identified by comparison with reference materials, and similarity analysis, cluster analysis, principal component analysis (PCA) and discriminant analysis by partial least squaremethod (PLS-DA) were used to identify and analyze the characteristic components in Mume Flos from three producing areas, and the quantitative analysis was carried out. Results Eight common peaks were calibrated from UPLC fingerprints of 30 batches of Mume Flos, and the similarity was between 0.816-0.969. Through clustering analysis, PCA and PLS-DA, each producing area of Mume Flos was better distinguished. Chlorogenic acid, rutin, hyperin and isoquercitrin were screened as Q-Markers by comprehensive analysis, and the mass fractions were 3.08%-4.71%, 0.41%-0.71%, 0.13%-0.25% and 0.25%-0.38%, respectively. Conclusion The method of fingerprint combined with chemical pattern recognition technology can effectively screen the quality markers of Mume Flos from different producing areas, and provide a reference for the quality evaluation of Mume Flos.

R286.2

国家重点研发计划（2018YFC1707004）；安徽省自然基金资助项目（2008085QH394）