In this study, Radix Paeoniae Alba with different appearance traits and from three geographical regions was analyzed using a combination of high-performance liquid chromatography fingerprints, chemometrics, and content determination. Hierarchically clustered heatmap was developed for sample visualization. Principal component analysis was used to classify 47 Radix Paeoniae Alba samples with different appearance traits and from three different geographical regions. Principal component analysis results revealed that the Radix Paeoniae Alba samples fell into three clusters, which were related to geographical location and appearance traits. In addition, orthogonal partial least squares discriminant analysis was established to predict samples classification and obtain the 14 most significant variables (chemical markers) by a variable importance in the project. The six chemical markers were gallic acid (1), oxypaeoniflorin (7), paeoniflorin (12), 1,2,3,4,6-O-pentagalloylglucose (19), benzoic acid (20), and benzoyl paeoniflorin (24). Finally, the content of gallic acid, oxypaeoniflorin, paeoniflorin, and benzoyl paeoniflorin was determined, and the total contents of the four components were class 2 > class 1 > class 3, which suggests that geographical origins and appearance traits can be used to determine the quality of Radix Paeoniae Alba.

中文翻译：

---

HPLC指纹图谱与化学计量学比较分析不同外观性状、不同产地的白芍

本研究采用高效液相色谱指纹图谱、化学计量学和含量测定相结合的方法，对来自三个地理区域的具有不同外观性状的白芍进行了分析。分层聚类热图是为样本可视化而开发的。采用主成分分析方法对来自三个不同地理区域的47份不同外观性状的白芍样品进行分类。主成分分析结果表明，白芍样品分为三个簇，与地理位置和外观性状有关。此外，建立正交偏最小二乘判别分析以预测样本分类，并根据项目中的变量重要性获得14个最显着的变量（化学标记）。六个化学标记物是没食子酸 (1)、羟基芍药苷 (7)、芍药苷 (12)、1,2,3,4,6-O-五没食子葡萄糖 (19)、苯甲酸 (20) 和苯甲酰芍药苷 (24) . 最后测定了没食子酸、羟基芍药苷、芍药苷和苯甲酰芍药苷的含量，4种成分的总含量为第2类>第1类>第3类，表明可通过地理来源和外观性状判断白芍的品质。和苯甲酰芍药苷 (24)。最后测定了没食子酸、氧化芍药苷、芍药苷和苯甲酰芍药苷的含量，4种成分的总含量为第2类>第1类>第3类，表明可通过地理来源和外观性状判断白芍的品质。和苯甲酰芍药苷 (24)。最后测定了没食子酸、氧化芍药苷、芍药苷和苯甲酰芍药苷的含量，4种成分的总含量为第2类>第1类>第3类，表明可通过地理来源和外观性状判断白芍的品质。