利用高效液相色谱全轮廓指纹图谱结合化学计量学方法对不同栽培地区的紫苏叶样品(共84个)进行区分.全轮廓色谱数据经自适应迭代加权最小二乘法(airPLS)和相关优化翘曲法(COW)校正后,线和保留时间漂移现象均得到明显改善.经预处理后的色谱数据采用主成分分析(PCA)进行解析,结果表明不同来源的样品能按其特性各自聚为一类;而分段间隔压缩变量后的色谱数据经主成分分析处理可得到与全轮廓色谱数据为输入变量时相一致的结果.此外,偏最小二乘判别分析(PLS-DA)对于紫苏叶样品分类的识别能力和预报能力分别为92.8
A total of 84 Perilla frutescens (L.) Britt. samples from three main geographical origins inChina were collected and investigated by using high performance liquid chromatography (HPLC) in-corporated with chemometrics method. Prior to investigation, the measured liquid chromatographicdata were subjected to pretreatments, including baseline correction and retention time alignment.Principal component analysis (PCA) was then applied in the aligned and compressed data sets, re-spectively, to evaluate the data quality, and three sample groups were observed in the PCA scoreplots. Moreover, partial least squares - discriminant analysis ( PLS - DA) was applied to classifythese herbal samples, providing better abilities of recognition(92. 896 ) and prediction(89. 6% ).