目的建立川产麦冬野生资源的化学模式识别方法。方法采用高效液相色谱法,建立川产麦冬野生资源的HPLC指纹图谱,使用SPSS17.0软件对26批不同来源川产麦冬样品的HPLC指纹图谱进行主成分分析,在主成分得分系数矩阵的基础上,对其进行系统聚类分析;并使用相似度评价软件进行相似度评价验证。结果 26批麦冬样品共提取出4个主成分,被分为4大类;与共有峰直接聚类相比,主成分-聚类分析更符合相似度评价结果。结论 利用SPSS软件对麦冬HPLC指纹图谱进行主成分-聚类分析,所建立的模式识别方法,操作简便,统计结果具有可靠性,可对麦冬化学计量学分类及其质量评价提供有效参考。

Objective To establish the method of chemical pattern recognition for wild germplasm resource of Ophiopogon japonicus in Sichuan Province.Methods To establish chromatographic fingerprint of germplasm resource of O.japonicus in Sichuan by HPLC,according to the HPLC fingerprint,26 wild germplasm resources of O.japonicus from different habitats in Sichuan were comprehensively compared by SPSS 17.0 software with principal component analysis and cluster analysis;and the correctness of "principal component-cluster analysis" was varified by the software of similarity evaluation.Results Four principal components were extracted from 26 samples of O. japonicus and the samples were divided into four major categories;Compared with the direct clustering according to common peaks,the results of "principal component-cluster analysis" were more closer to the truth of similarity evaluation.Conclusion The method of chemical pattern recognition is used for "principal component-cluster analysis" according to HPLC fingerprint of O.japonicus in Sichuan by SPSS 17.0 software.Its operation is simple and convenient,the result of statistics is reliable and able to provide a reference for the chemometrics classification and quality evaluation of O.japonicus.

四川省育种攻关项目“特色种质资源收集整理与创新利用”(2006yzgg12-12)