摘要
目的研究当归芍药散组方药味关联规律,为进一步研究当归芍药散组方原理提供基础。方法通过检索中药方剂数据库中与当归芍药散组方药味有关的方剂,构建当归芍药散药味配伍数据库,运用神经网络、决策树、关联规则、聚类分析等数据挖掘技术,利用SPSS Clementine和SPSS软件,研究其组方药味之间的关联规律。结果相互支持度,当归对芍药、芍药对当归为强支持,茯苓对白术、川芎对当归、当归对川芎、白术对茯苓、泽泻对茯苓为较强支持;对全方的贡献度,当归>茯苓>芍药>川芎>白术>泽泻;聚类分析表明泽泻对当归芍药散全方关联最强。结论在当归芍药散组成中,当归、茯苓和芍药为主要成分,泽泻为特征成分,川芎、白术在其中发挥重要作用。
Objective To study the relative rules among the ingredients of Dangguishaoyao Powder and provide a basis for further research. Method The relative rules among the ingredients of Dangguishaoyao Powder were studied by applying data mining technology, including neural networks, decision tree, association rules and clustering analysis, and software of SPSS Clementine and SPSS. The database of ingredient combinations of Dangguishaoyao Powder was established through retrieving the formulas related to the ingredients of Dangguishaoyao Powder from the Database of Chinese Formulas. Result In the aspect of mutual support rating, Radix Angelicae Sinensis to Radix Paeoniae Alba, and Radix Paeoniae Alba to Radix Angelicae Sinensis showed a stronger support rating; Poria to Rhizoma Atractylodis Alba, Rhizoma Ligutici to Radix Angelicae Sinensis, Radix Angelicae Sinensis to Rhizoma Ligutici, Rhizoma Atractylodis Alba to Poria, and Rhizoma Alismatis to Poria showed a strong support rating. In the aspect of contribution rating to the whole formula, Radix Angelicae Sinensis 〉 Poria 〉 Radix Paeoniae Alba 〉 Rhizoma Ligutici 〉 Rhizoma Atractylodis Alba 〉 Rhizoma Alismatis. The outcomes of clustering analysis showed that Rhizoma Alismatis had the strongest relevancy to the whole formula of Dangguishaoyao Powder. Conclusion In Dangguishaoyao Powder Radix Angelicae Sinensis , Poria and Radix Paeoniae Alba are the main ingredients, Rhizoma Alismatis is characterized ingredient, and Rhizoma Ligutici and Rhizoma Atractylodis Alba play important roles.
出处
《北京中医药大学学报》
CAS
CSCD
北大核心
2009年第5期299-301,共3页
Journal of Beijing University of Traditional Chinese Medicine
关键词
当归芍药散
组方
药味
关联规律
数据挖掘
Dangguishaoyao Powder
formula forming
formula ingredients
association roles
data mining