基于三维图形化数据挖掘方法的四物汤类方配伍规律研究

Study on Compatibility Laws of Siwu Decoction Categorized Formulae by 3-dimension Visualization Data Mining

目的:研究用于不同血瘀证证型的四物汤类方的配伍特点,从组方药物性味归经及功效等属性研究各类方之间的共同特征。方法:在本实验室开发的二维方剂图形化挖掘方法的基础上,进一步增加了药物的功效属性,分别以性味、归经、功效属性为坐标,将方剂信息表现在三维图形上。分别将用于各类血瘀证型的方剂转化为三维图形,并寻找各类血瘀证型方剂对图形中的共同节点,分析四物汤类方治疗血瘀证的用药特点。结果:通过适用于不同证型的各类方间的三维方剂图形比较,发现了部分四物汤类方治疗血虚血瘀证的部分共同特征。组方药味性味以辛、甘、温为多,大多归肝、心、脾经,功效以补血、活血、止痛为主。这类节点在各证型类方中均出现较多,平均每方出现0.5～1.5次。这些节点可代表四物汤类方治疗血瘀证的一般特点,与传统理论相符。结论:三维图形化数据挖掘方法具有直观、形象的特点,便于从复杂数据中发现规律,非常适用于类方数据挖掘研究。

Objective： To study on the compatibility characters of Siwu decoction categorized formulae for different Xueyu Zheng. The common elements of nature, flavor, channel tropism and effects from drugs in formulae were to be discovered. Method： Based on the 2-dimension （2D） visualization methods and formulae information database, the 3-dimension （3D） methods were developed and applied in this study. Result： Some common characters among categorized formulae were found by comparing different formulae information graphs for different Xuexu Xueyu Zheng. The drugs in formulae often had pungent, sweet and warm in flavor and nature. And these drugs most followed liver, heart and spleen channel tropism. Conclusion： The 3D visualization mining methods were verified suitable for the traditional Chinese medicine formulae data mining research.