[Application of cluster analysis to chemical identification of the drugs from fruits of Zingiberaceae].

The present paper deals with an application of cluster analysis to the chemical identification of the drugs from fruits of Zingiberaceae. Thirty numerical characters were obtained from thin layer chromatograms of volatile oils and flavonoids. The coefficient of Seuclid was used to show quantitative index of similarities among the samples. By methods of systematic cluster, 4 dendrograms from each method were created respectively. The results show that the fruit drugs from Amomum are similar to each other, but different from those from Alpinia; the results from median method are in accordance with the classification of fruit drugs from Zingiberaceae by their traditional effects. This paper may be helpful in seeking new resources as well as in identifying the fruit drugs from Zingiberaceae.