Study on classfication based on data of an traditional Chinese medcine by similarity network

Abstract

Traditional Chinese medicine (TCM) plays an important role in the world medical system. It is believed that the treatment of specific diseases may be the result of the joint action of multiple components. It is the basic hypothesis of traditional Chinese medicine on the disease treatment that multiple ingredients work together to treat diseases. As the most important theory in the discovery of active ingredients in traditional Chinese medicine, Dose-effect relationship theory holds that the effect of drugs is related to dose. Based on this theory, biological effect can correspond to multiple active ingredients as long as they have similar rules. Therefore, it is one of the key scientific problems for the discovery of active ingredients based on dose-effect relationship to explore the change rule of active ingredients in traditional Chinese medicine. The project team put forward a scientific hypothesis that the change rule can be explored based on the correlation network of active ingredients. Based on the data detected by mass spectrometry designed with the theory of dose¬effect relationship, correlation analysis was used to analyze the correlation of ingredients, and then the correlation network was constructed with correlated ingredients. Then, ingredients were classified into different subnetwork according by clustering analysis method, and the change curve of ingredients in the subnetwork was drawn to visualize change rule of ingredients in the subnetwork, and lay the foundation for the discovery of effective components of traditional Chinese medicine.