An method study on exploring potential ingredients with correlation network based on TCM data designed with dose-effect relationship

Abstract

Traditional Chinese Medicine (TCM) can effectively treat and improve diseases. It is currently believed that the therapeutic effect of TCM is the result of the synergistic effect of effective ingredients. According to the theory of dose-effect relationship, there is a correlation between effective ingredients and biological effects. Therefore, the theory of dose-effect relationship is one of the key ideas for exploring effective ingredients in traditional Chinese medicine. In addition to chemical separation techniques, gray correlation analysis is a common analysis method for exploring the effective ingredients of TCM or natural medicines. However, the collinear phenomenon in the dose-effect relationship data of TCM will reduce the ability of gray correlation analysis. Constructing an analytical method for the discovery of effective ingredients of traditional Chinese medicine is still the scientific problems that need to be solved urgently in the technology for discovery of TCM based on the theory of the dose-effect relationship. This study believes that the collinear phenomenon is the basic feature of the effective ingredients of traditional Chinese medicine, and correlation network analysis can solve the collinear problem among the effective ingredients of traditional Chinese medicine. In this study, a batch of experimental data of TCM injection designed with a dose-effect relationship was taken as an example, and then processed according to the following steps to demonstrate the method of analyzing the effective ingredients of TCM based on correlation network analysis. First, the correlation analysis method is used to explore the collinear ingredients, and then the correlation network is constructed using the ingredients as nodes and the correlations between the ingredients as edges. After that, the collinear ingredients are clustered and divided into different sub-networks. The feature nodes of the sub-network was extracted to construct a linear regression analysis to explore the effect-related ingredients, further the dose-effect curve of the effective ingredients was drawn to optimize the discovery method of the effective ingredients. Preliminary research results show that this method is suitable for the discovery of effective ingredient in TCM designed with the theory of dose-effect relationship experiments.