Study The Compatibility Law Of The Atrophic Lung Disease Based On Herb Community Discovery

Traditional Chinese medicine (TCM) is a holistic medical approach and the formula’s composition discipline is still a mystery. Traditional methods on mining TCM data usually only study the symptom-herb relationship, herb frequent item set and association rule. However, few people study the compatibility law of a prescription, which is also crucial in the development of traditional Chinese medicine. Detecting a formula’s structure and herb communities/clusters in TCM Formula networks (TCMF) is becoming a mainly existing problem in data mining of the data sets. In recent years, more and more attention are paid by the international community to TCM, which is also promoting the development of TCM.Traditional TCM data mining methods only study the relationship of herbs in a prescription while ignoring herbs’ own attributes(such as herb taste, tropism and efficacy). These attributes play a key role as hidden attributes in TCM research. In this paper, we devise a model considering herb attributes to construct the TCM Formula network. On the basis of this work, we choose to combine random walk model with hierarchical clustering method. We apply this method on atrophic lung disease data and get some compatibility law from herb communities. In this work, we also pretend to provide a prospective idea for other TCM researchers.The mainly work of this paper has the following several parts:1) Preprocessing of the consumptive lung disease data, solving problem of "the processing of drug synonyms" and "the standardization of effect terms" to get standardized data. This work lays the foundation for later research, and try to provide the basis for other TCM data mining.2) Extract three herb attributes including "taste", "tropism" and "efficacy", construct vector space model on these herb attributes. Define the similarity of herb attributes to detect similar herbs. Finally get some useful results:such as the replacement general herbs of expensive herbs.3) Devise an improved TCM network constructing method based on DMIM, define a novel network constructing model. Then construct the TCMF network and analyze the statistical characteristics of the network.4) Devise a novel community similarity calculating method in the process of clustering, which is called Random Walk Hierarchical Clustering (RWHC) algorithm, to identify herb communities by using clustering algorithms based on the formula network of atrophic lung disease. And we also use classic NG modularity function to evaluate the experimental results.The studies suggest that the TCM network clustering approach provides a new research paradigm for mining TCM data from an experience-based medicine, will accelerate TCM drug discovery, and also improve current drug discovery strategies.