Compatibility Analysis Of "Treatise On Febrile Diseases" Based On Complex Network

Objective: "Treatise on Febrile Diseases" was written by Zhang Zhongjing.It is Chinese first medical work with complete "the theory of principle,method,formula and herb".The internal Traditional Chinese Medicine(TCM)theory of "Treatise on Febrile Diseases" is based on the "Holistic Thinking" and "Treatment based on Syndrome Differentiation",involving many factors such as "syndrome","formulae" and "herb".which contains complex relationships.Complex networks have been widely used to discover the inherent laws of TCM.Based on the Complex Network theory and Bipartite Network theory,this topic constructs a network model of "Treatise on Febrile Diseases".And clarifies the relationship among the " the theory of principle,method,formula and herb " by analyzing the network topology and the structure of the community.Methods:1.Collect the formula information involved in the "Treatise on Febrile Diseases".2.Construct a Herb Nertwork,and introduce a definition “herb influence” to identify important herb nodes.And introduce a method of Herb Nertwork division based on overlapping communities.Use this method to extract the community structure contained in the Herb Nertwork to discover the structure of the herb relations.3.Construct a Multilayer Bipartite Network of "Treatise on Febrile Diseases".And calculate the network's node degree,degree distribution,average distance,clustering coefficient to confirm the complex network characteristics of "Treatise on Febrile Diseases".4.Introduce a method of bipartite network community discovery algorithm based on node importance and label propagation(I?LPAB algorithm).Use this method to extract the community structure in the real data sets such as Southern Women,Scotland,Crime,Graph,and compare the results with the IAA algorithm,BRIM algorithm,LPA?B algorithm,to verify the effectiveness of the I?LPAB algorithm.Apply the label propagation algorithm to the Multilayer Bipartite Network of "Treatise on Febrile Diseases",extract the community structure contained in it,and analyze the Zhongjing formulae and dialectical law contained in the community.Results:1.It was found that the Herb Nertwork can fully express the relationship between the herbs in the formulae,and herbs with high influence in this network are in line with Zhang Zhong-Jing's medication principle.And the method of Herb Nertwork division based on overlapping communities can extract the high-quality overlapping community structures in the Herb Nertwork,and most of the herbs in the same community belong to the common herbs group.2.It was found that the Multilayer Bipartite Network of "Treatise on Febrile Diseases" can fully express the relationship among the “the theory of principle,method,formula and herb”.And the network conforms to the power-law distribution,and has the small average distance and large clustering coefficient.3.Compared with the IAA algorithm,the BRIM algorithm,and the LPA?B algorithm,the community division effect obtained by the I?LPAB algorithm is more stable and higher quality.And the label propagation algorithm can extract the community structure in the Multilayer Bipartite Network of "Treatise on Febrile Diseases",and the nodes in the same community have certain similarities in herb properties,functions and formulations.Conclusion:1.The analysis of complex network topology indicators confirms that the "Treatise on Febrile Diseases" relationship has complex network characteristics among the “the theory of principle,method,formula and herb”.And the complex network method can be used to express the relationship among the “the theory of principle,method,formula and herb” in "Treatise on Febrile Diseases"2.The analysis of complex network community structure confirms that there is a high-quality community structure in the "Treatise on Febrile Diseases" network.And the analysis of complex network community structure can be used for the discovery of syndrome differentiation and treatment of "Treatise on Febrile Diseases" and the law of herb compatibility.