Research On The Integration And Reliability Analysis Of Traditional Chinese Medicine And Western Medicine Associated Data

Traditional Chinese Medicine(TCM)is a precious wealth of human beings.With the development of precision medicine model,the knowledge integration between TCM and Western Medicine has become an inevitable trend.The work based on the macro concept of TCM and the micro concept of Western medicine is a very important part.At present,the knowledge in the medical field is mostly scattered among various heterogeneous data sources,which brings great difficulties to the comprehensive system of knowledge query and analysis.In addition,in the data of Chinese and Western Medicine,some associations such as disease-symptom,prescription-herb and herbchemical composition are determined,while the reliability of disease-gene association needs further study because of its complex influence mechanism.In terms of the overall scale of disease-gene association data,the confirmed association is less than 2% of the total.Therefore,this paper has carried out the following research work around the above problems:(1)Six kinds of TCM and Western Medicine domain ontology were collected and integrated,including TCM syndrome ontology,herb ontology,human phenotype ontology,disease ontology,chemical entities of biological interest ontology and gene ontology.Nine kinds of TCM and Western Medicine domain associated relations were collected and integrated,including syndrome-symptom,syndrome-prescription,prescription-efficacy,prescription-herb composition,disease-symptom,disease-gene,herb composition-gene,TCM symptom-symptoms of Western medicine and herbchemical composition.On the basis of the above data,this paper uses the medical thesaurus matching method to build the integration model of Chinese and Western Medicine associated data based on domain ontology.This model uses the mapping of ontology and associated data at the conceptual level to integrate heterogeneous data sources.Through the B/S architecture,the integrated system of Chinese and Western Medicine associated data is developed,and the query of Chinese and Western Medicine associated data and domain ontology visualization are completed,These will provide the basis for discovering the potential knowledge between the cross field data of Traditional Chinese and Western Medicine,and help to further improve the scientific research of Traditional Chinese Medicine and the level of disease diagnosis and treatment.(2)Aiming at the reliability of disease-gene association in Traditional Chinese Medicine and Western Medicine associated data,In this paper,an automatic reliability discrimination method based on literature data is proposed.This method takes the richness,influence,clinical transformation and other indicators of disease-gene association corresponding to supporting literature as features,and uses the authoritative open database of disease-gene association to label the data set.On this basis,this paper uses a variety of classification algorithms(Logical regression,Support vector machine,Neural network and Random forest)to carry out the experiment of high reliable diseasegene association identification,The results show that the model based on Random forest is more effective,and the AUC of the model is stable at about 0.83 after cross validation.This shows that the reliability evaluation method of disease-gene association based on the literature data proposed in this paper has good discrimination ability,and also has the potential to quickly screen the large-scale disease-gene association with high reliability.At the same time,it is expected to provide important reference for medical researchers and clinicians to accurately grasp the disease mechanism.