Research On Visual Retrieval Method Of Biomedical Entity Relationship

With the rapid development of Internet technology and medical technology,a large number of biomedical literature and clinical data are produced every day.These data are of great significance for studying the relationship between biomedical entities.With the improvement of living standards,people pay more and more attention to their health.In 2015,former U.S.President Barack Obama put forward the "Precise Medical Plan" in his State of the Union address.He hopes that precision medicine can lead a new era of medicine.Precision medical treatment is a medical model and new medical concept that is based on individualized medical treatment and has been developed with the rapid progress of genome sequencing technology and the cross-application of biological information and big data science.At present,there are many popular biomedical databases,including but they all have their own flaws.First,each database contains only a small area.Second,the database retrieval capacity is limited,some databases only support each query returns a result,the user needs to manually integrate the results.Finally,the current database simply returns the result of a static list,and the user can not filter the result for the purpose of personalized search.In order to more quickly find the rapidly growing knowledge of biomedical literature,we developed a Biomedical Entity Relationship Retrieval Tool(BERST).It is also a biomedical knowledge integration framework and currently contains five popular databases(CTD,HMDD,BioGRID,GO and HPRD)and a genetic data set containing similarity data stored in medical literature Extracted medical concepts and network of relationships.Users can retrieve the converged knowledge network by entering keywords,and BERST will return a subnet that matches the keywords and represents the relationship between the keywords.Users can explore the returned graph of results in an interactive way to explore specific potential and interesting path relationships between any two nodes.We have developed a graphical user interface that allows users to view and research information more intuitively and comprehensively.The BERST framework can support the integration of biomedical knowledge from other sources.BERST is a biomedical entity indirect relation visualization tool implemented in Java Web technology.Building BERST is divided into the following major steps:(1)Data fusion: The data of medical entities from different sources need to be fused together.The biggest challenge is to solve the problem that the same medical entities have different names in different databases.(2)Entity Relationship and Network Construction: The entity relationship is stored in the database by the triplet.The network construction of the entity relation is to gradually load the different types of entity relation in the database into the memory.By means of recursive iteration,we bulid the network including different types of entity relationship.(3)Path Retrieval,Sorting: The retrieved path is scored by keywords input by the user,then sorted,and the sorted path is returned to the user according to the user's requirement.(4)Visual Retrieval: BERST returns the results to users in graphical form.Compared with the lengthy list form,BERST can intuitively see the relationship between medical entities,including direct and indirect potential connection relation.