Construction Of Applied Ontology For Crowd Cross-sectional Survey And Construction Of Biomedical Ontology Storage And Application Platform

Since 2000 when Gene Ontology was first built,ontology has played an increasingly important role in biomedical data integration as a means of data standardization.Many of the large international biomedical databases,such as UniProt and ENCODE DCC,have ontologies as their back-end semantic support.During the rapid development of biomedical ontologies,the establishment of ontology platforms greatly promotes ontology.On one hand,every platform collects ontologies based on their own rules,resulting in users’ access to ontologies according with different criteria.On the other hand,the platforms supply users various ontology application tools,facilitating their data annotation,data integration and ontology development.This research falls into two parts.First,establish a domestic biomedical ontology repository and platform,which collects ontologies focused on precision medicine,so as to provide ontology application services to domestic users.Second,based on the platform,construct a human cross-sectional study application ontology focusing on the cross-sectional study data stored in National Physique and Health Database to provide semantic support for the database.As for the first part,reusing NCBO BioPortal technology,we constructed MedPortal software framework.We selected precision medicine-related ontologies and built MedPortal ontology repository.In addition,the original code and ontology toolkit were further improved,thus enabling MedPortal to operate stably and process large-scale data automatically.MedPortal ontology repository has been successfully constructed.So far,42 mainstream biomedical ontologies have been imported into MedPortal,mappings among which have also been created.Several ontology services including term search,ontology mapping,data annotation are provided through websites and REST API,accessible at:http://medportal.bmicc.cn.Based on MedPortal ontology platform,we began to build cross-sectional study ontology(HCSSO).Following OBO Foundry ontology construction principles,we choose BFO as upper ontology,reuse terms and relations from related ontologies,use top-down class hierarchy building method and develop HCSSO in 3 steps:semantic modelling,ontology construction and ontology validation.HCSSO contains 676 terms and 21 relations.It uses BFO as its upper ontology and OGMS,IAO and OBI its middle-level ontologies,covering 78%of the database terms.HCSSO describes the main aspects of cross-sectional study such as study design,specimen collection and data gathering.Besides of term standardization,it supports complex queries and automated reasoning.MedPortal has provided services to domestic users for their data integration and ontology construction.Built based on MedPortal,Human Cross-Sectional Study Ontology will provide semantic support for National Physique and Health Database.