Figure 4

A simplified view of the results from the SDDM process. Five different ontologies are used in the construction of this multi-relationship graph: the Scenario Ontology (abbreviated with a colon, :), Bio2RDF (abbreviated with Bio2RDF_X where X is a namespace tag e.g. if X = PubMed, then Bio2RDF_PubMed = http://bio2rdf.org/pubmed), GeoNames, ProMed, and DBPedia. The SDDM process created all edges in the example, and all nodes in black and white. Bio2RDF (Blue), GeoNames (Green), and DBPedia (Purple) nodes already existed on the Semantic Web. Red nodes representing ProMed mail articles are integrated using an RDFizer as part of the SDDM process. The NDM-1 example created RDF individuals corresponding to Wikipedia articles, InterProScan models, PubMed articles, ProMed articles, Gene Ontology terms, a scenario, catalogs, events, outbreaks, NCBI sequences (gi numbers), and GeoName locations. The nodes and edges diagrammed in this figure are physically stored in 5 locations: Bio2RDF nodes marked in blue are stored on the main Bio2RDF server or an acceptable mirror, GeoNames nodes marked in green are stored on the GeoNames infrastructure, DBPedia nodes marked in purple are stored on the DBPedia servers or a mirror, ProMed nodes marked in red are stored in a local graph database (Neo4j), and nodes and relations identified in the scenario, marked in black, are stored in a local RDF database on the machine where the scenario is built.