We present the rst RDF representation of DisGeNET, a gene-disease database designed to integrate the current knowledge of human diseases. DisGeNET RDF data introduces a harmonized and semantically enriched description of the gene-disease association concept into the Semantic Web (SW) by means of the DisGeNET ontology. The centric view on gene-disease associations allows to nd links between diseases and genes easily and with added semantic content. The RDF representation follows the Linked Data (LD) principles that provides new opportunities for data integration, querying and crossing DisGeNET data with other external RDF datasets. The RDF version of DisGeNET has been developed in the context of the Open PHACTS project to provide disease relevant information to the knowledge base on pharmacological data.
DisGeNET RDF is presented as a new resource in the Linked Open Data space
to promote the discovery of key knowledge in the understanding of the
molecular mechanisms underlying a disease or an adverse drug event. DisGeNET [
The RDF version of DisGeNET is represented as a set of triples around the
gene-disease association concept. Information such as genomic variation or
scienti c publications supporting the association are related to this main
concept. Data is organized in a hierarchical manner around ve concepts: gene,
disease, pathway, disease class, and gene-disease association as the parent
concept (see DisGeNET RDF web interface http://rdf.imim.es/DisGeNET.html
for details on the RDF schema). The `RDF-ization' has been done using RDFS
and OWL languages, common ontologies and vocabularies, and following the
Linked Data principles (http://linkeddata.org/). To identify resources in
DisGeNET, URIs established by the identi ers.org e ort were selected
whenever possible to support its initiative to foster a uni ed use of URIs by the SW
community [
The DisGeNET gene-disease association ontology harmonizes the semantic
description of the di erent types of associations between genes and diseases. The
ontology provides foundational support for the DisGeNET database. This
ontology was integrated in the increasingly used Semanticscience Integrated Ontology
(SIO) [
It is considered good practice to provide provenance information to an RDF
resource. The provenance description of the DisGeNET RDF dataset declares
the database from which it is derived, the development date, the current version
and updates, the software used for its development, the license information,
the SPARQL endpoint location, the number of triples, etc... The provenance
description of the original database and each primary source is also tracked. This
information is provided using the Vocabulary of Interlinked Datasets (VoID) [
To map the relational database content into RDF triples, we used the D2RQ
platform (http://d2rq.org/). The dump les and the VoID description of
DisGeNET are loaded into the OpenLink Virtuoso RDF Quad Store [
DisGeNET RDF: a new member in the LD community
We present the potential of DisGeNET Linked Data with di erent uses cases. The rst use case is aimed at answering the following question: give me all the gene-disease associations from the Comparative Toxicogenomics Database (CTD) that have information on sequence variation and provenance. This question can be answered exploiting the information contained in DisGeNET, selecting the associations provided by CTD and that have annotations on genomic variation and literature provenance. This is translated in the following SPARQL query: SELECT DISTINCT ?gdassocIRI ?dName ?gName ?snpLabel ?PMIDLabel WHERE { } ?gdassocIRI sio:SIO_000628 ?disease,?gene . ?disease rdf:type ncit:C7057 . ?disease foaf:name ?dName . ?gene rdf:type ncit:C16612 . ?gene sio:SIO_000205 ?genenameURI . ?genenameURI rdfs:label ?gName . ?gdassocIRI sio:SIO_000001 ?snp . ?snp rdf:type ncit:C18279 . ?snp rdfs:label ?snpLabel . ?gdassocIRI sio:SIO_000253 ?source .
FILTER regex(?source,"ctd") ?gdassocIRI sio:SIO_000772 ?PubMedID .
?PubMedID rdfs:label ?PMIDLabel .
The inclusion of DisGeNET into the 'Web of linked data' using the most
advanced SW technologies brings the opportunity to integrate our gene-disease
data with other disparate data sources spread over the Web by performing
federated queries. The second use case is aimed at answering the following question:
give me all the gene-disease associations in which the association is linked to
changes in the expression of the gene, there is expression information, and
sequence variation linked to the disease. This question can be answered by querying
and integrating data from DisGeNET and GXA (http://www.ebi.ac.uk/gxa/)
databases. More speci cally, we ask for the DisGeNET gene-disease associations
labelled as 'AlteredExpression' and their related SNPs, and the GXA expression
values. See the SPARQL query at the DisGeNET RDF web interface. DisGeNET
SPARQL endpoint supports the syntax and semantics of SPARQL 1.1 for
executing queries distributed over di erent SPARQL endpoints. SPARQL queries
such these are aimed to be included in Bioqueries which is a portal aimed at
gathering SPARQL queries [
As the last use case, DisGeNET RDF has been implemented in the Open Pharmacological Space (OPS) discovery platform, which is a SW platform developed under the Innovative Medicines Initiative (IMI; http://www.imi.europa. eu) funded Open PHACTS project. Remarkably, the integration of DisGeNET in OPS is essential to answer important research questions such as which compounds could e ectively inhibit targets involved in a key pathway for the development of a disease. Aimed at exploring and querying DisGeNET data across the linked data in the platform, APIs are currently under development (see the API website for up to date information at http://dev.openphacts.org). This is expected to be fully operative in the upcoming OPS 1.5 release. 4
We present DisGeNET RDF, a new Linked Data dataset that provides genedisease association data to answer relevant scienti c pharmacological complex questions. Importantly, DisGeNET has been implemented in the pharmacological LD discovery platform developed within the Open PHACTS project. Acknowledgments. The research leading to these results has received support from the IMI Joint Undertaking under grant agreement n 115191, Open PHACTS, resources of which are composed of nancial contribution from the EU FP7 (FP7/2007-2013) and EFPIA companies' in kind contribution; and the Instituto de Salud Carlos III FEDER (CP10/005249). The Research Programme on Biomedical Informatics (GRIB) is a node of the Spanish National Institute of Bioinformatics (INB).