This CLI (command line interface) provides a generic RDFizer for any Intermine's endpoint See http://intermine.org/
You can install all the dependencies with the gem bundle
gem install bundle
bundle install
ruby intermine-rdfizer.rb -h2 mandatory arguments: -- endpoint and --output
No arguments will do both download and rdfize
ruby intermine-rdfizer.rb --endpoint http://flymine.org/flymine --output flymine-dataDownload only
ruby intermine-rdfizer.rb --endpoint http://flymine.org/flymine --output flymine-data --downloadRDFize only
ruby intermine-rdfizer.rb --endpoint http://flymine.org/flymine --output flymine-data --rdfizeTo create Linked Data - database cross references - you need to provide a CSV file with these fields :
- Type (Ontology or CrossReference) (need those tables in the object model see https://github.com/intermine/intermine/blob/master/bio/core/core.xml)
- DB/Ontology name
- URI-Prefix (for the target ontology/database)
- URI-relation (predicate)
Example of a CSV files :
CrossReference,Uniprot,http://purl.uniprot.org/uniprot/,skos:exactMatch
CrossReference,NCBIgene,http://bio2rdf.org/ncbigene:,skos:exactMatch
Ontology,GO,http://bio2rdf.org/go:,skos:exactMatch
Ontology,SO,http://purl.obolibrary.org/obo/SO_,owl:sameAsTo launch the creation of the dbxref triples creation use the options --dbxref . The output directory needs to contains the *.nq files generated with --rdfize, but can be use at the same time
ruby intermine-rdfizer.rb --endpoint http://flymine.org/flymine --output flymine-data --rdfize --dbxref mapping.csvruby intermine-rdfizer.rb# This script will query an intermine endpoint via their API and transform the data
# into RDF nquads files based on the instance's xml object model file :
# http://examplemine.example.org/mymine/service/model
intermine-rdfizer.rb --endpoint [URL] --output [dirname]
[COMMANDS] omit all will run --download and --rdfize ; --dbxref can be run afterward
--download will ONLY download table TSV files
--rdfize will ONLY RDFize TSV files in output/*.tsv
--dbxref <file> mapping for dbxref [ontology / crossreference] tables in a CSV file...
Need the core.xml object tables Ontology and/or CrossReference and
the NQuads files, generated with --rdfize
Columns :
Type[Ontology|CrossReference], DB/Ontology name, URI-Prefix, URI-relation
Example :
CrossReference,Uniprot,http://purl.uniprot.org/uniprot/,skos:exactMatch
CrossReference,NCBIgene,http://bio2rdf.org/ncbigene:,skos:exactMatch
Ontology,GO,http://bio2rdf.org/go:,skos:exactMatch
Ontology,SO,http://purl.obolibrary.org/obo/SO_,owl:sameAs
[OPTIONS]
--uri <baseuri> specify a base url for the URI (ex: purl.yeastgenome.org)
DEFAULT mymine.intermine.org
--lcoupled loosely coupled, when pass with --rdfize it will makes different
resources for a same record based on the table the info is extracted from.
Example : the same gene, cds, orf won't be merged into a single
resource but kept appart and linked together.
By default --lcoupled is off and the information from the tables will
be merged into a single resource that'd be typed accordingly.