Closed stain closed 9 years ago
Hi Stain,
That was an issue that was not fully resolved.
Ideally they should be merged/changed into a single URL There is no reason to have these different.
We originally started using: http://openphacts.cs.man.ac.uk:9090/ontology/DataSource.owl# as we had control of that
http://vocabularies.bridgedb.org/ops# fits more with the bridgedb branding but that is not a Manchester controlled site.
Christian
PS. I leave it to the current project team to pick the best way to clean what is agreeably a mess up.
My draft - in case it doesn't exist:
I assume there that the 'datasource' ontology is separate.. yet there are things in the constants that sounds like they should are also be in bdb, e.g. fullfillsLens
Thanks for the details, @brenninc.
OK, I'll complete the above ontology with the remaining datasource terms, and then we can call it the "Bridge DB ontology" - which match how it is described as "bdb" now. I might fix the odd spelling mistake.
I'll add support to the BridgeDB RDF code for both the old and new namespaces - probably some kind of warning for old ones?
@egonw - would you be the right person to (eventually) modify what the http://vocabularies.bridgedb.org/ops# points to - or who should I ask?
vocabularies.bridgedb.org. 7200 IN A 137.120.17.33
33.17.120.137.in-addr.arpa. 28800 IN PTR vocabularies.bigcat.maastrichtuniversity.nl.
Great work! @stain, here's one more file with terms to check for consistency with the ontology: datasources_headers.txt. You probably know better than me what these terms should be, but let me know if I can help update the file to be consistent with the BridgeDb ontology.
My current mappings from datasources_headers.txt to ontology terms can be seen here: https://github.com/bridgedb/bridgedbjs/blob/master/lib/dataset.js#L29
While you are looking at this one thing to consider is the use of bridgeDB:hasUrlPattern and bridgeDB:hasUriPattern The difference was because a classic BridgeDb DataSource only has one UrlPattern.
May be worth considering a better clearer naming here,
as in bridgeDB:DataSource_Chembl_Target a bridgeDB:DataSource ; bridgeDB:fullName "Chembl Target" ; bridgeDB:systemCode "ChemblTarget" ; ...... bridgeDB:hasUrlPattern http://data.kasabi.com/dataset/chembl-rdf/target/t$id ; bridgeDB:hasUriPattern http://chem2bio2rdf.org/chembl/resource/chembl_targets/$id ; bridgeDB:hasUriPattern http://linkedchemistry.info/chembl/target/t$id ; bridgeDB:hasUriPattern http://rdf.ebi.ac.uk/resource/chembl/target/$id . bridgeDB:hasUrlPattern http://data.kasabi.com/dataset/chembl-rdf/target/t$id ; bridgeDB:hasUriPattern http://chem2bio2rdf.org/chembl/resource/chembl_targets/$id ; bridgeDB:hasUriPattern http://linkedchemistry.info/chembl/target/t$id ;
Thanks, I was a bit confused of hasUrlPattern
vs hasUriPattern
- is there still a need for the first one?
Perhaps the primary
property on a pattern should be changed to a hasPrimaryUriPattern
subproperty of hasUriPattern
?
My draft for the BridgeDb DataSource ontology: https://github.com/bridgedb/BridgeDb/blob/OpenPHACTS/develop/org.bridgedb.rdf/resources/ontology.ttl
@AlasdairGray - perhaps you could have a brief look? For the terms from datadesc spec I included the documentation as is, but added the friendly skos:Concept
as range on hasDataType
, linksetJustification
, objectsDatatype
and subjectsDatatype
- as a kind of soft OWL/Full preventor as those would typically go towards other Classes and Properties.
TODO: rdfs:comment
for the terms that are not in http://www.openphacts.org/specs/2013/WD-datadesc-20130912/#bridgedb (but then I would need to understand them all.. :-)
@stain Said "Perhaps the primary property on a pattern should be changed to a hasPrimaryUriPattern subproperty of hasUriPattern?"
That would have my vote if I still had one.
So I changed hasUrlPattern
to hasPrimaryUriPattern
, a subproperty of hasUriPattern
.
The primary
property is something different, that I'll leave as is.
The "primary" property was not used by OPS but included for backward comparability and to ensure that the OPS version still implemented the existing BridgeDb API's
Assigned to @egonw who said he would have a look at publishing this at http://vocabularies.bridgedb.org/ops#
The result is online at the requested link. I started with the data source TTL files but in the end also included @stain's ontology.ttl terms.
@stain, is there anything left in the http://openphacts.cs.man.ac.uk:9090/ontology/DataSource.owl# namespace that was not in your ontology.ttl yet? If so, please reopen and let me know.
@ariutta, you want to use the ontology, you can perhaps use the https://github.com/alexmilowski/green-turtle HTML+RDFa parser in JavaScript?
Thanks, Egon. I will use that. On Sep 27, 2015 11:27 AM, "Egon Willighagen" notifications@github.com wrote:
@ariutta https://github.com/ariutta, you want to use the ontology, you can perhaps use the https://github.com/alexmilowski/green-turtle HTML+RDFa parser in JavaScript?
— Reply to this email directly or view it on GitHub https://github.com/bridgedb/BridgeDb/issues/13#issuecomment-143582731.
Two namespaces:
.. used for terms like
linksetJustification
,subjectsDatatype
and more BridgeDb-specific likehasIdentifiersOrgPattern
andsystemCode
.I don't know if these two namespaces are the same or not. But the fact is that none of them resolve (both give 404), yet they are used here:
Which one should it be? Has there been any attempt to create the ontology? Did
vocabularies.bridgedb.org
use to resolve to something else than an empty Drupal site?