Hi! I'm attempting to log the GSSO with a prefix at identifiers.org and part
of that process requires agnostic identifiers.
Basically, when I applied, the GSSO resolved individual identifiers with
URLs such as
instead of http://purl.bioontology.org/ontology/GSSO/GSSO:000096.
I attempted to update this from 2.0.0 to 2.0.1, changing the IDs from
http://purl.bioontology.org/ontology/GSSO/GSSO_006778, in order to mimic
those at the Gene Ontology, but it still requires something like
instead of http://bioportal.bioontology.org/ontologies/GSSO/GSSO:006778 or
What's the best/fastest way to resolve these URLs? i.e. what do I need to
change within the annotations themselves for this to work? I'm editing
~12,000 entries via Protege, so anything I can do non-manually and in bulk
would be incredibly useful!
Thanks so much!
Sent from: http://ncbo-support.2288202.n4.nabble.com/
bioontology-support mailing list
I started looking into this yesterday evening, and I’m not sure I have a complete answer for you yet.
When end users submit ontologies to BioPortal, an ontology PURL is created using this format:
… where your ontology acronym makes up the last fragment of the PURL. If you enter that PURL in a browser address bar, the BioPortal PURL server will resolve it to the following URL pattern:
This is the home page for your ontology in the BioPortal web application (developed in Ruby on Rails).
If you want a permanent link to any given class in your ontology that will resolve to the corresponding page in the BioPortal web application, you can use a URL pattern where you tack the URL encoded class ID onto the end of the ontology PURL using the “conceptid=“ parameter in the query string. For example, the permanent link to the “transgender” class from your GSSO ontology would be the following:
I did find an exception to the above information that occurs with all of the UMLS format ontologies in BioPortal, e.g., SNOMEDCT, etc. These ontologies are bulk-imported into BioPortal twice per year using a custom process that writes the ontology data directly into the RDF store, bypassing the file upload process normally performed by end users. For these ontologies, several formats for permanent class links work, e.g., all of the following PURLS will resolve to the “Body structure” class in the SNOMEDCT ontology:
I’m not yet able to provide more details about why this only works for UMLS format ontologies.
You mentioned that you’re editing a large number of entries in Protege. I’m not sure if you’re aware that we provide a REST API for programmatic retrieval of ontology data, documented here: http://data.bioontology.org/documentation. For example, you could retrieve data about the “transgender” class using the following REST call:
It’s a hypermedia API, and one of the links returned is a “ui” link with a URL that navigates to the page for said class in the BioPortal web application (http://bioportal.bioontology.org/ontologies/GSSO?p=classes&conceptid=http%3A%2F%2Fpurl.bioontology.org%2Fontology%2FGSSO%2FGSSO_000096).
We provide examples of using the API in several languages in one of our GitHub repositories: https://github.com/ncbo/ncbo_rest_sample_code.
Hope this information is helpful.
bioontology-support mailing list
|Free forum by Nabble||Edit this page|