Open cmungall opened 7 years ago
Looks interesting, although the focus of Ontolib ist narrower. Our main use case is in supporting ontology similarity searches.
I had no success finding the big ontologies on obofoundry in this format, so far. Could you point me there?
On 27 Jul 2017, at 14:32, Manuel Holtgrewe wrote:
Looks interesting, although the focus of Ontolib ist narrower. Our main use case is in supporting ontology similarity searches.
I see. The benefit for you is less as you have already implemented an obo parser, but one of the goals is to prevent people having to do this (you actually did this in a robust fashion with a grammar, but most people use regexes that break).
I had no success finding the big ontologies on obofoundry in this format, so far. Could you point me there?
It's not yet well advertised until we gather sufficient feedback from experience bioinformaticians and ontology users such as yourself, as we don't want to publicize and then make breaking changes. You can try for some ontologies by following the http://purl.obolibrary.org/obo/foo/foo.json pattern, e.g. http://purl.obolibrary.org/obo/go/go.json - in some cases this works: http://purl.obolibrary.org/obo/go.json.
Hm, ist there any documentation on allowed values in the strings?
The OBO format defines how dbxref etc. are structured and what allowed values are. This appears to be missing from the Json schemas.
OBO Format is legacy, this is the replacement: https://github.com/geneontology/obographs/
For context: https://douroucouli.wordpress.com/2016/10/04/a-developer-friendly-json-exchange-format-for-ontologies/
It should be much easier to support this than the existing obo parser here: https://github.com/Phenomics/ontolib/tree/master/ontolib-io/src/main/antlr4/de/charite/compbio/ontolib/io/obo/parser
No antlr grammar required, it's just json
Note: once you have implemented this you automatically have OWL support too, if you add the obographs jar