search

OBOFoundry / OBOFoundry.github.io

Metadata and website for the Open Bio Ontologies Foundry Ontology Registry
http://obofoundry.org
Other
160 stars 201 forks source link

Features for classifying an ontology - replace reference vs. application #400

Open ramonawalls opened 7 years ago

ramonawalls commented 7 years ago

The simple classification of references versus application ontology is ambiguous and no longer very useful. At the operations committee on 3/7/17, we discussed having a type of checklist for classifying axes along different axes. Below is a first draft of said list. Feel free to chime in with other suggestions.

Purpose (not necessarily mutually exclusive): -Intended to serve as a reference for a broad domain (e.g., animal anatomy, biomedical research) -Intended to serve as a reference for a specific domain (e.g., biobanks, plant phenology) -Intended for a specific application, but may be useful more broadly -Intended for a specific applicaiton, probably not useful more broadly

Term requests: -Accepts term requests. -Accepts term requests but may funnel them to a more general ontology -Accepts term requests but only funnels them to a more general ontology -Does not accept term requests

Terms in own namespace: -Yes -No -% terms in the ontology’s namespace

MIREOTs: -Yes -No -%of terms imported from other ontologies

Integration: Does the ontology add OWL axioms that connect external terms? -Yes -No -Number of added axioms connected to imported terms.

cmungall commented 7 years ago

This is great. What are the next actions?

nlharris commented 4 years ago

Is this something that still should be done?

nlharris commented 2 years ago

What is the status of this? Is this still something we want?

matentzn commented 2 years ago

Holy!!! I didnt realise this issue was discussed as early is 2017! And nothing ever happened. We absolutely should address this.. See also https://github.com/OBOFoundry/OBOFoundry.github.io/issues/1546

ramonawalls commented 2 years ago

This issue has been discussed since I can remember (which is about 2010).

"They've been fighting for 2,000 years. It can't go on much longer."

matentzn commented 2 years ago

Oh maaaaan! No dictators out there that are willing to take the plunge and just decide that there are reference ontologies and "other" ontologies, and "other" ontologies terms have to be ignored by term browsers and curation tools? It would be so useful to have these categories in OBO...

cthoyt commented 2 years ago

Bioregistry is pretty pedantic about this! I can be the dictator you always wanted! The only important thing in that context is whether an ontology mints its own identifiers or not. However I think the other dimensions that Ramona outlined are excellent and much more decidable than “application” ontology or “reference” or “other”

matentzn commented 2 years ago

"Application ontologies" will mint IDs. They just need to be ignored by downstream tools!

And we still need a name for this.

nlharris commented 2 years ago

(I probably shouldn't have reopened this can of worms, huh?)

matentzn commented 2 years ago

@cthoyt I would also be the dictator if anyone would care about my opinion, but we need the senior OBO people to make a decree. But since we cant even agree on a name for so many years about what exactly it is we are trying to prevent.. It will be hard.

nataled commented 2 years ago

The only input I can provide on this--with regard to a historical perspective--is that from the beginning of the current OBO governance structure (Operations, TWG, EWG, OWG), the EWG understood that the principles definitely applied to reference ontologies, but might not work for application ontologies, and that there might be tweaks (or even a few new ones) needed for the latter.

nlharris commented 1 year ago

I think this will be addressed by the new home page ontology tables, right?

nlharris commented 1 year ago

I believe this is now addressed by the new home page, right?

nataled commented 1 year ago

@nlharris I'd say not, as there's nothing about the new page that addresses the issue. If anything, it somewhat hides it. Besides, this goes beyond website considerations. Indeed, as noted above, the EWG long ago recognized that the differences will affect how ontologies are reviewed (automated or not).

nlharris commented 1 year ago

Ok.

nlharris commented 7 months ago

I still think this is now addressed by the new(ish) obofoundry.org page organization, but @nataled may still disagree.

cthoyt commented 7 months ago

I still don't think this is addressed

nataled commented 7 months ago

Yep, I still disagree. I'm not sure what it is about the new page organization that allows us to distinguish reference vs application. Indeed, since this is really a policy issue, there's no way a website tweak can solve it.

matentzn commented 7 months ago

The website does indeed not solve the issue. The only thing that will happen though is that starting 01.01.2024, cc @erik-whiting, we will visually separate ontologies that pass the dashboard from those that don't. This will draw another kind of line that perhaps will vaguely approximate reference vs application ontology (not conceptually, just empirically).

I still think this distinction should be made, I just don't have the energy to fight for it. I think what we all want is unmaintained ontologies that are not part of the grand unified vision of ours and are mostly application ontologies to get neatly separated from reference ontologies (ontologies we all import to reuse terms). It just needs an enthusiastic heroic member of OBO to push this idea somehow.