GO versions are very out of date

[dosumis]

GO: date: 03:11:2015 11:44
GO-plus: date: 27:08:2015 08:12

Can they be kept up to date?

Also, please could you display the version IRI where this is recorded? This is a much more reliable way to track versions that the date field. (Version IRI from OBO (foundry) ontologies should be resolvable to the original file).

Cheers

[leechuck]

This bug was due to a too small timeout (30s) for reloading the ontology; it's now fixed by setting timeout to 300s.

Version IRIs are a good idea, will add them asap.

[leechuck]

@dosumis We have two versions of GO-PLUS in AberOWL, http://aber-owl.net/ontology/go-plus and http://aber-owl.net/ontology/GO-PLUS. Both seem to be manually uploaded. I changed GO-PLUS (in capitals) to sync from http://purl.obolibrary.org/obo/go/extensions/go-plus.owl, and leave go-plus (lowercase) as it is.

[leechuck]

@dosumis New version of GO-PLUS has still one unsatisfiable class (down from 17 in the version we had before), where to report?

[leechuck]

Closing the issue and open a new one for version IRIs.

[dosumis]

I can't see any inconsistent classes in the latest:

http://purl.obolibrary.org/obo/go/releases/2016-05-05/extensions/go-plus.owl

Do you have a list of the classes you've detected and which versions they effect? These shouldn't make it through the release pipeline.

[leechuck]

In AberOWL, querying GO-PLUS for owl:Nothing returns http://purl.obolibrary.org/obo/GO_0042628, 'mating plug formation'. I have no access to Protege today, but will check manually when I have the opportunity.

[dosumis]

Thanks. Seems it didn't show up in Protege for me because of a missing disjointness axiom (between single and multi-organism process). Adding it back makes 'mating plug formation' inconsistent. Not clear to me why you're getting this axiom and I'm not. I checked local vs URI based resolution of x-disjoint.owl, where this should live, but couldn't find it in the latest version from the URI.

@cmungall - any ideas? Also, is it possible that inconsistent GO-plus versions sometimes get through the release process? I thought these should be blocked at the first Jenkins check of the editor's file + imports.

Anyway I'll add back the disjoint and fix the issue with 'mating plug formation'. There are no other new inconsistent classes This fix will be in tonight's release.

[leechuck]

@reality Would be good to add some documentation on using AberOWL for "Reasoning as a Service" for these kinds of quality checks as part of build processes. Should be part of #13

[cmungall]

I'm offline right now, but from my local possibly stale copies of the
file I see we have a

DisjointUnionOf(BP, SOP, MOP)

i.e.

BP = SOP or MOP
SOP DisjointWith MOP

If you were only looking for explicit DisjointWith axioms David you
might have missed this

Similarly, some reasoners (Elk?) may ignore DisjointUnionOf (even though
this is a shortcut that can be expanded to one EL++ axiom and one
non-EL++ axiom). This might explain discrepancies. Or maybe there is
some code somewhere that does the axiom expansion.

On 6 May 2016, at 2:59, David Osumi-Sutherland wrote:

Thanks. Seems it didn't show up in Protege for me because of a
missing disjointness axiom (between single and multi-organism
process). Adding it back makes 'mating plug formation' inconsistent.
Not clear to me why you're getting this axiom and I'm not. I checked
local vs URI based resolution of x-disjoint.owl, where this should
live, but couldn't find it in the latest version from the URI.

@cmungall - any ideas? Also, is it possible that inconsistent GO-plus
versions sometimes get through the release process? I thought these
should be blocked at the first Jenkins check of the editor's file +
imports.

Anyway I'll add back the disjoint and fix the issue with 'mating plug
formation'. There are no other new inconsistent classes This fix
will be in tonight's release.

---

You are receiving this because you were mentioned.
Reply to this email directly or view it on GitHub:
#45 (comment)