Ontology meeting 2014-07-24

From GO Wiki
Revision as of 07:37, 28 July 2014 by Rfoulger (talk | contribs) (→‎Overriding obsolete check in TG)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Attendees:

Minutes:

Why is assert inferences empty this week?

OWLtools build issues

Overriding obsolete check in TG

Do we want to relax the TG check for whether an obsolete term with the same name already exists? Can we allow editors to override?

Agreed that we don't want to relax this check completely since it's a very tiny number of cases where we'd want to allow this (e.g. when a curator submits a term in freeform TG and we obsolete and get them to re-submit via template TG). Work-around for now is to prefix the term name of the obsolete term (melanosome assembly) with obsolete (obsolete melanosome assembly). Protege also has a problem with identical term names, so that will fix that. In the future, we may want to put 'obsolete' on the front of all obsolete terms but for now I've (Becky) just changed the problematic term (GO:1990382).

Rationale for creating subtypes of protein complex

Did we agree some guidelines for creating 'buckets' under protein complex? I think we can all agree we don't want to persist with a mahoosive flat list of complexes (ala 'protein' in PRO) so we need some basis on which we choose groupings - did we decide MF groupings only? See [1].

Migration of GO annotation extensions relations editors file to OWL

Protein complex to GO term relations

  Rachael: "for annotation of a protein complex entity to a GO:protein complex term, the default annotation 
  part_of doesn't seem correct as you would be saying the complex is part_of GO:<x> complex. 
  Should we be using is_a in these cases?"

As we discussed last week, annotators need a mechanism for choosing is_a/part_of (located_in/colocated_with). I was asked about the mechanics of this by Rachael and Ruth yesterday but realised I wasn't completely clear about them myself.

In GAF/GPAD - presumably this choice has to be made in the qualifier column. The mapping to OWL is clear, but obviously we can't have an RO ID for is_a.