Ontology meeting 2012-03-07

From GO Wiki
Jump to: navigation, search

30 minute call


Announcement: TermGenie now sends an e-mail to the original term requester after the terms have been approved and committed. E-mail text is fine for now.

DISCUSSION ITEM I: Moving biological_process_xp_cellular_component into live GO

  • Chris to start discussions on moving the logical definitions for the internal cross products from scratch to the live GO.

  • To do:
    • check scratch file to make sure they are ok
    • option 1: hacky perl script exists to check the bridging files - obo-check - Jane knows how to run
    • option 2: load up corresponding files (importer file in /extensions for this purpose) into Protege, run ELK and look at inferences to see if they are normal
    • option 3: use OORT and look at inferences (no good way to look at why the inference was made) - main GO file, OBO file in CVS for bp_xp_cc file

See: https://sourceforge.net/tracker/index.php?func=detail&aid=3487992&group_id=36855&atid=440764

id: GO:0070072 ! vacuolar proton-transporting V-type ATPase complex assembly
intersection_of: GO:0022607 ! cellular component assembly
intersection_of: results_in_assembly_of GO:0016471 ! vacuolar proton-transporting V-type ATPase complex

DISCUSSION ITEM II: ChEBI changes: modified amino acid

From Marcus:

  I've been thinking again about our use of the term 'modified amino 
  acid', which both ChEBI and (from what you said earlier) GO seem to have 
  used as a repository for any derivative of an amino acid.  I now really 
  need to tidy this up within ChEBI, but I need to know whether the 
  changes that I'm thinking of would have an effect on the GO.
  I would like the term 'modified amino acid' to be renamed in accordance 
  with its original intended meaning, probably as 'modified canonical 
  amino-acid residue' and then to populate it accordingly.  The vast 
  majority of the entries currently under 'modified amino acid' could be 
  conveniently moved into one or other of two new disjoint classes of 
  'canonical amino-acid derivative' or 'non-canonical amino-acid 
  derivative', both is_a children of 'amino-acid derivative'.  (However, 
  just what the parent of 'amino-acid derivative' would be I'm not quite 
  sure about yet - we certainly couldn't use 'amino-acid derivative' is_a 
  'amino acid' because of the the inclusion of amides etc here.  The 
  'canonical amino-acid derivative' and 'non-canonical amino-acid 
  derivative' classes could have 'has_functional_parent' relationships to 
  'canonical amino acid' and 'non-canonical amino acid' terms 
  respectively, both children of 'amino acid'.
  <snip> What do you think? If I can make this work 
  within ChEBI, would it also work within the GO?
  One other thing that worries me, and that is with the terminology.  Is 
  'canonical amino acid' synonymous with 'proteogenic amino acid' and if 
  so which is generally preferred within the world of biology?  If they 
  are in fact subtly different, which would be the more appropriate term 
  to use within our ontologies?

TO_DO: David has this sorted and will talk to Jane who will walk down the corridoor and talk to CHEBI.


  • The annotators would like to see tags on terms that are pending obsoletion, particularly for large-scale edits where the obsoletions don't occur within a 2 week window.
  • Where can this information go, so it's immediately obvious to annotators? Can it go in existing fields?

After discounting the option of having it as a subset (because subsets aren't visible enough to annotators), decided to add this to the COMMENT filed with a syntax: PENDING OBSOLETION. [optional free text to explain to annotators why the term is up for obsoletion, and suggest using column 16 or alternative terms]. If the date of the obsoletion is known, this should be included in ISO format (2012-03-07).

  • AI: Paola to work out the syntax and send it round.

At next meeting, we may want to try and reach a consensus on "regulation by regulation"

See http://gocwiki.geneontology.org/index.php/Ontology_meeting_2012-02-15 and minutes.

TO_DO: We'll move this to the top of next weeks 60minute call, where we'll have a group-session of Protege to try these out. In the meantime, have a play around with Protege to come up with ideas.



  • David
  • Chris
  • Heiko
  • Tanya
  • Becky
  • Paola
  • Karen

TO_DO: David and Jane are collating a list of tasks to take to the GO-tops so they can decide project priority. Need to sort out how to describe 'maintenance'. Chris has already added the conversion to OWL.

AI: Send David a list of your tasks. The cut-off for granularity is that the task takes at least 1 FTE per month. David will discuss these with Jane when she's back from hols.