Ontology meeting 2018-06-04: Difference between revisions

From GO Wiki
Jump to navigation Jump to search
 
(3 intermediate revisions by the same user not shown)
Line 114: Line 114:
* If we are going to use the dbxrefs to seed an alignment between GO Rhea and Reactome, at least for MF, they should be 1:1.
* If we are going to use the dbxrefs to seed an alignment between GO Rhea and Reactome, at least for MF, they should be 1:1.
* Forward and reverse reactions will each have a 1:1 relation to the corresponding relation in Rhea
* Forward and reverse reactions will each have a 1:1 relation to the corresponding relation in Rhea
* Ontology editors would need to come up with exceptions where 1:1 alignment would NOT work


=== GO subsets (slims) ===
* Pascale has created a yaml file for all the slims
* The pull request is failing, though
* Which slims are ''actively'' maintained?
** This means someone uses it and someone is responsible for maintaining it
* What is the difference between the alliance slim and the generic slim?
* Pascale will check with Chris to make sure the format, i.e. fields, of the yaml is okay, but she has tried to standardize this





Latest revision as of 13:06, 4 June 2018

Conference Line

Agenda

GH project link

https://github.com/geneontology/go-ontology/projects/1

Editors Discussion

Ontology Developers' workshop possibility- Geneva in September?

  • David, Pascale and Kimberly will discuss possible dates and make a doodle poll.

Outstanding Pull Requests

There are now 29 open requests. We are getting better.

  • chris: 15
  • pascale: 5
  • dos: 4 (MF refactoring; Pascale will try to close)
  • eric: 2
  • jim: 1
  • karen: 1*
  • barbara: 1*
  • Need developers to fix


Regenerate Imports File

What is the status of Docker and do we have documentation on its use?

Obsoletion plugin for Protege

Does everyone have Jim's plugin? If not we should all install it and I will update the documentation. What is the status of the merge plugin?

DBxrefs

There was a recent thread about the scope of dbxrefs in the ontology. If we are going to use the dbxrefs to seed an alignment between GO Rhea and Reactome, at least for MF, they should be 1:1.

Did this get covered last week?

GO subsets (slims)

(pull request failing but we can sort this out another time)

  • List of subsets:
  • Active:
    • goslim_agr.yaml -> 59 classes - active
    • goslim_pombe.yaml -> 55 classes -active
    • goslim_synapse.yaml -> 344 classes- active
    • goslim_mouse.yaml -> 45 classes - active


    • goslim_aspergillus.yaml -> 85 classes - email xx
    • goslim_candida.yaml -> 89 classes - email xx
    • goslim_goa.yaml -> not used - email Tony and George
    • goslim_metagenomics.yaml -> 116 classes - who to contact ??
    • goslim_pir.yaml -> 462 classes - email Darren
    • goslim_plant.yaml -> 99 classes - email Tanya
    • goslim_virus.yaml -> 29 classes - email Philippe, Patrick and - others ?
    • goslim_yeast.yaml -> 169 classes - email Stacia and Suzi A



  • Check at ontology call:
    • goantislim_grouping.yaml -> 49 classes, mostly 'x part'
    • goslim_chembl.yaml -> 310 classes
    • goslim_generic.yaml -> 150 classes
    • gosubset_prok.yaml (status: obsolete)
    • mf_needs_review.yaml
    • termgenie_unvetted.yaml -> all terms are (obviously) obsolete. Is this useful ?
    • virus_checked.yaml -> 244 classes

Minutes

  • On call: Barbara, David, Eric, Jim, Karen, KImberly, Pascale, Tanya

Ontology Developers' workshop possibility- Geneva in September?

  • David, Pascale and Kimberly will discuss possible dates and make a doodle poll.
  • August will likely not work since so many people are away on vacation

Outstanding Pull Requests

  • Can we move towards merging and closing these?
  • Recommend that if 'DO NOT MERGE' is added to a pull request, the reason is also included so when we go back to the ticket, we understand why it's loitering

Multicellular Organism Signaling and ChEBI Import File

  • The unmerged request for obsoleting multicellular organism signaling is not working because there is a link to signaling coming from the chebi imports file
  • The link can be removed in Protege but then it will return when the chebi imports is loaded
    • This would lead to an error with the next ChEBI imports update
  • The only file changed locally is the go_edit.obo file
  • Why are we still importing relations from the imports file?
    • We shouldn't be getting these GO relations from ChEBI
  • Jim has been working on improving this workflow but it is still a work-in-progress
  • We need to communicate this change to ChEBI, but even so, this shouldn't prevent the GO editors from working
  • Pascale created a new ticket to move towards not importing GO relations from other ontologies
    • Jim has linked his proposal to this ticket
  • Reminder: bold text in Protege comes from opened file; unbolded text is from imports file(s)

Regenerate Imports File

  • What is the status of Docker and do we have documentation on its use?
  • Do the editors understand what to do here?
    • No
    • If there was a documented command for what to do, that would be helpful
  • How often do we want to rebuild the import files?
  • If every day, would this be problematic?
  • Would ontology editors remake the imports file each time they started an editing session?

Obsoletion plugin for Protege

  • Does everyone have Jim's plugin? If not we should all install it and I will update the documentation.
  • What is the status of the merge plugin?

DBxrefs

  • There was a recent thread about the scope of dbxrefs in the ontology.
  • If we are going to use the dbxrefs to seed an alignment between GO Rhea and Reactome, at least for MF, they should be 1:1.
  • Forward and reverse reactions will each have a 1:1 relation to the corresponding relation in Rhea
  • Ontology editors would need to come up with exceptions where 1:1 alignment would NOT work

GO subsets (slims)

  • Pascale has created a yaml file for all the slims
  • The pull request is failing, though
  • Which slims are actively maintained?
    • This means someone uses it and someone is responsible for maintaining it
  • What is the difference between the alliance slim and the generic slim?
  • Pascale will check with Chris to make sure the format, i.e. fields, of the yaml is okay, but she has tried to standardize this