Ontology meeting 2016-06-23

From GO Wiki
Revision as of 03:13, 29 June 2016 by Paola (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Attendees:

Regrets: Tanya

Minutes: Paola

Ontology editing and release blues

Attempting a summary (Wed morning, Europe time, so things may change).

What works ok:

  • Releases are now progressing. But we’ll need to change the pipeline (see below).
  • TermGenie templates (is this confirmed?).
Yes
  • Inference Changes Report is now sent, daily, to go-ontology only. It states the number of a) newly added inferences, b) deleted inferences, c) newly added terms. Is b) really needed?
 >  Is b) really needed?
 DOS: YES!

What doesn’t work:

  • Current release pipeline is suboptimal. Any time the release fails, go-plus gets ahead. The path forward is clear, but would rather not change pipeline now, so as not to delay synapse release. Proposals?
Will be re-assessed when Chris is back 
  • TermGenie freeform. Looks like our best bet is to announce this is not available. Or no longer available. Decision? We might want to talk to Birgit about her plans to make sure she has a route to get complexes in while TGFF is turned off. Tell her to open requests in the GH tracker and make sure that the editor-in-chief priorities them? DONE. Still pending: 'disconnect' TG FF; schedule revision of this decision for August?
There wasn't time to put TG FF offline. Curators have been warned not to use it. 
Should any FF term sneak in, editors must remember not to let them through, and to ask the curator to request via GH instead. 

Pending issues we should look into:

  • Can we use OBO-Edit now?
 DOS: I think so.  The new inference file (shared by Chris) doesn't have the trivial inferences of the one generated at the end of last week.  
 Saving from OE seems to work fine.
 There are currently 37670 tagged inferred is_a GO to GO.  Last tuesday there were 40161.  This fits ~ with the number of ETINES deleted.
 I can provide diffs. 
Update: there are still problems with OE, DOS looked into them (thanks DOS) but they're not solved yet. So DO NOT use OE yet.
  • Saving in Protege 5 (DH was having trouble; worked fine for TB)
 DOS: I'm finding Protege 5 more prone to freezing up than P4.  Sometimes takes a minute or 2 to unfreeze.
Up memory?
Has anyone tried saving from P5 to OBO yet and looking at the diffs?
Apparently all is well...
  • Setting_username_and_auto-adding_creation_date. Details from Chris:

http://wiki.geneontology.org/index.php/Protege_setup_for_GO_Eds#Setting_username_and_auto-adding_creation_date

But we never followed up on this: https://github.com/protegeproject/protege/issues/283#issuecomment-177656593

Matt made a nice configurable way for us to recapitulate an important OE feature. I assume we want to choose 'user supplied name' and stick with what we use in OE, with the created_by property. For the date use creation_date and ISO 8601?”

I think we skipped this. Bump to next week or when Chris is back?

Xrefs to Reactome

Looks like this wasn't fully addressed yet. See email thread here:

https://mailman.stanford.edu/pipermail/go-ontology/2016-June/007619.html

https://mailman.stanford.edu/pipermail/go-ontology/2016-June/007620.html

https://mailman.stanford.edu/pipermail/go-ontology/2016-June/007621.html

It shouldn't be done manually.
Peter D'E will be at MGI, David H and Harold can check with him. 
Revisit when Chris is back.