Oct 27th, 2015 Ontology Working Group Meeting

From Planteome.org
Jump to navigation Jump to search

Planteome Ontology WG Zoom Meeting

  • Time: Oct 27th, 2015 8:15 AM (GMT-7:00) Pacific Time (US and Canada)
  • Connection info: Join from PC, Mac, Linux, iOS or Android: https://zoom.us/j/672187671

attendees 

 - AM, MAL, LOL, PJ, JE

Planning for the Fall Workshop in Corvallis

Discussion of the Agenda: see link Corvallis Ontology Workshop Agenda- Nov. 2015- DRAFT

- Workshop - no poster session (presentation instead) - karthika, alfulla, omar, Vilma, rex, leo 

 - excel sheet
 - put together a survey for background on participants (google forms)
   - have people bring data set
 - major objectives
   - AmiGO 2
     - web services
     - enrichment search (MySQL version 1.8-1.3 -> move it over to AmiGO)
   - make a dummy TO for everyone to play with on GitHub
   - everyone will add terms and do updates with Protege
   - Everyone get comfortable
   - workflow between CO and Planteome (protege vs. excel)
     - do people work in excel, then convert to protege?
   - PJ's vision - everyone edits ONLY the TO (reference)  Then slices of it will be specific for each species.
   - determine a template for TermGenie to allow mapping cross products term creation.
   - Blow the TO up to reflect PATO.  (taxonomy vs. biology centric)
 - long term goals
   - terms won't be created by curators, but by users.
   - They will edit the crop ontology files and add 
   - use CO ID for crop ontology
   - use TO ID for them adding TO terms
   - all terms they want from CO should be in the TO (PJ)
   - Barry wants TO to stay as a reference ontology, and specific terms stay in CO

From Chris Mungall:

  • Git Hub training- command line vs Git tool or Sourcetree
  • GO training course was 2 full days

Ontology term discussion

- ontology (compound leaves)- no dennis = no discussion

Tabled for the next meeting

Progress on Updating AmiGO 2 Browser

  • Follow up call 10-27-15 with JE, LC, SC, CM

Questions about Customization of the Planteome AmiGO 2.x version:

From the email discussion:

  • Customizing templates: Is there a way to just specify a new template path? I can modify the files in the current templates folder, but this seems like the wrong way to do it.
    • SC:Yes there is, although it has not been specified as a config variable (we can change that)--we kept this usecase in mind during the last major refactor. That said, it might be doable to simply have a fork that changes the template pathing variable--that way you could stay current and still make arbitrary changes.
    • Seth is making the change in upstream AmiGO on git. As long as our fork is merged and we moved to match current master, this will happen soon.
  • xrefs: How can I add new xrefs so that the links in the "Mappings"tabs work?- In 1.5, I would modify Namemunger.pm, in 1.8 I would modify go-db-perl/t/data/GO.xrf_abbs then run a modified make_dblinks.pl that didn't pull the GO.xrf_abbs file from CVS new each time, but this always seemed like a terrible hack on my part. Even so, I don't see a comparable method in AmiGO2. I'm sure there is a better way to do it.
    • SC: We're all about yaml files now https://github.com/geneontology/go-site/blob/master/metadata/db-xrefs.yaml Depending on the exact use, the YAML files are only part of it: we also have all sorts of customizable low-level overrides for various display handlers, as well as easy overrides in the JS linker "class". Again, it's easy enough to isolate that particular YAML pick-up as a variable and generalize it all.
    • Discussion also pointed out that we can modify this in git and have it pushed to upstream. All xrefs would then be globally available.
  • What adds the text to "Last file loaded on"? On mine, it just has "???". I thought I found this once, but can't find it again, and thought this would be part of the golr/Makefile.
    • SC: This is generated by a specially structured (simple TSV) file generated by the loader; if found, as specified by a runtime variable, it will display the date, etc
  • Filters not being saved on clicking back- we already have issue tracker open: #50- Is this something that can be fixed easily downstream??
    • SC: No, it's something that I'm personally not a huge fan of behavior-wise. We have a fairly good bookmarking mechanism now that could help out with that, but some nice re-draw code would have to be added so it wasn't excessively "jerky".
  • Being able to pin multiple filters, i.e. being able to select both "plant_anatomy" and "plant_trait_ontology" from the list of "Ontology sources" at the same time, or being able to select multiple species in the annotation search. I suppose this works by selecting the "-" rather than the "+" and choosing all the ones we don't want, but it seems a bit more intuitive to choose the ones we do want rather than those we don't. Plus we have something like 50 species, and probably will add more moving forward.
    • SC: In addition to what Chris mentioned, there are a couple of options that could be explored at the manager API level that might not be too disruptive, but would require a little work, see: #5. While icky, this is something that can also be specified at the Solr level, but I wouldn't.

Points from the Meeting

  • Question and discussion about the best way to keep in sync with the revised version of AmiGO2- recommend making a new fork with the new version (2.3x) and adding in the customizations
  • Timeline for rolling out??
  • Using master vs a branch?
  • Goal is to move main Planteome page to the AmiGO 2.0
  • Do we need a separate SolR instance? Need to go through Sys admin- Chris Sullivan
  • term enrichment
  • Customizing the look and feel for Planteome
  • Need to see that the links + mappings are working e.g. MaizeGDB links go to archive site
    • Discussion of the YAML file


  • Issue Japanese characters not rendering properly-
    • Japanese fonts were fixed in master.

Revisions to the Trait Ontology

Tabled for the next meeting

Other upcoming meetings and workshops

    • PAG 2016
    • ICBO 2016

- other meetings (ICBO, PAG) 

 - Monday evening (jan 11th) working dinner meeting
 - have MAL reserve a room at PAG
Retrieved from "https://wiki.planteome.org/index.php?title=Oct_27th,_2015_Ontology_Working_Group_Meeting&oldid=2126"