LEGO January 4, 2016

From GO Wiki
Jump to: navigation, search

Bluejeans URL

https://bluejeans.com/969313231

Agenda

Review immediate action items from Geneva meeting

  • These were the tickets deemed most critical for wider adoption of the tool
    • LEGO Curation Documentation
      • Need GAF and GPAD outputs - see below on Annotation File Outputs. Need any additional testing/verification by curators?
      • Develop a system to specify isoforms and proteoforms
      • Enable copy-paste for enabled_by entities
      • Document how to get GAF and GPAD from Jenkins
      • Account for legal set of annotation relations

Software updates

  • Notes taken from Seth's emails on 2015-12-23 and 2015-12-24
    • Saving large models
      • As mentioned before, the issues that prevented saving large models should now be cleared (https://github.com/geneontology/noctua/issues/151). In case you manage to still find yourself in a situation where you cannot save, there is now also an emergency [Save] button under [Skunkworks] that should work pretty much no matter what.
    • Macromolecular complex creator
      • I had an error in the macromolecular complex creator workbench that got the edges turned around (https://github.com/geneontology/noctua/issues/265). This has been fixed, but those of you that used the workbench will want to revisit those models.
    • Blank Screens
    • Annotation File Outputs
      • We've added the ability to directly download GAFs, OWL files, and GPA files from the client (https://github.com/geneontology/noctua/issues/251). While be may do something else in the long run, links are now available on the landing page and in the graphical model editor. This is the temporary solution for people who want to add annotation to their pipeline. They look like:
 http://noctua.berkeleybop.org/download/gomodel:5362b2ed00000016/gaf
 http://noctua.berkeleybop.org/download/gomodel:5362b2ed00000016/owl
 http://noctua.berkeleybop.org/download/gomodel:5362b2ed00000016/gpad
    • GAF updates
      • Heiko has tweaked the GAF generation for symbol name, date format, and type.
    • RAS signaling template - JSON payload
      • David, for your model "RAS signaling template", while it is safe on disk (https://github.com/geneontology/noctua-models/blob/master/models/5667fdd400000802), there is something going on where it is a bit garbled coming back (the JSON payload is...odd). It's possibly due to some annotation input? We'll look at this when we get back, but editing other models and new models should be fine (https://github.com/geneontology/minerva/issues/34), just a warning/question is you've been trying anything new.
      • The issue here is that we found a loophole somewhere so that the model written to disk is not quite right, so cannot be directly read again from the filesystem. As long as your model was in memory, there was no issue, but with yesterday's restart we ran into the issue. There will be a discussion about this on #34, but it's not something that you probably need to worry too much about.

Set up schedule for presenting models on future calls

  • One idea: maximize curation efforts and coverage by signing up for specific pathways, processes - would not have to be a single curator

Other items?

Minutes

  • On call: Chris, Dan K., David H., David O-S., Heiko, Kimberly, Pascale, Paul T., Seth, Stacia, Suzi
  • GAF, GPAD, OWL files
    • Available for download from Noctua
    • Ultimately will be accessed from Jenkins or AmiGO with MOD-specific files generated by taxon ID
    • Outstanding issue - attribution of annotations
      • Can we have multiple assigned_by values if several people contribute to an annotation?
      • Don't want to retrofit to old system too much here if detailed information is already available in the OWL files
      • Each annotation has a date-last-modified tag, but not the full history
      • Form a working group to come up with specs for attribution - Pascale, Stacia ??
        • Use orcids?
    • Annotations to root terms
      • Are present in the output files if there is annotation context (i.e. Col. 16 information)
      • What evidence code to use? ND? IC?
      • Do we even want these annotations in the GAFs?
      • Will review specific examples from David
  • Representation of Cleavage Products, Isoforms, Proteoforms
    • Need to be sure we're accounting for this in the model and user interface
    • Right now, Noctua will autocomplete for an isoform if it's included in the database (i.e. previously annotated)
    • NEO contains information wrt gene-centric view of entities and a subset of isoforms
    • Need a working group to document specific use cases and make sure they are handled correctly
      • Cleavage products - Pascale
      • Isoforms - Kimberly
      • Proteoforms - David
  • Legal set of annotation relations for LEGO
    • David drafted a document with relations, definitions, and examples
    • Chris, Kimberly, Paul T. have commented
    • Paul suggests that we triage the list and provide justification for usage
    • This documentation will be discussed at the next LEGO call
  • Grant renewal - Suzi
    • What would we like to see for the tool and the annotations over the next five year period?
    • Send Suzi feedback
      • Transitioning curation process from journal articles to reviews
      • Identifying core, conserved components of pathways using OWL tools
      • Condition-dependent predictions for pathway outcomes - following on from Geneva discussions
      • Other ideas?