Projects stand up meeting 2021-02-24

From GO Wiki
Jump to navigation Jump to search

Attendees

  • Present: David, Huaiyu, Kimberly, Pascale, Seth, Suzi, Judy, Paul, Chris
  • Regrets:
  • Managers: David, Huaiyu, Kimberly, Pascale, Seth, Suzi, Judy, Paul, Chris

Projects presentation

Template:

  • Next deliverables + expected date
  • Progress
  • Issues / blockers

GAF2.2 migration status

  • https://github.com/geneontology/pipeline/issues/211
  • Kimberly manually testing GAF2.2 input, output WB files as well as GPAD1.2 output WB files
    • Still going to double-check Protein2GO output gpad files
      • Follow-up with UniProt
        • Annotations to root nodes (Pascale emailed Alex 2021-02-24)
        • Updated go-rule0000059 to account for annotations to root nodes (Kimberly will QC for WB files - should see this in rules report, when the code is updated - where/how do we know when go-rule0000059 is updated?)
        • Annotations for external2go - make a proposal for each mapping and submit to groups that create the mappings for explicit approval (Pascale, Kimberly, David)
  • David manually testing files that MGI ingest from all resources (e.g. GOC, UniProt) and subsequent release of MGI GAF2.2 (and also GPAD/GPI2.0)
    • David will set up a meeting David, Darren, Kimberly, Seth to discuss issues with GPI2.0 specs
  • Outstanding questions:
    • How will GO handle different kinds of errors that may occur with gp2term relations in GAF2.2 files?
      • Repair?
        • For example, the gp2term relation for a root node annotation is not correct according to our GAF2.2 specifications.
      • Remove?
        • For example, the gp2term relation for non-protein-containing complex CC terms is 'part of'.
          • This has not been addressed wrt rules yet. TBD
    • For future file testing, what can be automated? Is it worth it?
    • Testing these files raised some issues about how things actually happen wrt the pipeline, e.g. which PAINT files should be compared for input/output; we said we'd switch to GAF2.2 in March, what is happening with GPAD/GPI2.0 in March? If groups only produce a GPI2.0, will that break NEO production?
      • Chris has updated NEO build pipeline to consume GPI2.0 as well as GPI1.2.
    • Once the documentation is on the website, what should we do with the documentation page in github? Add status update for other GAF file formats? Delete?
      • Final documentation lives on the GO website. AI: ask Seth about deleting branch issue-2917 (will deal with GPAD/GPI2.0 specs when we get there)
    • Where else should we announce the file format change? Other forums? Send email again?
      • Yes, but we need to follow-up and decide where. Specifically add Alliance email to our announcements?
    • Double-check what GAF file format each submitting group is actually submitting (roll this into broader outreach task).

Pathways2GO manuscript

ART1.0/Noctua3.0 plan

Items not discussed at last meeting

Imports

UniProt

  • No plan yet

VEuPathDB

  • Got in touch this week and would like to start using Noctua

Other resources (Yeast pathways)

History

GPAD export

=

Next sprint

https://docs.google.com/document/d/1AIuq8xxgV6bZZHFbXW-s3ln8PdUfZGNUuZcNFX1n7eQ/edit#