Annotation Conf. Call 2018-11-13

From GO Wiki
Jump to: navigation, search

Meeting URL

GO Conference Calls

  • Tuesdays at 8am PDT
    • 1st Tuesday: Alliance Biological Function
    • 2nd Tuesday: GO Consortium
    • 3rd Tuesday: GO-CAM Working Group
    • 4th Tuesday: GO-CAM Working Group
    • 5th Tuesday: Alliance Biological Function
  • One Zoom URL for all - https://stanford.zoom.us/j/976175422

Agenda

Review Action Items, Decisions, and Future Plans from Montreal GOC Meeting

Annotation QC/QA Reports

Annotation Review

  • Discussed how priorities are set for annotation review
  • Use labels on go-annotation tickets to indicate which annotation reviews are high priority.
  • Some annotation review tickets are marked 'high priority'. This means that the Annotation Group would like curators to review these annotations as soon as possible. Tickets are marked high priority when they are relevant to major ontology projects (for example, in 2018, signaling and transcription), or when annotations are incorrect and should rapidly be fixed. The requestor of the annotation review can set the high priority level, but the manager of the annotation group can decide to increase/decrease the priority of any ticket.

ND Annotations

  • https://github.com/geneontology/go-annotation/issues/2045
  • Decision: For now, we will not automatically remove ND annotations, even when an experimental annotation also exists
  • If curators add an experimental annotation to a gene/gene product that has a root node ND annotation, if they made the ND annotation, then they should remove that when adding the experimental annotation.
  • If a curator didn't make the original ND annotation, please notify the group or curator that did, so they can remove the root ND annotation.
  • For MF annotations and genes/gene products that only otherwise have a 'protein binding' MF annotation, we will allow co-existence of root MF and protein binding (or a child term) but would like to review those annotations to see if more informative MFs could be made.
  • It would be helpful to have a report telling us what genes have both an ND annotation and an experimental annotation to better understand the scope of the issue and help decide how best to handle this going forward.
  • Proposal:
    • Have two child terms of Molecular Function: 1) binding and 2) molecular activity
    • If a molecular activity is not known, curators could annotate ND to molecular activity
    • If binding annotation exist, they could co-exist with an ND to molecular activity
    • We would no longer directly, manually annotate ND to Molecular Function
    • Implications?

GPAD/GPI as Main Annotation Exchange File Format

  • Goal is to move to GPAD as the main annotation exchange file format within the GOC
  • Some changes to the current GPAD specifications were discussed during one of the breakout sessions
  • A formal specifications document with proposed changes will be presented for comments
  • Timeline - roughly mid-to-late spring of 2019

Progress Reports

  • Due December 19th
  • Will get back to everyone on preferred file format, location

Minutes

  • On call: Chris, David, Dustin, Edith, Giulia, Harold, Kevin, Kimberly, Laurent-Philippe, Li, Marie-Claire, Michelle, Midori,Pascale, Patrick, Petra, Rob, Sabrina, Shur-Jen, Seth, Stacia, Stan, Suzi A, Tanya

Review Action Items, Decisions, and Future Plans from Montreal GOC Meeting

Annotation QC/QA Reports

  • We reviewed some of the content of the QC reports
  • Not all rules are currently listed in the html page; some are just contained within the more comprehensive report
  • Suzi A shared some examples where the error as described in the report could perhaps be made clearer
  • It would also be very helpful to curators to have information on exactly which annotation file was used for the report, as there may be synchronicity issues between when the report was made and the annotation file was generated

Annotation Review

  • We discussed how priorities are generally set for annotation review
  • We'll review the existing, open high priority tickets to see if they can be closed or if there are still outstanding issues that need to be addressed

ND Annotations

  • We discussed ND annotations
  • Agreed that we will not automatically delete an ND annotation when an experimental annotation also exists, but these cases will get flagged for curator review
  • MF presents a more complicated case as 'binding' annotations are not universally thought of as MFs, and so some groups are not comfortable with removing ND when only a binding annotation will remain
  • We discussed the proposal of creating two main branches of MF, roughly equivalent to binding and molecular activity, and allowing curators to make an ND annotation to molecular activity even if experimental annotations exists to binding
  • This statement would be interpreted as an indication that we don't know the molecular activity of a GP, but there is information about its physical interactions
  • David expressed concern about structuring the ontology in such a way as to allow for this.
  • Val also raised the issue of needing to have genes annotated to the root in order to have full representation of a genome in GO.
  • We did not resolve this issue/proposal, but will need to continue the discussion to find the best solution here.

GPAD/GPI as Main Annotation Exchange File Format

  • Work on the specs continues.
  • The goal is to have specs available for review by next month so groups can comment.
  • We are moving towards using GPAD/GPI as our main file exchange formats and would like to keep on track to do this by mid-to-late spring next year.

Progress Reports

  • Due December 19th
  • Will get back to everyone on file format, location, templates