Annotation Conf. Call 2016-08-09

From GO Wiki
Jump to: navigation, search

Bluejeans URL: https://bluejeans.com/993661940

Agenda

Meetings

Next GOC Meeting - USC, Los Angeles, CA, November 4-6, 2016

  • Please indicate on the Meeting Logistics Page if you plan to come.
  • We are also gauging interest in a one- or half-day Noctua/LEGO workshop at USC (either before or after the main consortium meeting) so have added two additional column to the table to see if people can also attend that, and if so, when.
  • Please add your information as soon as possible so we can make arrangements.

UK LEGO/Noctua Training

Annotation Consistency Exercise for August 2016

  • August 23 - SGD

http://www.ncbi.nlm.nih.gov/pubmed/23246437

Please send annotations to Stacia Engel by end of week prior (EOB F 19Aug2016).

Membrane Cellular Component Annotations and More Generally, Chain of Evidence Annotations

  • This annotation issue was raised by Rebecca and concerns annotation to membrane child terms, e.g. integral to synaptic vesicle membrane, given different types of experimental results.
  • More generally, though, the annotation issue here concerns what we have often referred to as chain-of-evidence annotations, which means annotations that result from curators considering the experimental result as well as additional information such as sequence motifs or domains.
  • There is a nice summary of the membrane annotation issue from Rebecca on the 2016-07-11 minutes.
  • For today's call, we would like to have a discussion on this general issue.

Minutes

  • On call: Chris, David H., Elena, Helen, Kimberly, Li, Judy, Midori, Paola, Petra, Rebecca, Ruth, Sabrina, Stacia

UK LEGO/Noctua Training

  • To all attendees who are staying at the Hinxton campus, please reply to Jean's email to confirm your arrival/departure dates
  • Meeting agenda is now up on the wiki
  • Anyone else who might be interested in attending, please add your name to the logistics page

Annotation Consistency Exercise for August 2016

  • SGD selected a paper (see link above) - please send Stacia your annotations by Friday, 2016-08-19

Membrane Cellular Component Annotations and More Generally, Chain of Evidence Annotations

  • We discussed how to make annotations where information from more than just the experimental result in a given paper could inform the choice of GO term
  • Lots of points to consider here:
    • Traceabiliity of annotation process - how would users know exactly what knowledge was used for annotation to a given term?
    • New combinatorial evidence codes - if we created new evidence codes to capture different combinations of knowledge, how many would we need? Would it add to the annotation workload for curators to use these additional codes?
    • Is use of author intent, or use of background biological knowledge, reasonable for selecting more granular GO terms than an experiment might show directly?
    • Which annotation process allows for the best representation of the biology?
    • Can we cite more than one reference for a given annotation?
    • How much do evidence codes really matter, anyway?
    • For use of a given evidence code, what are the effects on annotation propagation via PAINT or other sequence-based pipelines?
  • In addition to the specific examples that Becky cited for integral to membrane annotations, Midori will provide some examples
  • In general, though, here are the options we have so far:
    • Continue to use existing evidence codes, but allow curators to bring in external knowledge to select a term; would need to make sure our annotation documentation reflects this practice
    • Expand the use of IC to allow for use of sequence-based annotations from automated pipelines, such as InterPro2GO (also explore automated ways of making these annotations)
    • Work with ECO to create new combinatorial evidence codes that reflect the full annotation process (also explore automated ways of making these annotations)
    • Allow for more than one reference to support a given annotation
    • Explore LEGO representation of these use cases - how much does it help?
  • Midori will send David and Kimberly some use cases from pombe