wiki:CuratorMeeting

Version 636 (modified by vw253, 6 years ago) (diff)

--

PomBase Curator Meetings

PreviousCurationMeetings

Next Items, in rough order of priority

Curation priorities

  • Approval of community sessions
  • Processing phenotype screens
    • Down to one assigned to Jurg (community) and one where I have a couple of questions to discuss (midori 2015-03-25)
  • Identification of stalled community curation sessions and processing of any term requests in these
  • General curation priorities based on biology
  • Fixing FYPO specificity https://sourceforge.net/p/pombase/chado/419/ (in progress -midori)
  • Fixing GO specificity (I have done most of these)
  • Helpdesk tickets
  • logs, antonia 50, midori 51 Val 52 reminder, include the GO terms we don't annotate (canto only)

In Progress, parked

  • which other relations need to be included in Canto? ( so far only has_part, is this in the live tool yet or test only?) All keep a look out for this in the tool, and think about which others we want to see.
  • Discussion about search behavior for phenotypes (alleles, conditions, multigene, and GO extensions Decided:
    • For now we will only provide allele type filter (ignore conditions, penetance, expressivity) for phenotype, Midiri will do a mock up (null expression/overexpression)
  • Discussion for how to handle searches on GO extensions was postponed, and will be addressed when we go through the relations list and decide what type of queries might be useful
  • For multi gene phenotypes so far decided
    • will need to be able to search on a phenotype and then filter on one of i) single gene phenotype. Single gene phenotype only will allow the allele filter option. II) single or multi gene phenotype iii) multi gene phenotype only
  • we are not currently sure what the results page should show. Two types of output, geneotype list and genes. A summary of the geneotype list and the gene list? discuss this next.
    • The gene list will populate the current query history table. but the query history will need an additional table for geneotypes (what do we want to be able to do with geneotype lists in the history?)
    • need the ability to filter results to present genotypes with a specific gene

  • Do we want to group some phenotypes (normal?) https://www.ebi.ac.uk/panda/jira/browse/PB-1837 do we have a preferred ordering
    • Suggestion: group by putting all gene expression phenotypes together at the bottom, and possibly normal/abnormal
  • want to avoid terms which are not 'disjoint'

  • Did we decide for sure weather 'abolished' means (to the best of our knowledge..) 'does not start at all' or 'may start but never completes' (I tend to use it in the latter sense, but Midori seemed to favour the former?).
  • Things to do after multi-gene phenotypes available ==
    • update curation tool documentation
    • get some example multigene phenotypes visible on pomBase pages ( where is display mock-up )
    • update pombase web documentation to reflect new in
    • announce multi gene phenotype availability
    • make podcasts to demo curation tool (announce on pombelist)
    • send out session reminders, and new sessions
    • target individual groups who have not participated, or who have stalled sessions

  • annotate to protein complex ID (can get from PRO)?
  • example: PMID:18723846 has direct assays showing ORC binding to origin DNA, but doesn't distinguish subunits; from other work, it's known that some but not all ORC subunits actually make contact with DNA. Annotate all to DNA binding anyway, with or without contributes_to, or annotate to a complex ID? (I'd prefer the latter but will it break anything?)
  • same paper: similar issue for annotation extension - cdt1 or cdc18 increase ORC origin binding - put complex ID in has_regulation_target ext?
    • Next step here is to request complex IDs for known complexes from PRO
      • Then need to discuss what is needed to annotate to complexes (Canto, Complex pages- which data will be propagated to complex pages from gene pages and vice versa)

Next Tasks

  • conflicting extensions used for 'binding of promoter' 'binds' and 'at', see cdc10 and res2
  • dropbox

  • multgene phenotype format input/output (there is one dataset for this....which one?)

  • should we filter redundant ISS? (I think we should, but we would want them to take precedence over IEA, currently filtering is not specific for evidence code)

  • Q when we say ‘arrest’ what do we mean?, should all 'arrest' phenotypes distinguish between normal and abnormal (should cell cycle arrest phenotype, and some children, be not for direct annotation)

Ontology questions

Annotation issues, syntax, procedures, extensions etc

  • protein domain specific binding, see example in e-mail
  • discuss Rama's question about "response to" extensions
    • also discuss what is being regulated here .....I don't think it is -ve reg of the transition (this acts through cdc2 and not through trancription)
  • PENDING Val will look into this during G1/S cell cycle updates (not to self, background thread is in curation meeting folder)
  • longer term: decide what IDs we actually want to enter at curation time (and whether Canto or Chado should do any fancy mapping taking ontology term and extension relation into account)

Curation issues biology (by 'slim' or broad annotation topic)

Cell Cycle

Transcription

  • Sushil Tripathi's thesis and e-mail "2 MF transcription terms, sound the same"

Chromatin Organization

  • regulation of histone modifications., GO or phenotype? /curs/97535c919f7436fe thread

Establishment and/or maintenance of Cell Polarity

  • discuss establishment of cell polarity and the signalling network (MOR)

(see e-mail Created new session 31b82829d0c47e4d from PMID:23649273 )

Documentation

  • Canto documentation required
    • From Scott: provide even one paper and walk the user through entering the paper, the genes and a sample entry. Then you could also walk through an administrative approval and saving the data to Chado (Kim suggested we start with just a list of bullet points)
    • and https://sourceforge.net/p/pombase/curation-tasks/365/
  • Tidy wiki, move obsolete pages

Outreach

  • plan roadshow dates/places
    • Val will get UK lab list (done)
    • Midori will draft e-mail (done)
    • after BYGa and Toronto, will begin to pin down dates. Antonia will try to set up some whilst at BYG

  • Newsletter?
  • Gene naming 3578 genes primary name assigned (now 3671)
  • annotation complete genes
  • annotation amnestly
  • solicit more HTP datasets
  • unknowns
  • curation progress (numbers),
    • community, new tool, focussed efforts, areas, networks, resources allocation, reduced staffing numbers

PomBase website

  • Future Gene page changes
    • column sorting (click header etc.)
    • user configuration (page section order/ compact vs. expanded/show/hide page sections)

Community curation paper

  • Community curation spreadsheet: Community Curation errors spreadsheet replace
  • Talk about spreadsheets. We should be able to use Canto to report differences between curator and community now all annotations have recorded person. The only thing we won't be able to capture is when a curation is wrong an d is deleted and replaces so we will need to continue to monitor this?
  • Discuss paper outline (contents)

Lower priority

  • Discuss the item about capturing phenotypic observations vs. Phenotypes thread PMID:19033384 (postponed, needs more thinking, we don't completely know what to do about it)

  • collection of phenotype images
  • Low penetracne check (I think this is about population phenotype-> fully penetrant?)

  • Idea for accessing completed curation sessions add review links to paper section for community curated papers
  • Future Pombelist posts
  • gene name descriptions to phenotypes, add to phenotype ontology for searching?

  • Val- update on phenotype mapping, where it is, what to do with removed annotations
  • Antonia- Summarize community curation session quality/ ommissions
  • Val- viability exercise results (query intersection)
  • PHAF/GPAD/GPI files for export/import etc.
  • advanced search, better to have just query management ?
  • Genetic interactions ontology
  • how to represent the reintegration assays ? (see for example Alper/Partridge and Shanker/Partridge
    • decided for now just to represent the deletion phenotype, and then do a GO process for the 'reestablishment etc.