Difference between revisions of "WBConfCall 2020.02.20-Agenda and Minutes"

From WormBaseWiki
Jump to navigationJump to search
Line 10: Line 10:
  
 
== Help Desk ==
 
== Help Desk ==
* [https://github.com/WormBase/website/issues/7579 Estimate of number of C. elegans genes with no function]
+
* [https://github.com/WormBase/website/issues/7579 Estimate of number of C. elegans protein-coding genes with no function]
 
** What criteria would give us a reasonable estimate of that number?
 
** What criteria would give us a reasonable estimate of that number?
 
*** No phenotype, interaction, GO, expression?  Anything else?
 
*** No phenotype, interaction, GO, expression?  Anything else?
 
*** Answering this kind of question requires us to have "complete" curation on certain data types. Are we going to keep this goal with the move into AGR? If "complete" on all data types is not a practical goal, which data types should be targeted to "complete"?  
 
*** Answering this kind of question requires us to have "complete" curation on certain data types. Are we going to keep this goal with the move into AGR? If "complete" on all data types is not a practical goal, which data types should be targeted to "complete"?  
 +
 +
* Interesting discussion on what criteria could be used to define a set of C. elegans protein-coding genes with no known function
 +
* The automated description pipeline currently uses several types of data to write descriptions:
 +
** Orthology
 +
** GO
 +
** Expression
 +
** Disease
 +
* Genes lacking an automated description might constitute a reasonable set to start from, but since phenotype data is not included in the descriptions, there may be false negatives in the gene description set
 +
* Differential gene expression under a certain condition was generally not thought to be sufficient criteria to say a gene has a function
 +
* IEA GO annotations are predictors of function, but not sufficient evidence to say a gene definitely does enable that function (or participate in a process)
 +
* Analysis of the literature could also be informative as it may get at how many genes haven't even been studied.  TextpressoCentral searches could help get this data.
 +
** Even with literature searches, though, we'd need to have some sense of the context in which a gene is mentioned.
 +
* A matrix approach might be particularly helpful here.  Given a set of criteria that might inform function, how many does each C. elegans gene meet?
 +
*
  
 
* [https://github.com/WormBase/website/issues/7581 I was directed here from a link on the CGC]
 
* [https://github.com/WormBase/website/issues/7581 I was directed here from a link on the CGC]

Revision as of 22:11, 20 February 2020

Agenda

Project Meeting

  • Plan is to have Caltech host
  • Doodle poll for April and May dates
  • How many days do we need?
  • Will also host a virtual SAB meeting
  • AI: Start a draft agenda - flesh out by mid-March
    • Link to draft agenda sent out to staff email; please start adding topics

Help Desk

  • Estimate of number of C. elegans protein-coding genes with no function
    • What criteria would give us a reasonable estimate of that number?
      • No phenotype, interaction, GO, expression? Anything else?
      • Answering this kind of question requires us to have "complete" curation on certain data types. Are we going to keep this goal with the move into AGR? If "complete" on all data types is not a practical goal, which data types should be targeted to "complete"?
  • Interesting discussion on what criteria could be used to define a set of C. elegans protein-coding genes with no known function
  • The automated description pipeline currently uses several types of data to write descriptions:
    • Orthology
    • GO
    • Expression
    • Disease
  • Genes lacking an automated description might constitute a reasonable set to start from, but since phenotype data is not included in the descriptions, there may be false negatives in the gene description set
  • Differential gene expression under a certain condition was generally not thought to be sufficient criteria to say a gene has a function
  • IEA GO annotations are predictors of function, but not sufficient evidence to say a gene definitely does enable that function (or participate in a process)
  • Analysis of the literature could also be informative as it may get at how many genes haven't even been studied. TextpressoCentral searches could help get this data.
    • Even with literature searches, though, we'd need to have some sense of the context in which a gene is mentioned.
  • A matrix approach might be particularly helpful here. Given a set of criteria that might inform function, how many does each C. elegans gene meet?
  • I was directed here from a link on the CGC
    • Do we need to take any action to fix the broken Strain links from CGC to WormBase? (notify CGC or point the old URLs to the current strain URLs ...)
    • This may be an isolated case. KVA contacted the CGC to let them know and ask if they need any help from us.

Website Widgets

  • Any specific issues to discuss here?
  • Widgets Protein page motif_details and overview widget, Go page Associations widget, and Gene page Homology widget

Alliance

?Genotype class

  • Genotype Wiki page updated
  • Have removed strain associations, maternal/paternal genotypes and zygosity from proposal for now