Annotation Conf. Call 2015-11-24

From GO Wiki
Jump to navigation Jump to search

Meeting URL:


Annotation Consistency Exercise

Annotation consistency paper picked by TAIR: The MYB36 transcription factor orchestrates Casparian strip formation.

Choice of paper-motivation

Capturing downstream targets and their resulting effects, importance of reading term definitions

Annotations requested to be restricted to Figures 1-4 only, to keep discussion manageable.

Summary of the paper

MYB36 is a transcription factor that controls where the machinery for deposition/biosynthesis of lignin is expressed The location of the machinery determines where the Casparian strip can form.

Figure 1.

 leaf ionome mutant vs. wt
 root hair phenotype mutant vs. wt
 endodermal localization of MYB36 expression

Figure 2.

 Loss of Casparian strip and disruption of the apoplastic barrier in myb36 mutants.

Figure 3.

 Identifying target genes of MYB36 - qPCR
 MYB36 binds specific target gene promoters - ChIP

Figure 4.

 CASP1 localization
 Ectopic Casparian strip formation with induced ectopic expression of MYB36


Annotations received from five groups, TAIR included. (MGI, Pombase, WormBase, UCL)

Discussion points

Lots of new term requests. CASP1 and MYB36 annotated.

Molecular function

  • transcription factor activity/sequence specific DNA binding
  • review what to do with ChIP assays/chromatin binding
  • review the relationship types for AE (has_input promoter, regulates_transcription_of, has_direct_input)
  • IPI ok?

Cellular component

  • nucleus?

Biological process

  • secondary phenotypes? ionome and root hair differences - annotate?
  • regulation of transcription
  • regulation of protein localization?
  • regulation of lignin/suberin biosynthesis
  • (regulation of) Casparian strip formation


In attendance

  • Aleks, Alex, Chris, David H., Donghui, Edith, Elena, Kimberly, Leonore, Melanie, Midori, Pascale, Rachael, Robert, Sabrina, Stacia, Stan, Tanya, Val

Annotation Issues

Molecular Function - ChIP Experiments

  • Annotating ChIP experiments - still some discrepancy in how the results of these experiments are annotated
  • There were a number of annotations to some flavor of DNA binding term, but we seem to agree that ChIP experiments don't allow for that level of resolution, i.e. what, exactly, does the target of the IP bind?
  • Previously published guidelines for annotating transcription factors with GO terms have suggested not relying on in vivo experiments, like ChIP, to annotate to a DNA binding term:
    • "We have chosen not to rely on assays measuring in vivo TF–DNA interaction (e.g. the Chromatin ImmunoPrecipitation assay) because it is not possible to ascertain in these assays that the TF in question actually binds directly to DNA, or whether some other component in the in vivo system mediates the TF–DNA association."

Cellular Component - ChIP Experiments

  • If ChIP experiments are not suitable for DNA binding annotations, can they be used to make cellular component annotations?
  • Some annotations to GO:0000785 chromatin were made, but the feeling is that just making an annotation to chromatin loses the specific regulatory target information that the experiments hope to address
  • The solution to this is to use the annotation extension 'coincident_with' and make two, comma-separated, annotation extensions: the first to the SO term for promoter SO:0000167 and the second to the ID of the gene(s) identified in the ChIP experiment as a target of the interaction
 Example: GO:0000790 nuclear chromatin, IDA, coincident_with(geneID for MYB36),coincident_with(SO:0000167, promoter)).
 Do we have consensus here?  If so, we need to document it so it's easily found.  QuickGO has ability to link to documentation.
  • IPI for MF not encouraged, great for BP

Cellular Component - Annotations to Data in Figures

  • There was some discussion of annotating to data displayed in Figures, but not explicitly documented in text
  • For example, an image in a Figure may show fluorescence in a cell nucleus, but the authors don't state that result in the text
  • Some groups make these annotations, others don't
 Should GO have a policy on this?

Biological Process - Phenotypes, Regulation, Upstream and Downstream Processes

  • This paper documented a number of mutant phenotypes that result from mutations in a transcription factor
  • Aside from the mutant phenotypes that illustrated changes in gene expression, what other phenotypes could be used to make informative GO annotations?
  • There was some consensus that annotations to ion homeostasis, regulation of ion transport, and regulation of protein localization were too far removed from the action of the transcription factor to be suitable BP annotations
  • Annotations to terms (existing or suggested) that describe, in some way, a relation to Casparian strip formation seemed to have more support
  • We also discussed what the concept of 'regulation' in GO means - an effect on rate, but also an effect on the activation or beginning of a process?
  • Will LEGO help provide better annotations for papers like this?
  • This paper will be used to create a LEGO model at the upcoming LEGO training camp; we'll re-visit the paper in January 2016 to see how the LEGO model and the more traditional GO annotations compare

Upcoming Annotation Conference Calls