Annotation Conf. Call, January 14, 2014

From GO Wiki
Jump to: navigation, search

Agenda

Attendees:
SGD: Rama, Stacia, Diane, Edith
RGD: Tom, Shur-Jen
MGI: Judy, Li
EBI: Alex, Paola Becky
Pombe: Midori
WB: Kimberley
Dicty: Pascale, Petra
Reactome: Peter
Berkeley: Chris, Heiko, Moni, Suzi
Panther: Huaiyu

Jenkins Report

Are groups checking the Jenkins report for their GAF? Any questions?

Guidelines for Enzyme regulator activity

Should 'enzyme regulator activity ; GO:0030234' be limited to gene products that directly regulate (e.g. bind to) an enzyme. Should the definition be modified and the term also be a child of enzyme binding?

  • How much do curators need to know about the mechanism of regulation to use the MF term?
  • Do we require experimental evidence for a physical interaction to use the 'enzyme regulator activity' terms?
  • If binding is demonstrated for an ortholog, but inhibition for the protein of interest, is that enough to make the 'enzyme regulator activity' annotation?
    • Example: C. elegans MAI-2 shows inhibitory activity for the mitochondrial F0F1-ATPase, but direct binding between the inhibitor and catalytic subunit of the enzyme has been demonstrated in yeast.
  • Are there other enzyme regulator activities that don't involve direct binding to the enzyme?
  • If mechanism is not known, should curators only use the BP term?

Histone kinase terms (mentioning specific residues in term names)

Continued from last call: There are some histone kinase terms that include the name of the residue that is being phosphorylated in the term name itself (http://amigo.geneontology.org/cgi-bin/amigo/term_details?term=GO:0035174#lineage). Although histones are very well conserved, there are still some differences in the sequence and the serine is always not in the same position in all organisms, but often the context is the same (i.e it is +/- one or two residues).

  • It is not clear if curators should request granular term with specific residues in the term name.
  • how do we capture functional equivalence in GO term name?

UniProt-GOA species-specific files

We plan to stop submitting the GOA species-specific GAFs if the species is one owned by another group, that is the following files; http://viewvc.geneontology.org/viewvc/GO-SVN/trunk/gene-associations/

  • gene_association.goa_arabidopsis
  • gene_association.goa_mouse
  • gene_association.goa_rat
  • gene_association.goa_zebrafish

Users will still be able to get these files from the UniProt-GOA ftp site, or from the UniProt multispecies file.

We will announce this in our next release (10th December) and probably stop providing them early next year (end of January).

However, if you are one of the groups above, could you ensure that you are picking up annotations from other groups for your species so that you have a full set of annotations (remembering to consider the annotation extension column, your database may have the same annotation but without an annotation extension that another group has provided).

If your group is using Protein2GO then you will be getting the full set of annotations back from us anyway.

Minutes

  • Jenkins: Rule AR-13 is listed as "proposed" on the documentation page. It should be changed to 'Implemented'. Rama will fix it.
  • Enzyme Regulator activity:
    • making this term child of enzyme binding and modifying the definition is fine
    • while it is true that enzyme binding and regulation may not be shown in one assay, we agreed that we will use IC to infer this annotation from 2 different experiments from same paper or 2 different papers: a binding (IPI) assay where the enzyme and regulator are shown to interact, and an IMP or IDA experiment where the regulation is shown which would be a BP annotation. The scenario with 2 different papers will need a new GO_REF for IC. This is similar to our practice for annotating Txn factors.
    • groups should take time to review their annotations to comply with these guidelines. Detailed guidelines will be discussed at a future Annotation call


  • Histone Kinase terms
    • use the general histone serine kinase term, request a PRO term for the phosphorylated form of the protein and put that data in col-16 with has_output relationship
    • We should still consider keeping the specific terms for a subset of species because it is useful (ontology group has to decide where to draw the line)
  • GOA GAF files
    • GOA will not be supplying GAFs for species that already are owned by specific databases (TAIR, Zfin etc as shown above)
    • In addition, GOA (GOC) has started supplying the gene association files that are built from the one protein per gene set for human, dog, pig, cow and chicken (reference proteome set):

ftp://ftp.ebi.ac.uk/pub/databases/GO/goa/HUMAN/
http://viewvc.geneontology.org/viewvc/GO-SVN/trunk/gene-associations/

    • GOA will supply regular GAFs for these species as well