Annotation Conf. Call August 27, 2013

From GO Wiki
Jump to: navigation, search

Agenda

Mailing lists, SF tracker etc

Val asked me (Rama) to give a refresher on where one should post what messages/questions. The following is what I have so far.

  • SourceForge- for ontology term requests, annotation issues (mappings mostly)
  • GO Help desk- when you dont know how to get some data, report something is not working (like AmiGO or the website)
  • Mailing lists- make announcements, ask annotation related questions (which evidence code etc)
  • Seth et al are using github to collect AmiGO2 bugs (https://github.com/kltm/amigo/issues), but if you can't find this URL, feel free to send email to go-help or post a message on go-consortium mailing list

Annotation Extension exercise

Paper 1: PMID 8384702 (Title: MAP kinase-related FUS3 from S. cerevisiae is activated by STE7 in vitro)

File:PMID8384702.pdf

Annotation Extension to be discussed for the MAP kinase activity annotation for FUS3. GO annotations for FUS3- http://www.yeastgenome.org/cgi-bin/GO/goAnnotation.pl?dbid=S000000112

DB (Col 2) GO ID (Col 5) ev.code Reference (Col 6) Extension (Col 16)
FUS3, P16892 GO:4707 (MAP kinase activity) IDA PMID:8384702 has_direct_input: FAR1, requires_regulator:STE7
STE7, P06784 GO:4708 (MAP kinase kinase activity) IDA PMID:8384702 has_direct_input: FUS3, has_input: MF(ALPHA)1 (pheromone)
  • Questions for paper1:
    • Do we capture autophosphorylation?
    • Do we capture targets or substrates if it is an invitro assay
    • FUS3 has to be phosphorylated by STE7 for it to phosphorylate FAR1. Should the relationship be dependent_on or requires_regulator or requires_regulation_by?
    • Also capture the FUS3 phosphorylation of STE7?
    • What IDs should be used for entities in Column 16?
    • What are the corresponding BP annotations?

Paper 2: PMID 10882066 (Title: Cohesin's binding to chromosomes depends on a separate complex consisting of Scc2 and Scc4 proteins.)

File:PMID10882066.pdf

Annotation extension to be discussed for the term establishment of protein localization to chromatin (BP) for SCC2 gene. http://www.yeastgenome.org/cgi-bin/GO/goAnnotation.pl?locus=scc2

DB (Col 2) GO ID (Col 5) ev.code Reference (Col 6) Extension (Col 16)
SCC2, Q04002 GO:71169 IMP PMID:10882066 has_input: GO:0034990, happens_during: GO:6260/DNA replication
  • Questions for Paper 2:
    • If we use a GO complex term in the annotation extension, is there any way to also preserve the information about the specific proteins assayed in the paper, i.e. could we also include the gene product identifiers? For LEGO modeling, what is most useful?
    • Is this a case where a new GO term, e.g. establishment of protein complex localization to chromatin, would be appropriate?
    • Although not to be used for direct annotation, there is also a GO term for S phase (GO:0051320) or S phase of mitotic cell cycle (GO:0000084). Could this term also be used for the happens_during annotation extension? Should there be any relationship in GO between the BP terms S phase and DNA replication?


Discussion

Present: Rama, Paul, Edith, Janos, Petra, Kimberley, Judy, Harol, LI, Karen, Rachael, Jane, Midori, Donghui, Stan(RGD), Ruth


Col-16 exercise

  • When would you add information in col-16?
    • When you want to add/provide specificity to the annotation without having to add a new granular GO term. Another way to think about this is to see if you can say a story with the GO term and col-16 data. Will this be helpful in building LEGO models?
    • We want to capture physiologically relevant substrates. If a mutant form of a protein was used as a substrate then we should look to see why the mutant form was used (easy to make?) and if the authors make a statement about wild type.
  • Paper1:
    • Is had_direct_input the correct relationship to capture FAR1 as the substrate for FUS3?
    • You wan to use had_direct_input only when you are confident it is direct binding. This relationship is okay to use with Invitro enzyme assays.
  • How do we capture the fact that FUS3 has to be phosphorylated by STE7 in order for it to kinase FAR1?
    • Midori- I would capture the phosphorylated form of FUS3 in col-17 using a PRO id.
    • Rama- in that case we are loosing the detail that STE7 is required to modify FUS3 in order for FUS3 to act on FAR1.
    • What is wrong with requires_regulator: STE7 ?
    • What is the difference between requires regulator and requires regulation by.
    • No clear answers for these two questions.
    • For STE7, we could capture the following in col-16: had_direct_input FUS3, has_output: FUS3P(phosphorylated form of FUS3, with PRO ID)
    • has_input: pheromone wouldn't be right for the pheromone inducing part. It has to be a new relationship like in_the_presence_of.
  • How to capture auto-phosphorylation?

There is a specific GO term for autophosphorylation. Use that term, no need for col-16 in that case because you have captured the 'auto' part in the GO Term itself.

  • What IDs are allowed in col-16?
    • One should be aware that the goal of col-16 is to capture specificity, so it makes sense to capture the MODIDs or UniProt or PRO ids which can be tracked easily.
  • When to use has_regulation_target?
    • Transcription terms would be the most suited for this relationship. We should pick transcription related paper to see how this be captured.
    • happens_during: We should hold off on using this relationship. The ontology editors are working on overhauling cell cycle terms and are using happens_during relationship. We want to make sure we use this relationship consistently in the ontology and in col-16.
  • There will be a tutorial at the BarHarbor meeting. But it will be nice to collect some more questions before the BarHarbor meeting. We need a volunteer to select papers for the September consistency exercise.