May 10th 2010

From GO Wiki
Jump to: navigation, search

MINUTES

Present: Rebecca Foulger, Alex Diehl, Peter D'Eustachio, Sandra Orchard, Ruth Lovering


LINKING THE UPSTREAM CELL SURFACE RECEPTORS TO THEIR DOWNSTREAM SIGNALLERS

It was agreed that Becky should create the following terms:

  • intracellular signal transduction ; GO:NEW
    • is_a parent: signal transduction ; GO:0007165
    • is_a parent: intracellular signaling pathway ; GO:0023034
  • intracellular signal transduction involved in cell surface receptor linked signaling ; GO:
    • is_a parent: intracellular signal transduction ; GO:NEW
    • part_of parent: cell surface receptor linked signaling pathway ; GO:0007166

then as and when we need them, we can create children.

  • E.g. MAPKKK cascade involved in EGFR mediated signaling pathway ; GO:NEW
  • E.g. PI3K cascade involved in insulin receptor signaling pathway ; GO:NEW

The following terms will be given the additional parent: intracellular signal transduction ; GO:NEW

  • hippo signaling pathway ; GO:0035329
  • intracellular protein kinase cascade ; GO:0007243
  • phosphoinositide 3-kinase cascade ;GO:0014065
  • small GTPase mediated signal transduction ; GO:0007264
  • second-messenger-mediated signaling ; GO:0019932
  • TOR signaling pathway ; GO:0031929

let Becky know if there should be others.

Sandra requested that to make it clearer which are whole pathways (receptor-ligand binding to effect: eg steroid hormone receptor signaling pathway) and which are bits of pathways (eg MAPKKK cascade), we should rename all the 'cassettes' or 'bits of pathways' from pathway to cascade.

It was also noted that several of the 'pathway' terms are probably not entirely intracellular. Alex and Peter pointed out that it depends how you define intracellular.. is it restricted to the events within one cell only??

AI: Becky will move carbohydrate mediated signaling ; GO:0009756 to be a direct child of 'signaling pathway ; GO:0023033'.

When there's time, Becky will go through and see if any of the other terms should also be moved.


STEROID HORMONE RECEPTOR VS STEROID HORMONE SIGNALING

  • After quite a bit of discussion about redundancy, and whether the receptor or the ligand terms were most useful, it was decided that both steroid hormone receptor signaling pathway ; GO:0030518, and steroid hormone mediated signaling pathway ; GO:0043401 terms are probably needed, depending on the assay you use for the experiment.

Current annotations are:

  • steroid hormone receptor signaling pathway ; GO:0030518
    • 212 gene products, 34 direct associations
  • steroid hormone mediated signaling pathway ; GO:0043401
    • 36 gene products, 0 direct associations (mainly plants)

It seems like the two terms should be linked though. At the moment, they just share the receptor activity function children.

AI: Make steroid hormone mediated signaling pathway ; GO:0043401 a parent of steroid hormone receptor signaling pathway ; GO:0030518 for now. We might decide that this creates too many TPVs in the ontology, but it's the logical thing to do for now. Also need to add a comment to GO:0043401 to say that the non-receptor terms are for use in annotation where the method of action is not known.


RECEPTORS THAT BIND TO MORE THAN ONE LIGAND & LIGANDS THAT BIND TO MORE THAN ONE RECEPTOR

This is biology getting in the way of our ontology development. Ruth pointed out that we need to make sure that our receptor terms and signaling pathway terms describe a function/event and NOT a gene product.

At the moment, there's quite a bit of inconsistency in the definitions. The definition of a receptor-mediated pathway should be the series of events as a result of a receptor binding one of it's physiological ligands. The definition of a ligand-mediated pathway should be the series of events as a result of a ligand binding one of its physiological receptors. Becky will correct these as she comes across them.

There isn't a clear way in the ontology to deal with the problem, and it will have to be a case of using mutliple annotations for now.

We need to be careful when linking FUNCTION terms, and 'RESPONSE TO' terms, that we're not creating TPVs.




AGENDA

RECEPTORS THAT BIND TO MORE THAN ONE LIGAND & LIGANDS THAT BIND TO MORE THAN ONE RECEPTOR

  • How do we deal with these?
  • Need to link up signaling pathways with 'response to x' terms. This will affect this.

SIGNAL TRANSDUCTION

  • Do we want any of the signaling pathway terms to have the additional signal transduction parent?

STEROID HORMONE RECEPTOR VS STEROID HORMONE SIGNALING

  • Why are there terms for both steroid hormone receptor signaling pathway ; GO:0030518 and steroid hormone mediated signaling pathway ; GO:0043401? Should they be merged or links made between them?