search

geneontology / go-ontology

Source ontology files for the Gene Ontology
http://geneontology.org/page/download-ontology
Creative Commons Attribution 4.0 International
220 stars 40 forks source link

query about "protein autophosphorylation" term #11882

Closed gocentral closed 8 years ago

gocentral commented 9 years ago

Isn't this term a bit strange as a separate process? The only thing which differentiates it from "protein phosphorylation" is that the substrate is itself. This doesn't make it a different process though.

Ditto any similar terms.

Reported by: ValWood

Original Ticket: geneontology/ontology-requests/11720

gocentral commented 9 years ago

I agree. Col 16 materials.

Original comment by: pgaudet

gocentral commented 9 years ago

From the perspective of mechanism, you have a point. From the perspective of biologists trying to model pathways, knowing if a protein autophosphrylates can be very significant. While we could effectively represent this as an annotation extension, I worry that this would hide/obscure it from most users.

Original comment by: dosumis

gocentral commented 9 years ago

Diff:


--- old
+++ new
@@ -1,4 +1,3 @@
-

 Isn't this term a bit strange as a separate process? The only thing which differentiates it from "protein phosphorylation" is that the substrate is itself. This doesn't make it a different process though.

Original comment by: dosumis

gocentral commented 9 years ago

Thats a good point, but I would advise any modeller to use all of the molecular function and their extensions, which they would need anyway for the rest of the model.

That way they get everything in one go and they don't need to worry if its autophosphorylation or not.

In fact we already do this: http://www.esyn.org//builder.php?type=Graph&term=GO:0007346&interactionType=physical&source=pombase&includeInteractors=false

The 'autophosphorylations' here are the self/self loops.

Personally, I think locating a GO term for autophosphorylation as a special case is more obscure than a general "use the genes+ extensions" to model pathways guideline.

Original comment by: ValWood

gocentral commented 9 years ago

IntAct uses GO BP terms as xrefs on enzymatic reactions but we can't use C16. I think we export this data (or at least it's searchable within IntAct) and as DOS says, wet lab biologists will look for the difference between any old phosphorylation and autophosphorylation.

Food for thought...

Birgit

Original comment by: bmeldal

gocentral commented 9 years ago

Diff:


--- old
+++ new
@@ -1,4 +1,3 @@
-
 Isn't this term a bit strange as a separate process? The only thing which differentiates it from "protein phosphorylation" is that the substrate is itself. This doesn't make it a different process though.

 Ditto any similar terms.

Original comment by: ValWood

dosumis commented 8 years ago

Keeping auto-phosphorylation. No harm in annotators recording this in an extension anyway. Also worth noting that the extension can't distinguish cis from trans autophosphorylation. We have both terms in the ontology. I think that the ability to cis-autophosphrylate IS a property of a protein and so definitely belongs in the ontology. The term currently has no annotations, but that appears to be due to neglect rather then biology:

e.g. see

http://stke.sciencemag.org/content/2/54/pe4 http://www.ncbi.nlm.nih.gov/pubmed/9054577 http://www.ncbi.nlm.nih.gov/pubmed/23000554 http://www.ncbi.nlm.nih.gov/pubmed/9201908