Open cmungall opened 4 years ago
If I remember correctly, the short answer is I don't think GEFs are related to the GTPase activity. They swap out GDP for GTP and then Ras is in the GTP-bound form. This is its active state. Then GAPs stimulate the hydrolysis of the GTP to inactivate Ras. On its own, Ras has GTPase activity, but it is minimal and physiologically a GEF does the work of activating Ras.
I agree that the comment is misleading!
There are certainly biologists who see GEFs as regulating the activity of their cognate RAS-family GTPases, but in the world of GO, what the GEFs are doing is enabling the RAS-family proteins to bind substrate so they are closer to the acting-upstream-to-provide-input bucket.
I definitely think the GEFs are in the pathway. It's the GAPs and the actual GTPase activity I question. Are those part of the pathway or do they negatively regulate it? Note that the GTPase gene product would be included as an enabler in the pathway by virtue of its 'doing something' in the GTP-bound state. But like the other cases of regulation via small molecule binding, is the Ras-GTP binding the causal activity that is upstream of the next step in the signaling? The hydrolysis of the GTP stops the pathway. @huaiyumi, any thoughts on this?
When a biologist says that a GEF regulates a RAS-GTPase, are they talking about the molecule or the GTPase activity of the molecule? They are regulating the signaling function I think.
Not necessarily making the distinction, I suspect - but GO needs to.
Are they "GDP-binding inhibitors" I often need to describe the MF of a binding inhibitor, and can't...
What should the relation be ?
I propose the following:
[x] Change definition to be more explicit : old: "Stimulates the exchange of guanyl nucleotides associated with a GTPase. Under normal cellular physiological conditions, the concentration of GTP is higher than that of GDP, favoring the replacement of GDP by GTP in association with the GTPase." proposed: (edited) Stimulates the exchange of GDP to GTP on a signaling GTPase, changing its conformation to its active form. Guanine nucleotide exchange factors (GEFs) act by stimulating the release of guanosine diphosphate (GDP) to allow binding of guanosine triphosphate (GTP), which is more abundant in the cell under normal cellular physiological conditions.
[x] remove the comment "Note that guanyl-nucleotide exchange factor (GEF) activity alters binding affinity, but not GTPase catalytic activity directly. Therefore GO:0005085 is under 'regulation of GTPase activity' but is not a 'GTPase regulator', which is reserved for cases where the action of one gene product directly alters the rate of catalysis by another gene product." because as @cmungall points out, this is not true (the LD starts with is_a 'GTPase regulator activity'.
[x] change is_a 'GTPase regulator activity' to 'GTPase activator activity'. This seems direct enough.
Thanks, Pascale
But the GEF doesn't activate the GTPase activity does it? It will make the GTPase a 'positive regulator' but the antagonistic GAP activates the GTPase . We are conflating entity model and active it model here.
I notice the comments above allude to this...
I think I see what you mean.
For ATPases we have split the 'ATPase protein' versus the ATP hydrolysis activity, see #20876 We will discuss this today on the annotation call, it'd be great if you could join.
We were thinking the same probably should be done for GTPases. The idea is that an ATPase is a protein that has_part ATP hydrolysis activity. So, here if we state that the GEF makes the GTPase in its active form, we can say it activates the GTPase 'protein', while the GAP activates the GTP hydrolysis activity.
Does that make sense ?
Note that GO:0005085 JSON guanyl-nucleotide exchange factor activity has a comment Comments Note that guanyl-nucleotide exchange factor (GEF) activity alters binding affinity, but not GTPase catalytic activity directly. Therefore GO:0005085 is under 'regulation of GTPase activity' but is not a 'GTPase regulator', which is reserved for cases where the action of one gene product directly alters the rate of catalysis by another gene product.
which is no longer true .
In the past it used to be directly under 'molecular function'
but if the GEF is antagonistic to the GAP should the just be annotated a "signal transducer activator activity"
and the GAP a "signal transducer inactivator activity" and a GTPase activator activity
These don't seem to be 'coupled' in the same way as an ATPase that is providing the energy for a physical change (movement). These are separable steps?
remove the comment "Note that guanyl-nucleotide exchange factor (GEF) activity alters binding affinity, but not GTPase catalytic activity directly. Therefore GO:0005085 is under 'regulation of GTPase activity' but is not a 'GTPase regulator', which is reserved for cases where the action of one gene product directly alters the rate of catalysis by another gene product." because as @cmungall points out, this is not true (the LD starts with is_a 'GTPase regulator activity'.
If (but note: big if) I recall correctly, the comment is biochemically accurate. Therefore, if it's not "correct" in that it doesn't describe the actual ontology structure, it's the ontology that needs to change. (I also think the ontology structure was consistent with the comment at some point but that memory is getting fuzzy.)
I'm not going to wade into entity model vs. activity model here. But I'm going to stake out some old-school enzymology turf and say that if it were up to me, I would use "catalytic activity regulator" terms only for things that alter catalytic activity directly.
We were thinking the same probably should be done for GTPases. The idea is that an ATPase is a protein that has_part ATP hydrolysis activity. So, here if we state that the GEF makes the GTPase in its active form, we can say it activates the GTPase 'protein', while the GAP activates the GTP hydrolysis activity.
Sorry, no, to me this borders on gibberish. But if it's just me, and you have evidence that biologist users will be happy with it, carry on.
The comment is misleading as in fact there is NO relationship to
regulation of GTPase activity
What in fact is the relationship? This should be encoded in the ontology. Negative information can also be encoded using disjointness axioms, rather than just as text, allowing for automated checks