How to represent that phosphorylation of a protein activates it?

[cthoyt]

trafficstars

In BEL, it's common to see chains like:

# protein X phosphorylates protein Y
act(protein(X)) -> protein(Y, pmod(Ph))

# protein Y becomes active when it's phosphorylated (the part I'm actually asking about)
protein(Y, pmod(Ph)) -> act(protein(Y))

# Phosphorylated protein Y (active) phosphorylates protein Z
act(protein(Y)) -> protein(Z, pmod(Ph))

I apologize if this isn't idiomatic, but I hope it gets the idea across. What would be the best relation / set of relationships to express this using RO?

CC: @cmungall @dexterpratt

[cmungall]

Is there a BEL toolkit for interconverting between these?

presumably another way is

act(p(X), ma(GO:"kinase activity")) -> act(protein(Y))

This would have the most natural mapping to an activity-centric GO-CAM view. Here the semantics are on the occurrent class, and a more general directly activates relation could be used

@goodb, is the inference of the modifications something that could fall out of the RHEA axiomatization work?

https://docs.google.com/document/d/10xi10oyISribbssR9CYgz6YVgC_rAVR3W2RvSt7Vdqg/edit

[goodb]

I'm not seeing how the RHEA work would impact this, but might be missing something. Could you give a strawman of what you are thinking might be working here - in terms of the GO MF definitions? The RHEA-based axiomitization of catalytic activities will be useful for classification based on the inputs and outputs of a reaction/function. (See https://github.com/geneontology/go-ontology/issues/14984 ).

On Mon, Oct 1, 2018 at 10:58 PM Chris Mungall [email protected] wrote:

Is there a BEL toolkit for interconverting between these?

presumably another way is

act(p(X), ma(GO:"kinase activity")) -> act(protein(Y))

This would have the most natural mapping to an activity-centric GO-CAM view. Here the semantics are on the occurrent class, and a more general directly activates relation could be used

@goodb https://github.com/goodb, is the inference of the modifications something that could fall out of the RHEA axiomatization work?

https://docs.google.com/document/d/10xi10oyISribbssR9CYgz6YVgC_rAVR3W2RvSt7Vdqg/edit

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/oborel/obo-relations/issues/263#issuecomment-426157954, or mute the thread https://github.com/notifications/unsubscribe-auth/AB_U6lq0j6mYOe6hwScZ5ldZmte_-6_Iks5ugwBxgaJpZM4W8TPl .

[nlharris]

Is this still needed?

[cthoyt]

This is still unresolved. #522 will hopefully renew interest in this