Question MAP kinase kinase activity and protein tyrosine kinase activity

[ValWood]

I'm picking up a mapping

GO:0004713 | protein tyrosine kinase activity for MAP kinase kinase activity GO:0004708

defined Catalysis of the concomitant phosphorylation of threonine (T) and tyrosine (Y) residues in a Thr-Glu-Tyr (TEY) thiolester sequence in a MAP kinase (MAPK) substrate.

so it isn't in correct, but it isn't a parent of MAP kinase activity.

should the parent be added? Or should the mapping be fixed to protein serine/threonine/tyrosine kinase activity

[pgaudet]

Hi Val,

Where is this mapping coming from?

[ValWood]

sorry it was https://www.pombase.org/gene/SPBC106.01 from Rhea, https://www.rhea-db.org/rhea/10596 will transfer to the curation tracker

[ValWood]

How do I tag Rhea?

[ValWood]

Actually @pgaudet i'm still confused.

The "MAP kinase kinase: def implies it is tyrosine/threonine/serine.

So, in these cases should we not make annotations to "GO:0004713 protein tyrosine kinase activity

I'm not really sure how this works.] I could annotate to MAP kinase kinase protein tyrosine kinase activity protein serine/threonine/tyrosine kinase activity protein serine/threonine kinase activity

[pgaudet]

The GO term for the dual specificity kinase is protein serine/threonine/tyrosine kinase activity

You're right that in GO we would not annotate to "GO:0004713 protein tyrosine kinase activity

[pgaudet]

In fact we maintain GO-> RHEA mappings !! (sorry I was also confused for a bit).

You can change them in the ontology.

[ValWood]

Will transfer back to the GO tracker

[pgaudet]

@ValWood I dont see this mapping; I'll close for now, please reopen if you find the source of the incorrect mapping.

[ValWood]

I still see it here https://www.uniprot.org/uniprotkb/P10506/entry Rhea

[ValWood]

and here https://www.uniprot.org/uniprotkb/Q9Y884/entry from a SPKW

[ValWood]

It seems that MAP kinase kinase do phophorylate tyrosine, and this now has the correct parent, do this was fixed. All good.

[pgaudet]

I guess the KW is https://www.uniprot.org/keywords/KW-0723? @Antonialock Can you please check?

[ValWood]

Ah yes, mappings should be to protein serine/threonine/tyrosine kinase activity not annotate to "GO:0004713 protein tyrosine kinase activity sorry ignore my previous comments

[Antonialock]

it's a rhea mapping, not a keyword

[Antonialock]

e.g. https://www.rhea-db.org/rhea/10596

Ir emoved myself and the label but I don't know who to assign

[ValWood]

There's a KW too,

[Antonialock]

hmmm in uniprot, anything that phoshphorylates "ser/thr" are mapped to "protein serine/threonine kinase" and anything that phosphorylates "tyrosine" is mapped to protein tyrosine kinase activity.

I would need to move the mapping to "protein kinase activity" so it would have a very large effect

• 1,022,666 entries would lose the specificity to protein ser/thr kinase
• 148,515 entries would lose the specificity to protein tyr kinase

I understand why you want to keep them separate because they are different enzymes, but is it correct ontologically?

[ValWood]

I'm not sure why we have a term for "protein serine kinase activity", anything which phosphorylates serine also phosphorylates threonine ?

Isn't it possible to remap the dual specificity kinases to "protein serine/threonine/tyrosine kinase activity" instead. That would be the most accurate.

I agree it's a bit weird, but it does capture specificity correctly. In the mean time , I'm going to block the mapping at PomBase. we'll need to wait for @pgaudet to return for any different solution.

[Antonialock]

No I cannot apply EC's to the keywords. Anything that phosphorylates a serine/threonine gets that keyword, and anything that does tyrosine is assigned that keyword. MAPKK gets both since it does both. It feels above my payfrade to take the decision to remap all those entries......it doesn't feel technically correct, but I can see why you want the distinction too for MFs.

I think serine kinase exists because of cases like this: https://www.uniprot.org/uniprotkb/Q9Y2I7/entry it looks like it does phospholipids + serine residues ....so this should get a different MF too for dual specificity phospholipid and protein serine kinase?

[ValWood]

I guess we could ignore this one for now in the big scheme of things. I filtered it anyway...

[pgaudet]

I'm not sure why we have a term for "protein serine kinase activity", anything which phosphorylates serine also phosphorylates threonine ?

@ValWood you can check the term in Protege, I do try to add the ticket explaining why a term is there - in this case, please see https://github.com/geneontology/go-ontology/issues/20114 - specifically here: https://github.com/geneontology/go-ontology/issues/20114#issuecomment-909198642

It seems there are serine-specific kinases. This is also in the term comment for protein serine kinase activity

[pgaudet]

WRT problematic RHEA mappings (https://www.uniprot.org/uniprotkb/P10506/entry):

RHEA splits each activity WRT substrate, ie a Ser/Thr kinase is annotated to 2 RHEA terms, one for Ser, one for Thr, since they are reaction-centric and in principle these are 2 reactions.

In GO we represent the activity of a protein, and we say it acts on both. We should not have EXACT mappings to RHEA.

[pgaudet]

@Antonialock MAPKK should be mapped to dual specificity kinase activity: protein serine/threonine/tyrosine kinase activity, shouldn't it?