causal models for binding and binding inhibition

[ValWood]

from https://github.com/pombase/canto/issues/2482#issuecomment-878120963

Note to self: We also need to consider if this will work for protein A directly inhibits [protein B protein-binding protein C] Maybe not, but I am not convinced there are not better ways to capture/infer this. https://www.pombase.org/gene/SPAC1F5.04c might be a good example to model this (but needs new GO terms).

Cut12 shows a different way of modelling this type of binding dependent pathway choice: https://www.pombase.org/gene/SPBC649.05 although this isn't just competitive binding, it is also modification dependent so it's a bit different, (but I suspect modification will nearly always underlie this differential binding). Anyway I think 'regulating' binding is something we might need to figure out.

[ValWood]

Was reading this today: Because Cdk1 phosphorylation of the formin Cdc12 inhibits its interaction with Cdc15 (Willet et al., 2018), we tested whether Cdk1 phosphorylation inhibits Pxl1 binding to Cdc15 similarly. Cdk1-Phosphorylated form of pxl1 does not bind to F-bar domain

I am not curating this as 'cdc2 inhibits binding of Pxl1 to cdc15' (actually it' s a bit more complicated anyway because Pxl1 binds cdc15 in 2 regions and this only prevents binding to one of them)

But I will request the phosphorylated form and un-P form of Pxl1 so that I can model which one is bound to Cdc15 (un-P). The fact that cdc2 will be annotated as adding these residues is enough to infer that cdc2 inhibits the binding between Pxl1 and cdc15.

So we will eventually have something like: Cdc2 phosphorylates (inhibitory) pxl1-P negatively regulates contractile ring assembly Cdc15/UnPhos cytoskeletal adaptor for Pxl1-unP part of contractile ring assembly

Then we can also do away with curating the 'localizations which aren't really independent 'processes' , they are just a consequence of the different modifications (i.e signalling). This is much neater.... it seems to fix the regulation of binding and regulation of localization problems and makes things much more consistent.

We will see this pattern over and over. For example, this will be common for cytokinesis as cdc2 inhibits proteins involved in cytikinesis/mitotic exit during metaphase and then decrease in cdc2 activity and dephosphorylation of the ring components promotes ring assembly We can already see this pattern here: https://www.pombase.org/gene/SPAC20G8.05c cdc12 https://www.pombase.org/gene/SPBC11B10.09 cdc2

[ValWood]

Based on the above this cdc2 annotations seems surprising: cdc2 phosphorylates hcn1 involved in positive regulation of exit from mitosis during mitotic telophase it is the only one that doesn't fit.

It seems to be regulating ace2 which is post ring-contraction so it is probably OK (might even be in G1/next cycle because S-phase is begun by the time cell separation is complete.