Add several post-translational modification relations under RO:0002436

[cthoyt]

As an addition to RO:0002447 (phosphorylates) and RO:0002480 (ubiquitinates), that occur under RO:0002436 (molecularly interacts with), I would suggest adding relationships for the following post-translational modifications, which all have corresponding functions with the Biological Expression Language (BEL) as well as corresponding Gene Ontology terms in the molecular function namespace:

BEL Name	GO	GO Label
chap	GO:0044183	protein binding involved in protein folding
gtp	GO:0005525	GTP binding
kin (already present)	GO:0016301	kinase activity
pep	GO:0008233	peptidase activity
phos	GO:0016791	phosphatase activity
ribo	GO:0003956	NAD(P)+-protein-arginine ADP-ribosyltransferase activity
tscript	GO:0001071	nucleic acid binding transcription factor activity
tport	GO:0005215	transporter activity
gef	GO:0005085	guanyl-nucleotide exchange factor activity
gap	GO:0032794	GTPase activating protein binding

There are also several other entries in cellular protein modification process (e.g., amination, nitration, polyamination, nitrosylation, neddylation, sumoylation, etc.) but let's just start by thinking about a couple concrete ones :)

I think many of these make sense following the form proposed by @cmungall (see: https://github.com/oborel/obo-relations/issues/259#issuecomment-424472190) with the addition of mf to represent one of these molecular functions that are a cellular protein modification process:

?a1 enabled_by ?gp1
?gp1 a ?mf
?mf a GO:'cellular protein modification process'
?a1 directly-regulates ?a2
?a2 enabled-by ?gp2
->
?gp1 XXX ?gp2

One caveat: a couple of these aren't "activities" but come from the "binding" tree. I'm also curious how one would handle those.

[nlharris]

Are these still needed?

[cthoyt]

Yes I'm still interested in adding additional relationships (and reverse relationships). I'll update the issue header with some more information

[nlharris]

Is this now done?

[cthoyt]

It appears there are some issues with the build and these new terms are not being propagated. Unfortunately, of the issues/PRs/commits that were a follow-up to #521 are very cryptic:

• 586

• 587

• c2157c9f6a9e885bfe619196bee298066ec77edd

The new termsin the ro-edit.owl don't appear to be in the release files. e.g., a search in https://github.com/oborel/obo-relations/blob/master/ro.obo for myristolates currently shows nothing. Does this also means there's an overall problem with the build?

[matentzn]

cc @anitacaron

[anitacaron]

I'll look into the ODK migration branch I've been working on if the issue is still there.

[anitacaron]

@cthoyt I checked the IDs created in PR #586 and I could found in the last RO release. Try to search for myristoylates.

[cthoyt]

Yes, but that one was done by hand - I mean the remaining ones that we curated in https://github.com/oborel/obo-relations/blob/master/src/patterns/ro-interaction-relations-src.yaml to be generated via template are not getting propogated

[anitacaron]

Ok, I see it's a LinkML template, which we don't have in ODK. I could move the targets in the src/patterns/Makefile to ro.Makefile and generate the component. What do you think, @matentzn?

[cthoyt]

In case you missed it, the components are already generated in https://github.com/oborel/obo-relations/blob/master/src/patterns/ro-interaction-relations.ofn and https://github.com/oborel/obo-relations/blob/master/src/patterns/ro-interaction-relations.owl as an example

[anitacaron]

Yes, they need to be moved to the components folder properly.