search

geneontology / go-ontology

Source ontology files for the Gene Ontology
http://geneontology.org/page/download-ontology
Creative Commons Attribution 4.0 International
223 stars 40 forks source link

NTR: non-membrane-bounded organelle formation #21939

Closed pgaudet closed 3 years ago

pgaudet commented 3 years ago

Please provide as much information as you can:

LD 'cellular component assembly' and (results_in_assembly_of some non-membrane-bounded organelle)

pgaudet commented 3 years ago

+[Term] +id: GO:0140694 +name: non-membrane-bounded organelle assembly +def: "The aggregation, arrangement and bonding together of a set of components to form a non-membrane-bounded organelle." [PMID:28225081] +synonym: "non-membrane-bounded organelle assembly" EXACT [] +synonym: "non-membrane-bounded organelle formation" EXACT [] +synonym: "non-membrane-enclosed organelle assembly" EXACT [] +synonym: "non-membrane-enclosed organelle formation" EXACT [] +intersection_of: GO:0022607 ! cellular component assembly +intersection_of: results_in_assembly_of GO:0043228 ! non-membrane-bounded organelle +property_value: term_tracker_item https://github.com/geneontology/go-ontology/issues/21939 xsd:anyURI +created_by: pg +creation_date: 2021-08-09T15:30:06Z

ValWood commented 3 years ago

Children terms (if applicable) Should any existing terms that should be moved underneath this new proposed term? (inferred):

No 'bacterial-type flagellum assembly' 'centriole assembly' 'kinetochore assembly' 'myofibril assembly' 'ribosome assembly' 'septin ring assembly' 'spindle assembly'

Yes 'P granule assembly' 'P-body assembly' 'proteasome storage granule assembly' 'stress granule assembly'

Don't know 'cellulosome assembly' 'lipid droplet formation' 'podosome assembly' 'terminal web assembly'

As an aside, I worry a little about these "formation terms" being used for phenotypes for any member of a condensate/stress granule etc which if not present the granule does not form correctly. This is a bit like annotating a mutated subunit of the proteasome to proteasome assembly. IMHO to use such terms something should be either a) a chaperone (not part of the final complex/condensate) or be shown to be a substrate in some regulation (i.e. modification dependent).

pgaudet commented 3 years ago

As an aside, I worry a little about these "formation terms" being used for phenotypes for any member of a condensate/stress granule etc which if not present the granule does not form correctly.

Right, we discussed this- this is a general problem for all assembly terms. In this case there is a MF, molecular condensate scaffold activity, that serves to assemble to condensate, at least in some cases. The review that @esalladini provided explains this quite nicely PMID:28225081

pgaudet commented 3 years ago

WRT the terms you say should not be children: No 'bacterial-type flagellum assembly' 'centriole assembly' 'kinetochore assembly' 'myofibril assembly' 'ribosome assembly' 'septin ring assembly' 'spindle assembly'

they classify automatically because they themselves are non-membrane-bounded organelles. Why do you think they are not relevant there?

ValWood commented 3 years ago

I don't think they are "condensates" at least flagellum, ribosome, and spindle have a defined stoichiometry, or defined structure. I'm less sure about kinetochore but I don't think it is a condensate- but perhaps it is- I will ask.

We already said "condensate" is narrower than "are non-membrane-bounded organelles", so maybe we really do need to define and add the term.

ValWood commented 3 years ago

The flagellum, kinetochore, ribosome, spindle have a very defined, regulated, step-wise assembly pathway.

I thought that "condensate" formation was more of an intrinsic property of the proteins. Although I guess this is "fuzzy". I still think we need to be wary of calling everything that is not membrane-bound a condensate.

ValWood commented 3 years ago

Right, we discussed this- this is a general problem for all assembly terms. In this case there is a MF, molecular condensate scaffold activity, that serves to assemble to condensate, at least in some cases. The review that @esalladini provided explains this quite nicely PMID:28225081

Right. But I would not annotate a scaffold that was part of a complex to a BP "protein complex assembly" if the complex is self-organizing. If activity needs to happen (i.e. a modification) to make the scaffold function as a scaffold then I think that it is part of an 'active process of assembly'. Otherwise, it is only a scaffold involved in a signalling pathway (or some other pathway or process). ~Also we requested scaffold terms before but these were prevented. We had to use molecular adaptor (even though I think they are slightly different things, and some proteins like MAP kianse scaffolds are both scaffolds and adaptors). I'll add these comments to the scaffold ticket.~ I see this is under molecular adaptor. I'm not convinced that scaffolds and adaptors are the same but let's level that one for now.

ValWood commented 3 years ago

Maybe I am over-thinking this. It's just that most of the annotation I look at for these "assembly" processes the evidence is poor. In some cases the expression level drives condensate formation. However in others there is some active step-wise process. That seems to be what the review says too. It's important only to annotate the active processes (and that it aligns with what we mean by a GO BP). I guess in most cases I don't see a difference in the practice that has been discussed for protein complex assembly.

For example we don't annotate MAP-kinase scaffold activity The binding activity of a molecule that functions as a physical support for the assembly of a multiprotein mitogen-activated protein kinase (MAPK) complex. Binds multiple kinases of the MAPKKK cascade, and also upstream signaling proteins, permitting those molecules to function in a coordinated way. Bringing together multiple enzymes and their substrates enables the signal to be transduced quickly and efficiently.

to a term "MAPK complex assembly" (there is no such term), but we annotate it to the signalling process. This was decided back at the signalling meeting in 2019, but maybe we want to do it some other way? Whatever, we should be consistent -I find it very confusing that assembly of condensates and complexes would be treated differently. I'm also a bit disturbed that we pull in so many inferred annotations to "assembly of non membrane-bound complexes" to gene products for which there seems to be no evidence for a direct role in this process. (Note that these concerns predated any of these recent new term request in this area, I just didn't get around to reporting them yet -these changes actually make things clearer, and I think the scaffold MF term will be useful, it's the process I have a problem with in some cases).

I guess it does no harm to let this roll along for a bit but this seems to be one of those areas where everything which interacts with something within a complex or condensate during its formation would become annotated to its "assembly".

pgaudet commented 3 years ago

In this specific case all I did was to create a grouping term for already existing 'non-membrane bounded organelle assembly' terms.

But I agree evidence is often lacking, and certainly knockouts should not be used as evidence for this process.

ValWood commented 3 years ago

In this specific case all I did was to create a grouping term for already existing 'non-membrane bounded organelle assembly' terms.

I know, I'll leave this for now....

ValWood commented 3 years ago

Based on this paper, I would include kinetochore in the condensate list https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7341897/