search

geneontology / go-annotation

This repository hosts the tracker for issues pertaining to GO annotations.
BSD 3-Clause "New" or "Revised" License
31 stars 10 forks source link

cell component annotation for transmembrane proteins #2929

Open amandamackie opened 4 years ago

amandamackie commented 4 years ago

The histidine kinases of E. coli two-component systems are inner membrane proteins. They often contain a large periplasmic domain for sensing an external signal and they always contain cytoplasmic domains for interacting with their cognate response regulators which are usually transcription factors.

Given that GO defines 'cellular component' as the 'location occupied by a gene product when it carries out a molecular function', should I annotate these proteins with cell component terms for all three locations ie GO:0005829 cytosol, GO:0005887 integral component of plasma membrane and GO:0030288 outer-membrane bounded periplasmic space.

My apologies if this is the incorrect forum; I checked the GO wiki but couldn't find anything relating to this Thanks Amanda (EcoCyc project)

vanaukenk commented 4 years ago

Hi @amandamackie

Thanks for your question; this is actually one that we've discussed a lot over the years.

For annotating these types of transmembrane proteins, we recommend that curators annotate to the appropriately specific child of 'integral component of membrane': http://amigo.geneontology.org/amigo/term/GO:0016021

The current documentation we have on the GO wiki is here, but we can certainly add to that if you think we need to be more explicit: http://wiki.geneontology.org/index.php/Transmembrane_Proteins

Note also that we'd really like to do some work on the ontology, relations, and annotations for capturing this type of spatial information, but for now, a single annotation to an 'integral component of membrane' child term would be acceptable.

@ukemi @pgaudet

amandamackie commented 4 years ago

Thanks for the clarification.

On the GO wiki, I think it would be helpful to specifically mention that some integral membrane proteins will have functional domains extending into adjoining compartments but that GO recommends annotating these with a single 'integral component of membrane' term (or appropriate child term)

Amanda

ValWood commented 10 months ago

@pgaudet is this still in date?

pgaudet commented 10 months ago

There is no documentation on the wiki, let's keep that one open.