search

geneontology / go-annotation

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

UniProtKB-KW:KW-0762 carbohydrate metabolism #1237

Open gocentral opened 9 years ago

gocentral commented 9 years ago

This UniProt Keyword is sometimes applied to gene products which are involved in carbohydrate derivative metabolism, rather than carbohydrate metabolism.

yea4 O74750 NST UDP-N-acetylglucosamine transporter (predicted) rft1 O94302 Man5GlcNac2-PP-Dol translocation protein Rft1 (predicted) pet3 Q10354 phosphoenolpyruvate transporter Pet3 vrg4 Q9UTK8 Golgi GDP-mannose transporter Vrg4 (predicted)

Reported by: ValWood

Original Ticket: geneontology/annotation-issues/1257

ValWood commented 6 years ago

see https://github.com/geneontology/go-annotation/issues/1503

ValWood commented 4 years ago

Still present:

UniProtKB-KW:KW-0762 S. pombe O74750 yea4 carbohydrate transport   incorrect annotation, correct- carbohydrate derivative  
UniProtKB-KW:KW-0762 S. pombe Q10354 pet3 carbohydrate transport   incorrect annotation, correct- carbohydrate derivative  
UniProtKB-KW:KW-0762 S. pombe O94302 rft1 carbohydrate transport   incorrect annotation, correct- carbohydrate derivative  
pgaudet commented 7 months ago

@Antonialock can you have a look at this?

Antonialock commented 7 months ago

hmm I think in this case it's best to remap the keyword 'carbohydrate transport' to organic substance transport

I can't do the same for carbohydrate metabolism as the GO term 'organic substance metabolic process' is "not for direct annotation". The next GO term up is "metabolism". I have deleted the mapping. Do you have a suggestion what can be done here? There are over a million trembl entries with this keyword and I suppose it would be good if they can have at least a high level GO term.

can we have a new grouping term carbohydrates and carbohydrate derivative metabolism CHEBI_78616 ?

Some GO terms have both carbohydrate and carbohydrate-derivative metabolism as parents anyway if that is relevant? e.g GO:0061621 canonical glycolysis GO:0034232 ascospore wall chitin catabolic process GO:0061613 glycolytic process from glycerol

ValWood commented 7 months ago

Why not see if the KW can be removed only from the nucleotide-sugar transporters, rather than from everything?

Antonialock commented 7 months ago

I don't think I could remove the lw sugar transport from carbohydrate-derivative transporters https://www.uniprot.org/uniprotkb?query=%28keyword%3AKW-0762%29+AND+%28reviewed%3Atrue%29

ValWood commented 7 months ago

The problem here is probably more CHEBI than GO. BY some classifications a nucleotide sugar would be a carbohydrate. Maybe don't worry about this. Athough I'm not sure if we are discussing metabolism terms or transport terms?

ValWood commented 7 months ago

wow this ticket is 2014, let's just. close it?

Antonialock commented 7 months ago

well I reviewed both the carbohydrate metabolism and carbohydrate transport GO term mappings.

I remapped KW sugar transport to a less specific term I removed KW sugar metabolism mapping

the removal of the mapping means we have lost 1 million annotations to trembl entries.....it might be good to put something back? can we make a parent for both?

ValWood commented 7 months ago

I think a grouping term for carbohydrate/carbohydrate derivative would be quite useful, because it is often difficult to specify. I would support this, and because it is in CHEBI it should be OK. @pgaudet can we add this parent ?

pgaudet commented 7 months ago

Sure - but I think we'll need to add them in all branches where we have both 'carbohydrate' and 'carbohydrate-derivative': 'carbohydrate binding' http://purl.obolibrary.org/obo/GO_0030246 'carbohydrate biosynthetic process' http://purl.obolibrary.org/obo/GO_0016051 'carbohydrate catabolic process' http://purl.obolibrary.org/obo/GO_0016052 'carbohydrate derivative transmembrane transporter activity' http://purl.obolibrary.org/obo/GO_1901505 'carbohydrate homeostasis' http://purl.obolibrary.org/obo/GO_0033500 'carbohydrate mediated signaling' http://purl.obolibrary.org/obo/GO_0009756 'carbohydrate transmembrane transporter activity' http://purl.obolibrary.org/obo/GO_0015144

Right?

deustp01 commented 7 months ago

Coming in very late here, but I wonder if a better fix is more modularity. Off the top of my head, most carbohydrate derivatives are synthesized by joining a pre-existing carbohydrate (mono- or oligosaccharide) to a preexisting something else, e.g., a protein, to make a carbohydrate derivative. And in the other direction an early step in catabolism breaks the covalent connection between the carbohydrate and the something else, leaving two pieces that are catabolized further, separately. (Full disclosure: a textbook exception, possibly unique but important, is purine biosynthesis where the input is ribose, to which bits are added to construct a purine ribonucleoside.)