search

geneontology / go-ontology

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

GO:0003997 acyl-CoA oxidase activity children #28733

Open sjm41 opened 3 weeks ago

sjm41 commented 3 weeks ago

We currently have this:

GO:0003997 acyl-CoA oxidase activity (EC:1.3.3.6, RHEA:38959)
    |_GO:0044535 very-long-chain fatty acyl-CoA oxidase activity -no xrefs (no annotations)
    |_GO:0016401 palmitoyl-CoA oxidase activity (no xrefs)- 4 EXP (rat Acox1, rat Acaa1a, mouse Acadl, human ACAA1)
    |_GO:0016402 pristanoyl-CoA oxidase activity (RHEA:40459) - 3 EXP to rat Acox3

GO:0044535 could get the xref RHEA:78847. I don't see a RHEA reaction corresponding to GO:0016401.

But note that RHEA:38959 (xref on the parent GO:0003997 term) lists 22 additional specific forms of that reaction - I don't think we want to add all of them.

I think the best solution here is to keep the very-long-chain term (GO:0044535/RHEA:78847) and add sister terms for short (RHEA:78859) /medium (RHEA:78855)/long chain (RHEA:78851).

This is in-line with what we've done in other areas of fatty acid biology.

If we did that, then GO:0016402 pristanoyl-CoA oxidase activity and GO:0016401 palmitoyl-CoA oxidase activity would be merged into the new long-chain term (CHEBI says they are both long-chain fatty acyl-CoAs)

So we'd end up with this:

GO:0003997 acyl-CoA oxidase activity (EC:1.3.3.6, RHEA:38959)
    |_GO:0044535 very-long-chain fatty acyl-CoA oxidase activity (RHEA:78847)
    |_NTR short-chain fatty acyl-CoA oxidase activity [RHEA:78859]
    |_NTR medium-chain fatty acyl-CoA oxidase activity [RHEA:78855]
    |_NTR long-chain fatty acyl-CoA oxidase activity [RHEA:78851]
pgaudet commented 3 weeks ago

That seems consistent with how we have handled other fatty acid terms - do you want to create the new terms?

pgaudet commented 3 weeks ago

@sjm41 - for reference: I think GO:0016401 palmitoyl-CoA oxidase activity corresponds to RHEA:40167 (RHAA calls palmitoyl-CoA, "hexadecanoyl-CoA", see https://www.ebi.ac.uk/chebi/searchId.do?chebiId=CHEBI:57379

So that would merge into NTR long-chain-chain fatty acyl-CoA oxidase activity [RHEA:78851] (as you suggested)

pgaudet commented 3 weeks ago

However for GO:0016402 pristanoyl-CoA oxidase activity, it's not a child of long chain fatty acid in ChEBI, it's a branched chained (and they dont classify it WRT length), so should that not mere into the parent GO:0003997 acyl-CoA oxidase activity ?

sjm41 commented 2 weeks ago

That seems consistent with how we have handled other fatty acid terms - do you want to create the new terms?

Will do, though I'm out of office for a couple of days now.

sjm41 commented 2 weeks ago

+[Term] +id: GO:0120522 +name: short-chain fatty acyl-CoA oxidase activity +namespace: molecular_function +def: "Catalysis of the reaction: a short-chain 2,3-saturated fatty acyl-CoA + O2 = a short-chain (2E)-enoyl-CoA + H2O2." [RHEA:78859] +comment: While there is not universal consensus on the lengths of short-, medium-, long- and very-long-chain fatty acids, the GO uses the definitions in ChEBI (see CHEBI:26666, CHEBI:59554, CHEBI:15904 and CHEBI:27283). +xref: RHEA:78859 +is_a: GO:0003997 ! acyl-CoA oxidase activity +property_value: term_tracker_item "https://github.com/geneontology/go-ontology/issues/28733" xsd:anyURI +created_by: sjm +creation_date: 2024-08-27T08:19:22Z + +[Term] +id: GO:0120523 +name: medium-chain fatty acyl-CoA oxidase activity +namespace: molecular_function +def: "Catalysis of the reaction: a medium-chain 2,3-saturated fatty acyl-CoA + O2 = a medium-chain (2E)-enoyl-CoA + H2O2." [RHEA:78855] +comment: While there is not universal consensus on the lengths of short-, medium-, long- and very-long-chain fatty acids, the GO uses the definitions in ChEBI (see CHEBI:26666, CHEBI:59554, CHEBI:15904 and CHEBI:27283). +xref: RHEA:78855 +is_a: GO:0003997 ! acyl-CoA oxidase activity +property_value: term_tracker_item "https://github.com/geneontology/go-ontology/issues/28733" xsd:anyURI +created_by: sjm +creation_date: 2024-08-27T08:21:49Z + +[Term] +id: GO:0120524 +name: long-chain fatty acyl-CoA oxidase activity +namespace: molecular_function +def: "Catalysis of the reaction: a long-chain 2,3-saturated fatty acyl-CoA + O2 = a long-chain (2E)-enoyl-CoA + H2O2." [RHEA:78851] +comment: While there is not universal consensus on the lengths of short-, medium-, long- and very-long-chain fatty acids, the GO uses the definitions in ChEBI (see CHEBI:26666, CHEBI:59554, CHEBI:15904 and CHEBI:27283). +xref: RHEA:78851 +is_a: GO:0003997 ! acyl-CoA oxidase activity +property_value: term_tracker_item "https://github.com/geneontology/go-ontology/issues/28733" xsd:anyURI +created_by: sjm +creation_date: 2024-08-27T08:24:03Z

sjm41 commented 2 weeks ago

@pgaudet I've made the new terms, so the remaining tasks for you are:

deustp01 commented 2 weeks ago

then GO:0016402 pristanoyl-CoA oxidase activity and GO:0016401 palmitoyl-CoA oxidase activity would be merged into the new long-chain term (CHEBI says they are both long-chain fatty acyl-CoAs)

But handling branches requires distinct biochemistry, so would it be useful to have a branched-chain versus straight-chain distinction in the MF hierarchy?

sjm41 commented 2 weeks ago

But handling branches requires distinct biochemistry, so would it be useful to have a branched-chain versus straight-chain distinction in the MF hierarchy?

I see what you mean.
But adding that term to GO would result in two orthogonal classification systems for these terms - one classified by fatty acid length, and one classified by straight vs branched chain. We have that kind of thing elsewhere in the ontology, so it's possible, but can get complicated. Given there's currently only 3 EXP annotations to a single gene (rat Acox3) for the sole branched chain term "GO:0016402 pristanoyl-CoA oxidase activity", I wonder if it's worth adding this complication at the present time?

deustp01 commented 2 weeks ago

I see what you mean. ... I wonder if it's worth adding this complication at the present time?

I will dig out the biochemistry books tomorrow when I'm in my office, to see if this can be simplified. In particular, it may be that branches are only found on fatty acids of certain lengths (longer ones perhaps), and if so one or a few additional siblings - straight long-chain, branched long chain, etc. might do the job. Maybe also a weeds issue for Thursday? @ukemi

pgaudet commented 2 weeks ago

For GO, one major decision factor has been enzyme specificity. If the enzymes don't distinguish between branched and linear substrates, then we would not create/keep the terms.

deustp01 commented 2 weeks ago

If the enzymes don't distinguish ...

But that in fact, if I remember right, is exactly what the relevant enzymes do - the standard ones choke on the branch points and a separate group deals with the branches, all as part of the overall process of converting a large fatty acid to the maximum possible number of molecules of acetyl-CoA.

deustp01 commented 2 weeks ago

Here is the most recent review I can find on the enzymology of branched-chain fatty acid breakdown, from the authority in the field: PMID: 21683154. If a paywall keeps you from getting the article, let me know and I will e-mail you a copy.

Looking into the paper a bit, I see what may be the key bit: after some initial maneuvering, cycles of beta-oxidation at branch points yield propionyl-CoA (instead of canonical acetyl-CoA) and the propionyl-CoA can go into its own textbook catabolic pathway to yield methylmalonyl-CoA and ultimately succinyl-CoA, a TCA cycle intermediate.

So while GO terms are needed for alpha-oxidation as described here (they may well already exist - too distracted here to look for this), it may well be that everything else can be covered by the terns Steven has already proposed as well as the ones for methylmalonyl-CoA catabolism.