Why is GO:0006676 mannosyl diphosphorylinositol ceramide metabolic process is_a GO:0006505 GPI anchor metabolic process

[ValWood]

GO:0006676 mannosyl diphosphorylinositol ceramide metabolic process
The chemical reactions and pathways involving mannosyl diphosphorylinositol ceramide, any lipid with a phosphodiester bridge between an inositol residue and the ceramide group which contains two phosphoryl (-P(O)=) groups and a mannose derivative.

is_a

GO:0006505 GPI anchor metabolic process

but I can't see why?

13 annotations, 8 EXP from
Complex Portal 5
CGD 2
SGD1

genes involved are IPT1, SUR2, CGS2, CSH1

[ValWood]

I have the same genes annotated to

mannosyl-inositol phosphorylceramide biosynthetic process (GO:0051999)

The chemical reactions and pathways resulting in the formation of mannosyl-inositol phosphorylceramide, any lipid with a phosphodiester bridge between an inositol residue and the ceramide group which contains a phosphoryl (-P(O)=) groups and a mannose derivative.

which is a subclass of glycosphingolipid biosynthetic process
This makes much more sense?

maybe a confusion between "glycosphingolipid" and "glycosylphosphatidylinositol "?

@edwong57 @suzialeksander could you check this at SGD?

Is this term required?

[edwong57]

I'll look into it

[edwong57]

I think this GPI anchor metabolic process parentage might be coming from the ChEBI ontology. If you look in Protege and the explanation for why GO:0006676 is a subclass of GPI anchor metabolic process, you see:

In ChEBI, mannosyl diphosphorylinositol ceramide (ChEBI:74997) is conjugate base of mannosyl diphosphorylinositol ceramide (CHEBI:61006). mannosyl diphosphorylinositol ceramide (CHEBI:61006) is a phosphatidylinositol mannoside (CHEBI:59466) and phosphatidylinositol mannoside (CHEBI:59466) is a glycosylphosphatidylinositol (CHEBI:24410), which has been used in the definition of GPI anchor metabolic process.

Maybe cleaning that up would clean up the parentage?

If you look at this sphingolipid pathway (https://pathway.yeastgenome.org/YEAST/new-image?type=PATHWAY&object=SPHINGOLIPID-SYN-PWY-1&detail-level=2), maybe parentage of GO:0051999 and GO:0006676 should be similar?

[ValWood]

[edwong57]

@ValWood, could you please clarify what message you are trying to convey with this screenshot? I was suggesting that there may be an issue with ChEBI which is why GO:0006676 is a child of GPI anchor metabolic process. If that's the case, then fixing ChEBI might fix that issue. I also think it makes sense that GO:0006676 should be changed to have similar parentage to GO:0051999

[ValWood]

Hi @edwong57 ignore that.
I will ned to request a change from CHEBI but I'm not sure what , I will ask @pgaudet to show me.