GMP synthase activity terms

[sjm41]

ID: GO:0003922 Term: GMP synthase (glutamine-hydrolyzing) activity Def: Catalysis of the reaction: ATP + xanthosine 5'-phosphate + L-glutamine + H2O = AMP + diphosphate + GMP + L-glutamate. EC xref: 6.3.5.2

ID: GO:0003921 Term: GMP synthase activity Def: Catalysis of the reaction: ATP + xanthosine 5'-phosphate + NH3 = AMP + diphosphate + GMP. EC xef: 6.3.4.1

According the EC, 6.3.4.1 has been transferred to 6.3.5.2. https://enzyme.expasy.org/EC/6.3.4.1

So I think the two GO terms be merged, retaining the xref to 6.3.5.2. Actually, the relationships between the two current GO terms doesn't look right (see attachment), so some sort of edit does look necessary.

[sjm41]

Tagging @hattrill

[ukemi]

This certainly seems reasonable considering the merge in EC. I will do the due diligence with other xrefs, but will most likely merge.

[ukemi]

Looks like these refer to Rhea 18301 (no ec xref, but asserted as part of 11680) and Rhea 11680 (EC 6.3.5.2 xref). @alanbridge we should coordinate how we proceed here.

[ukemi]

@deustp01 A Reactome query shows R-HSA-56085 --> GO:0003922 and GO:0003921, R-HSA-73792 --> GO:0003921 and --> Rhea:11681.

[deustp01]

This looks like a simple curation mistake in Reactome. The reaction we have annotated, R-HSA-73792, is XMP + L-Glutamine + ATP + H2O => GMP + L-Glutamate + AMP + pyrophosphate, which corresponds to GO molecular function GO:0003922 GMP synthase (glutamine-hydrolyzing) activity "Catalysis of the reaction: ATP + xanthosine 5'-phosphate + L-glutamine + H2O = AMP + diphosphate + GMP + L-glutamate." (RHEA:11680).

The use of GO:0003921 GMP synthase activity "Catalysis of the reaction: ATP + xanthosine 5'-phosphate + NH3 = AMP + diphosphate + GMP." (RHEA:18301) to annotate R-HSA-73792 is wrong. It's fixed in the internal database and will propagate to the public site with our next release in December.

[ukemi]

Thanks @deustp01. In an ideal world the new alignment project would be able to spot these kinds of inconsistencies and flag them for review. @goodb just as a heads up for you to think about.

[goodb]

Yes @ukemi looks like a great example. We should pin this and use it as a test case.

[pgaudet]

We can go ahead and merge.

[ukemi]

Looking at this again and based on discussions from last week. To leverage from the Rhea reactions we should:

GO:0003921- remove the EC xrefs (they aren't in Rhea), and tweak the definition to match Rhea exactly. Make this term a part_of GO:0003992 to match the Rhea structure.

GO:0003992- Retain the EC xref to 6.3.5.2 as it is in Rhea and tweak the definition to match Rhea.

[sjm41]

This sounds good, thanks David.