The blinded OpenScientist stress test of GO:0030267
[file:PSEPK/ptxD/ptxD-hypotheses/function-hypothesis-go-0030267/openscientist.md
"By independent global pairwise alignment the 320-aa query is 53.4% identical to the biochemically characterized E. coli GhrB"]
judged the TreeGrafter graft sound and recommended keeping GO:0030267 and
adding GO:0016618. It was blinded to gene context.
Checks made in this re-review (scratch alignments, Biopython PairwiseAligner,
global, BLOSUM62, gap open -10 / extend -0.5; identities as % of the 320-aa
query; sequences from UniProt REST):
UniProt P37666 (GhrB, Swiss-Prot) lists EC 1.1.1.79, EC 1.1.1.81 and
EC 1.1.1.215 (2-ketogluconate reductase, RHEA:16653), so the subfamily is
promiscuous across 2-oxoacids including a 2-ketoaldonate.
KEGG (rest.kegg.jp, ppu): PP_3376 is K00032 phosphogluconate 2-dehydrogenase
(EC 1.1.1.43); PP_3377 putative 2-ketogluconate transporter; PP_3378 putative
2-ketogluconokinase (K11441); PP_3379 putative epimerase; PP_3380
2-ketogluconate utilization repressor (PtxS). This is the kguEKTD-ptxS
arrangement described for Pseudomonas PMID:29607620.
Decision: the report and the earlier review agree on the facts (GhrB-type,
active, not PtxD) and differ on interpretation (glyoxylate/hydroxypyruvate
reductase vs KguD). GO:0030267 and GO:0016618 stay UNDECIDED; the GO:0050609
REMOVE is reinforced. No new MF term is needed: GO:0008114 already covers the
KguD reaction (EC 1.1.1.43, synonym "2-keto-6-phosphogluconate reductase
activity").