UBC9 (P50623) subfamily placement and active-site check

Script: subfamily_and_active_site.py (run with uv run --script subfamily_and_active_site.py;
raw output in results-run.txt). Data were pulled live from the PANTHER geneinfo
service and UniProt REST on 2026-10-21.

PANTHER subfamily placement

accession	protein	PANTHER subfamily	identity to P50623	residue at yeast C93
P50623	S. cerevisiae UBC9	PTHR24067:SF248	157/157 (100.0%)	C93
P40984	S. pombe hus5	PTHR24067:SF248	95/157 (60.5%)	C93
Q1K772	N. crassa NCU04302	PTHR24067:SF248	91/156 (58.3%)	C92
P63279	H. sapiens UBE2I	PTHR24067:SF280	88/157 (56.1%)	C93

S. cerevisiae Ubc9 P50623 is placed in PANTHER subfamily PTHR24067:SF248, together with S. pombe hus5 P40984 and Neurospora NCU04302 Q1K772; human UBE2I P63279 is in PTHR24067:SF280.

The yeast catalytic cysteine Cys93 (UniProt ACT_SITE, glycyl thioester intermediate) aligns to a cysteine in all three other proteins.

Interpretation

SF248 (fungal Ubc9 proteins) and SF280 (vertebrate UBE2I) are sister subfamilies.
Both descend from the PAINT Ubc9 node PTN000629675, and SGD:S000002222 (this gene)
is one of that node's experimental seeds (see interpro/panther/PTHR24067/PTHR24067-paint.tsv).
Global identity to the fission-yeast and Neurospora proteins (58-61%) is only a little
higher than to human UBE2I (56%). The subfamily split does not reflect a large difference
in sequence and does not show any change of function. The node-level terms (SUMO
conjugating enzyme activity, protein sumoylation, nucleus) apply to both branches.
The catalytic cysteine is retained in every protein checked, so nothing here points to a
pseudoenzyme. Retaining the residue does not by itself prove activity: the S. pombe
Hus5.62 mutant still forms a thioester but is defective in conjugation.