Sulfide Oxidation Children — Obsoletion & Replacement

Bottom line: Mitochondria detoxify hydrogen sulfide by oxidising it to
sulfite and sulfate, starting with sulfide:quinone oxidoreductase (SQOR).
GO has obsoleted three children of GO:0019418 sulfide oxidation that
named the enzyme used (GO:0070221 via SQOR, GO:0070222 via sulfide
dehydrogenase, GO:0070223 via sulfur dioxygenase), because they were
more specific than any known gene product and two had no annotations.
All three ontology PRs are merged, so annotations now belong on the
parent. We recorded the upstream changes, listed the direct annotations
that must move (human SQOR IDA, TSTD1 TAS, SLC25A10 TAS, plus rodent
Sqor ISS/ISO rows), and queued SQOR, the one well-characterised enzyme
on the list, for review. Partly done: human/SLC25A10 has since been
reviewed in another workstream, which the impact section below
predates; its GOA file already carries the Reactome row on GO:0019418,
and the review marks it MARK_AS_OVER_ANNOTATED because the carrier
exports sulfate but does not oxidise sulfide. SQOR and TSTD1 are still
unreviewed.

Overview

A GO obsoletion has retired three molecular-mechanism-specific
children of GO:0019418 sulfide oxidation, because they were more
specific than any known gene product and 2 of the 3 had no
annotations. All affected annotations should be migrated to the
parent term GO:0019418 (with the appropriate substrate/specificity
captured in evidence rather than the term ID).

• GO:0070221 sulfide oxidation, using sulfide:quinone oxidoreductase (BP, obsolete)
• GO:0070222 sulfide oxidation, using sulfide dehydrogenase (BP, obsolete)
• GO:0070223 sulfide oxidation, using sulfur dioxygenase (BP, obsolete)

Replaced by:

• GO:0019418 sulfide oxidation (BP)

The textual definition of GO:0019418 was simultaneously updated to
"The chemical reactions and pathways resulting in the conversion of
sulfide to sulfite or sulfate." The MetaCyc cross-references on the
obsoleted children were redirected to GO:0019418
(MetaCyc:P222-PWY, MetaCyc:PWY-5274, MetaCyc:PWY-5285,
MetaCyc:PWY-7927).

Upstream tickets

• Annotation tracker: geneontology/go-annotation#6388
• Ontology ticket: geneontology/go-ontology#31842 — closed 2026-05-11
• Ontology obsoletion PRs (all merged):
• geneontology/go-ontology#31949 — obsoleted GO:0070222 and GO:0070223; moved MetaCyc xrefs and added synonyms (merged 2026-04-22)
• geneontology/go-ontology#32025 — obsoleted GO:0070221 (merged 2026-05-04)
• geneontology/go-ontology#32068 — changed textual definition for GO:0019418 (merged 2026-05-11)

Obsoletion plan (per upstream)

Affected experimental / direct annotations (GO:0070221 only)

GO:0070222 and GO:0070223 had zero direct annotations at
obsoletion time. All affected annotations are on GO:0070221.

A larger pool of IBA (ECO:0000318) annotations propagated from
PANTHER will be remapped to GO:0019418 automatically once the next
PAINT/IBA pipeline run picks up the obsoletion. These span SQOR
orthologs across vertebrates (Bos taurus, Gallus gallus, Xenopus
tropicalis, Danio rerio, Anolis carolinensis, gorilla, chimp, etc.),
fungal/protist orthologs (Paramecium tetraurelia, Aspergillus
nidulans, Neurospora crassa, Cryptococcus deneoformans, S. pombe
hmt2, S. japonicus hmt2, Dictyostelium purpureum), bacterial homologs
(Pseudomonas aeruginosa PA2345, Staphylococcus aureus, Chloroflexus
aurantiacus), and a Monosiga brevicollis SQOR. These are
auto-migrated and do not need per-annotation work here.

InterPro2GO / mapping cleanup (per upstream issue)

• InterPro:IPR042457 Thiosulfate:glutathione sulfurtransferase,
  mammalian → GO:0070221 — mapping removed from InterPro (will
  appear in InterPro release 109.0; confirmed 2026-05-15 in
  go-annotation#6388 comments).
• Reactome has already migrated its 2 affected annotations
  (confirmed 2026-04-22 in upstream issue comments).
• No ARBA / UniRule mappings to the obsoleted terms were listed.

Impact on this repo

No genes annotated to GO:0070221/222/223 are currently reviewed in
this repo:

• genes/human/SQOR/ — does not exist
• genes/human/TSTD1/ — does not exist
• genes/human/SLC25A10/ — does not exist
• genes/PSEAE/PA2345/ — does not exist
• genes/SCHPO/hmt2/ — does not exist

So no existing reviews need refresh for the obsoletion itself.
The opportunity here is that SQOR is otherwise unreviewed in
this repo despite being a well-characterized human mitochondrial
sulfide-oxidizing enzyme directly relevant to H2S signaling, sulfide
detoxification, and the inherited mitochondrial disorder caused by
SQOR deficiency (PMID:31108281; affects 1 of the 3 known annotations
queued for migration).

Scope

• Organisms: primary candidates are human (existing genes/human/)
  and S. pombe (existing genes/SCHPO/). All other affected genes
  are IBA propagations and will migrate automatically.
• GO branches: BP only — the obsoletion is a
  mechanism-specificity collapse to the broader parent term. No MF
  or CC changes.
• Type of fix: terminological — the biology (sulfide → sulfite/
  sulfate) is unchanged; reviews would evaluate whether GO:0019418
  is the correct BP for SQOR's core function or whether more
  appropriate descendants (e.g. GO:0070221 is gone, but the
  cognate MF GO:0047804 sulfide:quinone reductase activity is
  intact) better capture SQOR's mechanism.

Candidate genes for initial review

Verify each with just fetch-gene <organism> <gene> before starting
and confirm UniProt accessions. None are currently in the repo.

Tier 1 — direct experimental annotation, well-characterized

1. SQOR / Q9Y6N5 (Homo sapiens) — mitochondrial sulfide:quinone
   oxidoreductase, the canonical human enzyme for sulfide oxidation
   in the first step of mitochondrial H2S detoxification. The IDA
   annotation on GO:0070221 is the 1 EXP that the upstream issue
   flags for migration. SQOR also carries an IBA on GO:0070221. A
   review here should anchor the BP on GO:0019418 (post-obsoletion)
   and confirm the MF anchor remains GO:0047804 sulfide:quinone
   reductase activity. Clinically relevant (SQOR deficiency causes
   Leigh-like mitochondrial encephalopathy).

Tier 2 — TAS, less direct

1. TSTD1 / Q8NFU3 (Homo sapiens) — thiosulfate sulfurtransferase-
   like domain-containing 1; carries a TAS annotation on GO:0070221
   plus an IEA. Its primary characterized activity is rhodanese-type
   thiosulfate:glutathione sulfurtransferase (PMID:24107290), and
   InterPro has already removed the IPR042457 → GO:0070221 mapping
   per the upstream issue. A review would assess whether GO:0019418
   (the obsoletion replacement) or a different sulfurtransferase BP
   is the right placement.
2. SLC25A10 / Q9UBX3 (Homo sapiens) — mitochondrial
   dicarboxylate carrier; TAS on GO:0070221. The annotation is on a
   pathway role rather than direct enzymatic activity; review would
   determine whether GO:0019418 is appropriate or whether this
   annotation should be retracted in favor of transport-centric BPs.

Tier 3 — opportunistic, non-human

1. hmt2 / SPBC2G5.06c (Schizosaccharomyces pombe) — fission yeast
   ortholog carrying an IBA on GO:0070221. Would extend SCHPO
   coverage but lower priority since IBA migration is automatic.

Proposed approach

1. Obsoletion has landed. All three terms are obsolete in the
   ontology and the upstream ontology ticket is closed. The
   annotation-migration work in go-annotation#6388 is the only
   remaining loose end upstream (still open as of 2026-05-15, with
   the InterPro mapping removed the same day).
2. Start with human SQOR. Run just fetch-gene human SQOR and
   confirm the GOA pull picks up the obsoletion: GO:0070221 should
   appear as an obsolete term (or already have been remapped to
   GO:0019418). Review per CLAUDE.md, paying attention to:
3. The cognate MF GO:0047804 sulfide:quinone reductase activity
   (intact) is the right MF anchor.
4. The BP anchor should be GO:0019418 sulfide oxidation
   (post-obsoletion).
5. SQOR also participates in CoQ/electron-transport context;
   check that any CC/BP claims about mitochondrial ETC
   integration are evidence-backed.
6. Follow with TSTD1 then SLC25A10. Both are TAS-only on the
   obsoleted term and the upstream mapping (IPR042457 → GO:0070221)
   has already been removed; review should evaluate whether
   GO:0019418 is even the right BP or whether the original TAS was
   on shaky ground.
7. Defer S. pombe hmt2 and the bacterial homologs unless they
   come up via other workstreams; the IBA propagations will be
   auto-migrated.
8. Do not create reviews for the IBA-only orthologs listed in
   the impact table — they will be remapped automatically.

Priority

Medium-low. Only 3 direct (non-IBA) annotations are affected
and the ontology change is already in production. The migration
upstream is small. However, human SQOR is a clinically and
biologically important gene with no review yet in this repo, so
this obsoletion is a reasonable trigger to add it. The other two
human entries (TSTD1, SLC25A10) are TAS-only and lower value.

Status

• 2026-05-15 — Project file created. Upstream annotation issue
  #6388 still open; latest activity was InterPro confirming
  removal of the IPR042457 → GO:0070221 mapping. All three ontology
  obsoletion PRs are merged (#31949, #32025, #32068) and the
  parent ontology ticket #31842 is closed. No gene reviews started
  in this repo for SQOR/TSTD1/SLC25A10.

Slides

• Slides (Marp source: SULFIDE_OXIDATION_OBSOLETION-slides.md) — AI generated