Nitrogen Cycle Metabolic Process — do_not_annotate / Regulation Term Obsoletion

IN_PROGRESS OBSOLETION

Nitrogen Cycle Metabolic Process — do_not_annotate / Regulation Term Obsoletion

Overview

A GO obsoletion proposal will:

The motivation (per upstream) is that "nitrogen cycle metabolic process" is an
ecosystem-level grouping term: real biology happens in its children
(denitrification, nitrification, nitrogen fixation, mineralization). Direct
annotation to the parent erases mechanistic detail and — more importantly — has
attracted over-annotations of mammalian genes that have nothing to do with the
ecological nitrogen cycle.

Upstream tickets

Obsoletion / do_not_annotate plan (per upstream)

Term ID Action Replacement strategy
nitrogen cycle metabolic process GO:0071941 do_not_annotate (stays as grouping term) Re-route to a descendant: e.g. GO:0019333 denitrification pathway, GO:0009399 nitrogen fixation, GO:0008152 / a nitrification or mineralization child as appropriate
regulation of nitrogen cycle metabolic process GO:1903314 obsolete No replacement currently slated — only relevant if a child regulation term exists for the specific pathway
negative regulation of nitrogen cycle metabolic process GO:1903315 obsolete Same
positive regulation of nitrogen cycle metabolic process GO:1903316 obsolete Same

Term labels verified in OLS on 2026-05-13. All four terms are still live.
The replacement term GO:0019333 (denitrification pathway) was also verified.

Affected experimental annotations

The upstream issue cites 5 MGI + 1 CACAO experimental annotations.
QuickGO (exact-term query against GO:0071941, evidence codes IDA / IMP / IGI /
IPI / IEP) on 2026-05-13 returns 6 direct experimental annotations — matching
the upstream count:

# Gene Accession Taxon Qualifier Evidence PMID Assigned by Replacement plan
1 Prkcsh UniProtKB:O08795 Mus musculus (10090) acts_upstream_of_or_within IGI PMID:21685914 MGI REMOVE — Prkcsh is glucosidase II β / Pkd-network polycystic-liver gene; no nitrogen-cycle role
2 Pkd1 UniProtKB:O08852 Mus musculus (10090) acts_upstream_of_or_within IGI PMID:21685914 MGI REMOVE — Polycystin-1, cystogenic Ca²⁺ signalling; no nitrogen-cycle role
3 Pkd1 (dup) UniProtKB:O08852 Mus musculus (10090) acts_upstream_of_or_within IGI PMID:21685914 MGI REMOVE — Duplicate row; flag both annotations
4 Apc UniProtKB:Q61315 Mus musculus (10090) acts_upstream_of_or_within IMP PMID:16740478 MGI REMOVE — Adenomatous polyposis coli (Wnt pathway / liver zonation); no nitrogen-cycle role
5 Sec63 UniProtKB:Q8VHE0 Mus musculus (10090) acts_upstream_of_or_within IGI PMID:21685914 MGI REMOVE — Sec63 translocon component; no nitrogen-cycle role
6 napA UniProtKB:O88111 Cereibacter sphaeroides f. sp. denitrificans (39723) involved_in IMP PMID:10227138 CACAO MODIFY → GO:0019333 denitrification pathway — Periplasmic nitrate reductase; legitimate denitrification gene

Five of the six annotations look like classic over-annotation of mammalian
kidney/liver genes (PKD1, PRKCSH, SEC63 are the canonical autosomal dominant
polycystic kidney/liver disease genes; APC drives a liver-zonation phenotype).
The 1999 paper supporting napA (PMID:10227138, Bisci. Biotechnol. Biochem.) is
explicitly about denitrification in Rhodobacter (Cereibacter) sphaeroides, so
the bacterial annotation is biologically sound but should be moved to the
specific child term GO:0019333 (denitrification pathway), not stranded on the
parent.

The three regulation children (GO:1903314 / 1903315 / 1903316) currently have
zero experimental annotations in QuickGO (verified 2026-05-13), so their
obsoletion is uncontested from an annotation-impact standpoint.

Impact on this repo

None of the affected genes (mouse Prkcsh / Pkd1 / Apc / Sec63 or
Cereibacter napA) currently have an *-ai-review.yaml in this repo
(verified by find genes -iname '*Prkcsh*' -o -iname '*Pkd1*' -o -iname '*Apc*' -o -iname '*Sec63*' -o -iname '*napA*' and a directory scan of
genes/mouse, genes/human on 2026-05-13). The project is therefore a
queueing exercise plus a clear documentation of the over-annotation pattern
this term has accumulated.

Scope

Candidate genes for initial review

Listed in priority order. Each should be set up with
just fetch-gene <organism> <gene> before review begins. Mouse genes are
written with the MGI symbol casing; substitute the species code as appropriate
for the fetch-gene invocation (e.g. mouse for MGI orthologs, or pull the
human ortholog if better-characterised).

  1. PKD1 (mouse Pkd1, UniProtKB:O08852; human PKD1) — Polycystin-1. The
    IGI annotation to GO:0071941 from PMID:21685914 is a clear over-annotation
    and a useful case study: a kidney-disease genetic-interaction paper has
    propagated an ecological-nitrogen-cycle term onto a mammalian
    mechanotransduction gene. The human PKD1 review would carry the most
    downstream value because PKD1 is heavily studied; flag the mouse-MGI
    annotation in the existing-annotations review.
  2. PRKCSH (mouse Prkcsh, UniProtKB:O08795; human PRKCSH) —
    Protein kinase C substrate 80K-H / glucosidase II β subunit. Same paper
    (PMID:21685914), same critique. PRKCSH's actual core function is N-linked
    glycoprotein quality control in the ER (glucosidase II regulatory
    subunit); reviewing it would let us cleanly remove the nitrogen-cycle
    over-annotation in the same pass.
  3. SEC63 (mouse Sec63, UniProtKB:Q8VHE0; human SEC63) —
    Translocon subunit, ER protein translocation. Same paper, same critique.
    Human SEC63 is the canonical autosomal dominant polycystic liver disease
    gene and a clean target for a focused over-annotation removal.
  4. APC (mouse Apc, UniProtKB:Q61315; human APC) —
    Adenomatous polyposis coli, β-catenin destruction complex scaffold. The
    IMP annotation from PMID:16740478 (Apc as "zonation-keeper" of mouse
    liver) does not support a nitrogen-cycle role; the paper's phenotype is
    metabolic zonation in hepatocytes, which involves urea-cycle and
    ammonia-handling enzymes downstream — not a nitrogen-cycle role for APC
    itself. Recommend REMOVE of the GO:0071941 annotation in the APC review.
  5. napA (Cereibacter sphaeroides f. sp. denitrificans, UniProtKB:O88111) —
    Periplasmic nitrate reductase. Biologically the cleanest case: the
    PMID:10227138 IMP evidence supports involvement in denitrification, so
    the GO:0071941 annotation should be modified to GO:0019333
    denitrification pathway. Use the species code corresponding to taxon
    39723 (Cereibacter sphaeroides f. sp. denitrificans). This is a
    non-canonical organism for the repo and may not have an obvious species
    subdirectory yet; check before invoking just fetch-gene.

Proposed approach

  1. Pattern-document the mammalian over-annotation first. Add a short
    section to projects/OVER_ANNOTATION_PATTERNS.md (or a new entry there)
    noting that GO:0071941 has attracted polycystic kidney/liver genes via
    IGI/IMP propagation from PMID:21685914 and PMID:16740478. This is a
    reusable insight even if the obsoletion is delayed.
  2. Prioritise the human orthologs over the mouse MGI entries. Human
    PKD1, PRKCSH, SEC63, APC are well-studied and likely to attract more
    downstream consumers of any review we publish. The mouse-MGI annotation
    can be cited within those reviews under existing_annotations (or under
    references with a brief note).
  3. Move napA last. It is biologically correct and only needs a term
    refinement (MODIFY); it does not have the diagnostic value of the
    mammalian cluster. If the Cereibacter species directory does not yet
    exist, this is a small enough single-annotation case to defer.
  4. Wait for the do_not_annotate decision before bulk-rewriting. GO
    ontology ticket #27220 may still be in discussion. The action codes
    (REMOVE vs MODIFY) in any reviews started now are not affected by
    the eventual ontology decision; do not block on the upstream timing.

Priority

Medium-low. The total annotation set is tiny (6 rows), but the over-annotation
pattern is illustrative — five out of six annotations on a microbial-ecology
term are on mammalian kidney/liver genes. Reviewing PKD1 / PRKCSH / SEC63
fits the repo's existing emphasis on cleaning up over-annotations of pleiotropic
genes (see projects/OVER_ANNOTATION_PATTERNS.md).

Status