Neuroblast Proliferation / Division — Vertebrate vs Protostome Disambiguation

Overview

The word neuroblast denotes two biologically distinct cell types:

Vertebrates (mouse, rat, human): a neuroblast is a postmitotic cell
that does not divide further. It migrates and then differentiates into a
neuron. (Cf. Cell Ontology "neuroblast (sensu Vertebrata)".)
Protostomes / flies: a neuroblast is a mitotic, self-renewing neural
stem cell that divides asymmetrically to produce ganglion mother cells.
(Cf. "neuroblast (sensu Nematoda and Protostomia)".)

GO terms in the "neuroblast proliferation / neuroblast division" branch were
written around the protostome (dividing stem-cell) sense — e.g. GO:0007405
neuroblast proliferation is defined as "The expansion of a neuroblast
population by cell division. A neuroblast is any cell that will divide and give
rise to a neuron." That definition is incompatible with the vertebrate
postmitotic neuroblast, yet the QuickGO download attached to the upstream issue
shows the branch is annotated almost entirely with vertebrate genes (mouse,
rat, human). In vertebrates the thing that actually proliferates is the
neural precursor / progenitor cell (GO:0061351 neural precursor cell
proliferation, verified in OLS 2026-05-18), not the neuroblast.

This is therefore not a simple obsoletion. The upstream plan is to separate
the two senses: keep the division branch for protostomes, migrate the
vertebrate annotations to vertebrate-appropriate proliferation/neurogenesis
terms, then rename the terms and re-map the Cell Ontology so the two senses can
never be confused again. This repo's contribution is the annotation-level
review: for each affected vertebrate gene, decide the correct replacement
term from the cited literature (mostly MODIFY → GO:0061351 or a more specific
neural-precursor-proliferation / neurogenesis child).

Upstream tickets

Annotation tracker: geneontology/go-annotation#6393
— "Review mouse annotations under 'neuroblast proliferation'" (curator
@sjm41; last upstream activity 2026-05-07).
No GO-ontology ticket yet. The upstream curator is still in the
scoping/discussion phase ("before I make the tickets for the ontology
changes and annotation migrations, I wanted to check if you think this is
the correct path"). The ontology renaming + CL re-mapping is out of scope
for this repo; our annotation MODIFY recommendations are an independent
input that does not need to wait for the ontology decision.

Affected terms (from the upstream QuickGO export, 42 rows)

GO ID	Label	Rows	Sense as written
GO:0007405	neuroblast proliferation	26	protostome (division-centric)
GO:0055059	asymmetric neuroblast division	4	protostome
GO:0021849	neuroblast division in subventricular zone	4	mixed / vertebrate SVZ
GO:0021847	ventricular zone neuroblast division	4	mixed / vertebrate VZ
GO:0045200	establishment of neuroblast polarity	2	mixed
GO:0021873	forebrain neuroblast division	2	vertebrate forebrain

Term IDs/labels are taken verbatim from the upstream QuickGO TSV; GO:0007405
and the proposed replacement GO:0061351 were independently re-verified in OLS
on 2026-05-18.

Scope analysis

Organisms (annotation rows): Mouse (NCBITaxon:10090) 35 rows, Human
(9606) 4 rows, Rat (10116) 3 rows = 42 rows total — i.e. 100% vertebrate,
confirming the conceptual mismatch. No fly/protostome annotations appear in
this export.
Distinct genes: ~33 (some mouse genes carry multiple annotations across
the 6 affected terms; ARHGEF2 additionally appears in both mouse and human).
The gene-count is approximate pending the per-gene review pass; the
authoritative figure is the 42-row annotation count.
Evidence: All experimental (IMP / IGI / IDA / IEP). Many are
acts_upstream_of_or_within MGI annotations (regulatory/upstream genes that
modulate progenitor proliferation, not neuroblast division per se).
Common problem: a protostome stem-cell term applied to vertebrate genes
whose cited papers describe neural precursor/progenitor proliferation,
cortical neurogenesis, or progenitor polarity/spindle orientation — not
the division of a (postmitotic) vertebrate neuroblast.
Expected action distribution: predominantly MODIFY
(proposed_replacement_terms → GO:0061351 neural precursor cell
proliferation or a VZ/SVZ/forebrain-specific neural-precursor child, chosen
per-gene from the cited PMID), with occasional REMOVE where the paper does
not support any proliferation/neurogenesis claim, and KEEP_AS_NON_CORE
where proliferation is a genuine but peripheral effect of a pleiotropic gene.

Candidate genes for initial review (priority order)

Prioritised by (a) already present in this repo, (b) human ortholog directly
annotated, (c) high-profile, well-characterised vertebrate neurogenesis genes
where the correct term is unambiguous.

#	Gene	Acc	Taxon	Term annotated	Ev	PMID	In repo?	Likely action
1	ASCL1 (Mash1)	Q02067	mouse	GO:0007405	IGI	PMID:15976074	human ortholog only (genes/human/ASCL1); human ASCL1 P50553 has no GO:0007405 annotation	MODIFY → GO:0061351 on the mouse Mash1 annotation (Mash1 drives progenitor proliferation/neurogenesis)
2	FGFR2	P21802	human	GO:0021847	ISS	GO_REF:0000024 (with/from UniProtKB:P21803 mouse)	genes/human/FGFR2 (already KEEP_AS_NON_CORE)	revisit: existing review action is KEEP_AS_NON_CORE; under this project consider MODIFY → vertebrate VZ neural-precursor proliferation
3	SOX5	—	human	GO:0055059	IGI	PMID:23946438	no (genes/human)	review — likely MODIFY/REMOVE (asymmetric neuroblast division is protostome)
4	ARHGEF2	Q60875	human+mouse	GO:0055059	IDA/IMP	PMID:28453519	no	MODIFY → progenitor mitotic spindle / asymmetric progenitor division
5	DOCK7	—	human	GO:0045200	IMP	PMID:16982419	no	review progenitor-polarity context
6	TEAD3	—	human	GO:0055059	IGI	PMID:23946438	no	review (co-cited with SOX5)
7	Pafah1b1 (LIS1)	P63005	mouse	GO:0007405	IMP	PMID:12629176	no	MODIFY → neural precursor proliferation (LIS1 = classic neurogenesis/migration gene)
8	Aspm	Q8CJ27	mouse	GO:0021873	IMP	PMID:16798874	no	MODIFY → forebrain neural-precursor proliferation (ASPM = microcephaly/progenitor gene)
9	Shh	Q62226	mouse	GO:0007405	IDA	PMID:15337776	no	MODIFY → neural precursor proliferation
10	Nde1	Q9CZA6	mouse	GO:0007405	IMP	PMID:15473967	no	MODIFY → neural precursor proliferation (NDE1 = microcephaly/progenitor)

Other affected mouse genes (defer to a second batch): Plxnb2, Numb, Numbl,
Neurod4, Kcna1, Fgfr1, Lef1, Dct, Id4, Nfix, Eml1, Zzef1, Dagla, Daglb, Hhip,
Ckap2l, Akna, Frs2, Acsl6, Racgap1, Fzd9; rat Gh1, Ifrd1, Rab10.

Proposed approach

Use the two repo genes (ASCL1, FGFR2) to demonstrate the pattern, but
note that neither carries the offending annotation on the human gene as
listed in the human GOA:
genes/human/ASCL1 has no GO:0007405 annotation; the upstream
annotation is on mouse Mash1 (Q02067) via IGI in PMID:15976074. The
existing_annotations block on the human ASCL1 review cannot be used
here; the appropriate channel is a new mouse-ASCL1 review under
genes/mouse/ASCL1 (or to record the recommendation in the project doc
and propagate upstream via geneontology/go-annotation#6393).
genes/human/FGFR2 already reviews GO:0021847 with
action: KEEP_AS_NON_CORE (ISS via GO_REF:0000024, with/from mouse
P21803). Under this project the action should be revisited — the
candidate replacement is GO:0061351 or a VZ-specific neural-precursor
proliferation child confirmed via OLS — but this is an action update
on an existing review, not a new existing_annotations entry.
Then the four human-annotated genes (SOX5, ARHGEF2, DOCK7, TEAD3) —
these live directly in genes/human and are the upstream-impacted
human set. just fetch-gene human <SYMBOL> then /review.
Then the high-profile mouse progenitor genes (LIS1/Pafah1b1, Aspm,
Nde1, Shh) as mouse reviews (genes/mouse/ exists). These are textbook
neural-precursor-proliferation genes; the MODIFY target is unambiguous and
they make a clean illustrative cluster.
Per-gene rule: read the cited PMID. If it describes proliferation of a
progenitor/stem population (VZ/SVZ radial glia, intermediate
progenitors) → MODIFY to GO:0061351 or the appropriate
neural-precursor-proliferation child. If it describes spindle
orientation / asymmetric progenitor division → MODIFY to the relevant
progenitor-division term, not "asymmetric neuroblast division". If
the paper does not support any proliferation/neurogenesis claim →
REMOVE. Never retain GO:0007405/GO:0055059 etc. on a vertebrate gene as
core.
Do not block on upstream ontology timing. The MODIFY/REMOVE action
codes are independent of whether/when GO renames the terms or re-maps the
Cell Ontology. Record the upstream rationale in each review's
existing_annotations[].review.summary.
Cross-link the emerging pattern into
projects/OVER_ANNOTATION_PATTERNS.md (protostome stem-cell term applied
wholesale to vertebrate genes) once 3–4 reviews are done.

Priority

Medium. 42 annotations / ~33 distinct genes — larger than the typical
obsoletion-tracking set, with a crisp, repeating conceptual error (a
protostome neural-stem-cell term applied 100% to vertebrate genes). Two genes
(ASCL1, FGFR2) are already in the repo, and the cluster includes textbook
neurogenesis genes (LIS1, ASPM, NDE1, ASCL1, SHH) where the correct vertebrate
term is unambiguous, making this a strong, well-evidenced illustrative
batch. Upstream is still in discussion, so there is no deadline pressure — but
the annotation review is valuable input to that discussion now.

Status

2026-05-18 — Project file created. Tracking upstream
geneontology/go-annotation#6393 (last upstream activity 2026-05-07). The
attached QuickGO export (42 rows, 6 GO terms, 100% vertebrate) was fetched
and parsed for scope. GO:0007405 and the proposed replacement GO:0061351
verified in OLS. No reviews started yet; ASCL1 and FGFR2 are the only
affected genes currently in the repo (both genes/human/).