PEX13 curation notes

2026-09-04 — PAINT no-IBA project finishing pass (AI-assisted)

First journal entry for this gene; the review was drafted with all actions assigned but no
core_functions block and no notes file. Work done in this session:

Authored `core_functions`

One core function, built around the SH3 domain rather than around localisation:

molecular_function: GO:0030674 protein-macromolecule adaptor activity — PEX13's
cytoplasmically oriented SH3 domain engages the PTS1 receptor PEX5 and also PEX14, so it
physically couples the cargo-loaded receptor to the rest of the translocon.
PMID:8858165
and PMID:9653144
contributes_to_molecular_function: GO:0008320 — see below.
directly_involved_in: GO:0016558 / GO:0016560 / GO:0016561;
locations: GO:0005778; in_complex: GO:1990429.

The loss-of-function phenotype is what makes the adaptor reading rather than a generic
"structural component" reading defensible: the defect is in receptor engagement, not in
membrane assembly. PMID:8858165

Material changes to existing annotations

GO:0008320 (IDA, PMID:28765278): MARK_AS_OVER_ANNOTATED → ACCEPT. The draft argued
that transmembrane transport is an emergent property of the whole DTM and so should not
be ascribed to PEX13. That argument is out of date. Ravindran et al. show that Pex13
itself phase-separates with Pex5-cargo, that import depends on the aromatic residues of
the Pex13 IDR, and that import coincides with transient Pex13/Pex14 focusing at the
membrane — i.e. PEX13 is conduit-forming, not a peripheral scaffold. PMID:37165185 The activity is still only exercised within a multi-subunit
translocon, so it sits in contributes_to_molecular_function, not molecular_function.
Also fixed the term label (GOA writes "transmembrane protein transporter activity"; the
ontology snapshot used by the term validator writes "protein transmembrane transporter
activity" — the latter is what validates).
GO:0005515 IPI rows for PMID:8858165 (PEX5) and PMID:9653144 (PEX14):
MARK_AS_OVER_ANNOTATED → MODIFY, with proposed_replacement_terms: GO:0030674. Project
guidance is to replace bare protein binding with an informative MF term for adapter
function rather than merely flagging it. The three PEX19-related protein-binding rows
(PMID:10704444, PMID:20531392, PMID:32296183) stay MARK_AS_OVER_ANNOTATED — there PEX13
is PEX19's cargo, which is not a molecular function of PEX13.
GO:0034614 cellular response to ROS (IDA, PMID:26344566): kept
MARK_AS_OVER_ANNOTATED, reasoning rewritten. The cached record for this paper is full
text (full_text_available: true) and a search of the entire body returns no occurrence
of PEX13; the ROS-responsive participants characterised are ATM, PEX5, the
PEX2/PEX10/PEX12 E3 and p62, with PEX1 and PEX14 as the pexophagy readouts. So the usual
caution — "the abstract foregrounds another gene but the full text may assay the target" —
does not apply here. Still not REMOVE, because supplementary material is outside the
cache and PEX13 loss is independently reported to raise peroxisomal ROS.
NEW GO:0016558: reason previously cited GO:0140888, which is not the term being
proposed and does not belong here; rewritten to state the actual rationale (GOA carries
the two sub-steps but not their parent, and the loss-of-function phenotype is failure of
import per se).

Deep research

Cited file:human/PEX13/PEX13-deep-research-falcon.md on the GO:0008320 and GO:0016561
rows and in core_functions, where it genuinely informed the decision — it is the source
that frames PEX13 as the conduit rather than a docking-only factor
["PEX13 contributes the core translocation conduit in the peroxisomal
docking/translocation module (DTM)"], which is exactly the point on which I reversed the
GO:0008320 call.

Notable