PEX2 (P28328) curation notes

2026-09-04 — Finishing pass (PAINT no-IBA project)

Completed the review pass over all existing annotations; status moved IN_PROGRESS → COMPLETE
(validation clean, zero warnings). Changes made:

  1. Resolved the GO:0000425 (pexophagy) action inconsistency. The IDA annotation from
    PMID:26344566 was UNDECIDED on the grounds that the paper "does not directly test PEX2".
    The cached full text shows it does: siRNA knockdown of the RING peroxins was performed and
    reduced PEX5 ubiquitination in the ATM/ROS pexophagy pathway
    PMID:26344566 and
    PMID:26344566.
    Changed to ACCEPT, consistent with the ACCEPT for the PMID:27597759 pexophagy IDA.

  2. Added structured propagation_review to the IBA REMOVE of GO:0016593 (Cdc73/Paf1
    complex).     The GOA WITH/FROM for this IBA is PANTHER:PTN008299004|UniProtKB:P28328,
    i.e. the phylogenetic assertion is seeded by PEX2's own IDA from PMID:18987311 — the
    parafibromin/PAF1 transcription-complex paper. That IDA is a homonym confusion between
    PEX2's historical synonym PAF1 (Peroxisome Assembly Factor 1
    PMID:1546315) and the unrelated PAF1/RNA-polymerase-II-associated factor.
    Classified root_cause: SOURCE_BAD, failure_mode: SOURCE_MISCITATION. The four
    PMID:18987311-based annotations (Cdc73/Paf1 complex, protein destabilization, and the two
    proliferation terms) all carry REMOVE for the same reason. Note this is not
    second-guessing an experimental annotation on full-text grounds: the miscitation is a gene
    identity error, evident from the paper itself.

  3. Added core_functions (previously absent — was a validation warning). Three functions:

  4. E3 ubiquitin ligase (GO:0061630) within the PEX2-PEX10-PEX12 ubiquitin ligase complex
    (GO:0000151) monoubiquitinating PEX5 for receptor recycling (GO:0016562) during matrix
    protein import, at the peroxisomal membrane
    PMID:35768507.
  5. Pexophagy trigger: PEX2-specific ubiquitination of peroxisomal membrane proteins during
    amino acid starvation PMID:27597759.

  6. Lipolysis regulation: ROS-stabilized PEX2 polyubiquitinates ATGL at K92 (K48-linked)
    for proteasomal degradation PMID:34903883.

  7. Added two action: NEW annotations so the core_functions terms are reflected in
    existing_annotations: GO:0000151 (ubiquitin ligase complex; IDA, PMID:35768507 — GOA has
    no complex-membership CC term for PEX2) and GO:0050995 (negative regulation of lipid
    catabolic process; IDA, PMID:34903883).

  8. Cited the deep research file (file:human/PEX2/PEX2-deep-research-falcon.md) in the
    references list and in the NEW GO:0000151 annotation — its synthesis of the 2024
    Kumar et al. and Pandey reviews genuinely informed the framing of the heterotrimeric
    ligase/retrotranslocon complex as PEX2's core context.

Final action tally: 41 ACCEPT, 9 KEEP_AS_NON_CORE, 5 REMOVE (1 IBA + 4 IDA/IMP, all from the
PAF1 homonym miscitation), 4 MARK_AS_OVER_ANNOTATED (bare protein binding IPIs), 2 NEW,
0 UNDECIDED, 0 PENDING.

Notable curation finding worth flagging upstream: PEX2 receives no IBA for its actual core
functions (peroxisomal membrane / peroxisome organization IBAs exist, but nothing for the E3
ligase activity or receptor recycling), while its only complex IBA (Cdc73/Paf1) is a
propagation of its own homonym-confused IDA. See interpro/panther/PTHR48178/PTHR48178-review.yaml.

2026-09-17 — PR #2956 review response