pmp20

Summary: IBA annotation for cytoplasmic localization. UniProt confirms cytoplasmic localization.

Reason: Cytoplasmic localization is confirmed by high-throughput localization data (PMID:16823372) and consistent with the UniProt annotation.

Summary: Over-annotated; no evidence places S. pombe Pmp20 in the peroxisome, and PAINT now withdraws the inference.

Reason: Two independent lines point the same way. The target's experimental locations are cytoplasm, cytosol and nucleus. Current PAINT places an IRD (NOT) for this term at PTN000824099 (2026-07-15), a fungal Pmp20 clade that includes S. pombe pmp20, so the IBA's ancestral support at PTN000046537 is withdrawn and the row remains only in the cached GOA file. The C-terminal SSL is not a consensus PTS1, but that is weak evidence on its own: human PRDX5 ends in the equally non-consensus SQL and is imported by PEX5, so the tripeptide does not exclude import. Under the root-organelle policy (projects/ORGANELLES.md, rule 3) a synthesis row whose organelle claim is not supported is MARK_AS_OVER_ANNOTATED. It is not REMOVE, because no assay excludes a minor or condition-dependent peroxisomal pool; the name PMP20 is not localization evidence.

Summary: Retain rejection of inferred hydrogen peroxide catabolism based on target experiments.

Reason: The complete PMID:20356456 paper reports no Trx-coupled peroxide reduction, no GSH-dependent activity (data not shown), and no protection in the thiol-driven metal-catalyzed oxidation assay. These are target-specific negative biochemical findings, reinforced by the existing experimental NOT annotations, rather than absence of a second cysteine alone. The two OpenScientist reports are corrected on GSH assay coverage. Untested partners or conditions are not exhaustively excluded, but no positive target evidence offsets these contradictory experiments; the observed holdase activity does not itself detoxify oxidants. Current PAINT places an IRD (NOT) for this term at PTN000824099 (2026-08-28), a fungal Pmp20 clade that includes S. pombe pmp20, so this IBA is no longer propagated by PAINT and remains only in the cached GOA file. The IRD agrees with this verdict. The experimental seeds of the blocked PTN001625584 IBD (A. fumigatus Asp f3 O43099, A. nidulans prxA Q5ASN8, human PRDX5) lie in other subfamilies (SF30, SF16), outside the PTN000824099 clade (SF39), so the IRD does not deny the activity to any assayed protein.

Summary: Retain the broad oxidative-stress response inference.

Reason: A cellular response to oxidative stress need not be peroxide reduction. Pmp20 retains an experimentally demonstrated protein-protective holdase activity, compatible with stress protection despite loss of detectable canonical peroxidase activity. No target evidence excludes this broader inherited response. Acceptance is an evolutionary process inference; the heat-aggregation assay alone is not claimed to be a direct in-vivo oxidative-stress experiment. Retain the inherited oxidative-stress process as non-core relative to the directly demonstrated holdase activity; the latter was measured under heat aggregation and is not itself an oxidative-stress assay. Current PAINT places an IRD (NOT) for this term at PTN000824099 (2026-07-15), a fungal Pmp20 clade that includes S. pombe pmp20, so this IBA is no longer propagated by PAINT and remains only in the cached GOA file. This review keeps the broader stress response on the strength of the holdase activity, which the IRD does not address; the disagreement with PAINT is recorded rather than resolved.

Summary: Redox homeostasis is broader than direct peroxide catalysis.

Reason: Target biochemical experiments contradict inferred peroxide detoxification, but GO:0045454 also permits regulatory or protective contributions to redox balance. Loss of a catalytic route does not establish absence of every process role. A holdase activity alone is not sufficient proof of redox maintenance either. The ancestral process inference remains unresolved, with the existing reports identifying the same mechanism/physiology gap. Current PAINT places an IRD (NOT) for this term at PTN000824099 (2026-07-15), a fungal Pmp20 clade that includes S. pombe pmp20, so this IBA is no longer propagated by PAINT and remains only in the cached GOA file. That is a PAINT judgement of loss for the clade, not target-specific evidence, so the verdict stays UNDECIDED.

Summary: Over-annotated; no evidence places S. pombe Pmp20 in the mitochondrion, and PAINT now withdraws the inference.

Reason: Three independent lines now point the same way. The target's experimental locations are cytoplasm, cytosol and nucleus. Current PAINT places an IRD (NOT) for this term at PTN000824099 (2026-07-15), a fungal Pmp20 clade that includes S. pombe pmp20, so the IBA's ancestral support at PTN000046537 is withdrawn and the row remains only in the cached GOA file. The protein also lacks the mitochondrial transit peptide that targets its mammalian family members: human PRDX5/P30044 has a residues 1–52 transit peptide, whereas canonical Pmp20 is 156 residues and lacks that extension. Under the root-organelle policy (projects/ORGANELLES.md, rule 3) a synthesis row whose organelle claim is not supported is MARK_AS_OVER_ANNOTATED. It is not REMOVE, because no assay excludes an alternative import route or a minor mitochondrial pool.

Summary: IEA annotation for nuclear localization from UniProt subcellular location mapping.

Reason: Nuclear localization is confirmed by high-throughput data (PMID:16823372) and consistent with the HDA annotation below.

Summary: IEA annotation duplicating IBA for cytoplasmic localization.

Reason: Consistent with the IBA annotation and HDA data from PMID:16823372.

Summary: Retain the broad oxidative-stress response inference.

Reason: A cellular response to oxidative stress need not be peroxide reduction. Pmp20 retains an experimentally demonstrated protein-protective holdase activity, compatible with stress protection despite loss of detectable canonical peroxidase activity. No target evidence excludes this broader inherited response. Acceptance is an evolutionary process inference; the heat-aggregation assay alone is not claimed to be a direct in-vivo oxidative-stress experiment. Retain the inherited oxidative-stress process as non-core relative to the directly demonstrated holdase activity; the latter was measured under heat aggregation and is not itself an oxidative-stress assay.

Summary: HDA annotation for nuclear localization from high-throughput localization study.

Reason: Nuclear localization determined by systematic ORFeome-based protein localization study in S. pombe (PMID:16823372). Consistent with UniProt annotation.

Summary: HDA annotation for cytosolic localization from high-throughput study.

Reason: Cytosolic localization determined by systematic protein localization study (PMID:16823372). Consistent with the cytoplasm IBA annotation.

Summary: Retain rejection of inferred thioredoxin-dependent peroxidase activity based on target experiments.

Reason: The complete PMID:20356456 paper reports no Trx-coupled peroxide reduction, no GSH-dependent activity (data not shown), and no protection in the thiol-driven metal-catalyzed oxidation assay. These are target-specific negative biochemical findings, reinforced by the existing experimental NOT annotations, rather than absence of a second cysteine alone. The two OpenScientist reports are corrected on GSH assay coverage. Untested partners or conditions are not exhaustively excluded, but no positive target evidence offsets these contradictory experiments; the observed holdase activity does not itself detoxify oxidants. Current PAINT places an IRD (NOT) for this term at PTN000824099 (2026-08-28), a fungal Pmp20 clade that includes S. pombe pmp20, so this IBA is no longer propagated by PAINT and remains only in the cached GOA file. The IRD agrees with this verdict. The experimental seeds of the blocked PTN001625584 IBD (A. fumigatus Asp f3 O43099, A. nidulans prxA Q5ASN8, human PRDX5) lie in other subfamilies (SF30, SF16), outside the PTN000824099 clade (SF39), so the IRD does not deny the activity to any assayed protein.

Summary: Broader oxidoreductase catalysis remains unresolved after the negative peroxidase assays.

Reason: The target has a conserved Prx5-like peroxidatic region, but direct Trx- and GSH-dependent peroxidase assays were negative. This materially argues against the inferred canonical reaction without experimentally excluding every redox substrate/partner. The missing resolving cysteine alone is not proof of universal inactivity (single-Cys Candida Pmp20 is active), and AlphaFold geometry is not proof of turnover. Retain uncertainty for this much broader MF, with no new positive catalytic assertion.

Summary: Retain rejection of inferred cellular oxidant detoxification based on target experiments.

Reason: The complete PMID:20356456 paper reports no Trx-coupled peroxide reduction, no GSH-dependent activity (data not shown), and no protection in the thiol-driven metal-catalyzed oxidation assay. These are target-specific negative biochemical findings, reinforced by the existing experimental NOT annotations, rather than absence of a second cysteine alone. The two OpenScientist reports are corrected on GSH assay coverage. Untested partners or conditions are not exhaustively excluded, but no positive target evidence offsets these contradictory experiments; the observed holdase activity does not itself detoxify oxidants.

Summary: Retain the experimentally supported NOT annotation.

Reason: The primary negative Trx-dependent and GSH-dependent peroxidase results and negative oxidative protection assay support the curated NOT assertion. The GSH result is reported as data not shown. Acceptance does not claim that every substrate, external partner or physiological condition has been excluded; a contrary target assay would require reassessment.

Summary: Pmp20 has directly measured weak holdase activity.

Reason: Primary PMID:20356456 shows concentration-dependent protection of citrate synthase against thermal aggregation, approximately 30% at a 16:1 Pmp20:substrate ratio, with loss of protection after heat pretreatment. This supports holdase rather than active refolding. Physiological clients and the in-vivo contribution remain unestablished; the second provider report incorrectly frames this already measured activity as only a proposal.

Summary: The systematic ORFeome localization study observed a cytoplasmic pool of Pmp20. This is consistent with its cytosol HDA and cytoplasm IBA annotations.

Reason: Retain the experimental cytoplasmic localization; it independently corroborates the more specific cytosolic pool.

Summary: Retain the experimentally supported NOT annotation.

Reason: The primary negative Trx-dependent and GSH-dependent peroxidase results and negative oxidative protection assay support the curated NOT assertion. The GSH result is reported as data not shown. Acceptance does not claim that every substrate, external partner or physiological condition has been excluded; a contrary target assay would require reassessment.