Accession: Q7SAD4; original locus: NCU06296. Source method: ProtNLM2. Source version: live API snapshot 2026-09-09T03:00:51.831347+00:00.
The complete prediction record is preserved in NCU06296-protnlm-source.json; claim metadata are also preserved in NCU06296-protnlm-provenance.json. Original UniProt SL identifiers are retained; these are not substituted with GO terms.
Original text: Endoplasmic reticulum membrane
Original evidence:
[
{
"evidenceCode": "ECO:0008006",
"id": "ProtNLM2",
"properties": [
{
"key": "model_score",
"value": "0.72"
},
{
"key": "phmmer_accession",
"value": "P49109"
},
{
"key": "phmmer_score",
"value": "37.5"
}
],
"source": "Google"
}
]
Assessment: UNC (confidence score 1).
No localization experiment exists for NCU06296 or for any member of its PANTHER subfamily (PTHR23023:SF292, a Pezizomycotina clade in which every member is uncharacterised). The donor P49109 is mammalian FMO5, a co-ortholog of the whole fungal clade at speciation distance, and the phmmer score of 37.5 is modest. Mammalian FMOs reach the ER membrane through a hydrophobic C-terminal anchor; Q7SAD4 has no predicted anchor of that type, and its C-terminus is charged, so the mammalian mechanism of ER-membrane association does not transfer. Q7SAD4 is the only product of the gene and spans the full FMO-like domain, so this is not a truncation relative to a canonical product. Yeast FMO1, which sits peripherally on the cytoplasmic face of the ER (PMID:10077572), belongs to a different PANTHER family (PTHR43539) and does not ground transfer either. The absence of an anchor does not exclude peripheral ER association, so the claim is neither supported nor refuted.
Evidence:
COR denotes a biologically supported location newly specified relative to the frozen target annotation snapshot. LSP denotes a less precise location than an established target annotation. Neither category establishes model training-set membership.