Current FlyBase lists Q9VVS1 as the 302-residue PA/PC product; the separate PB product is 173 residues. The prediction-time sequence is absent from the release, so the current exact accession is the evaluated product. Dic4-flybase.txt and frozen UniProt JSON preserve these records.
PMID:21130726(https://pubmed.ncbi.nlm.nih.gov/21130726/), “No transport activity was observed for DmDic4p.” The abstract also directly reports mitochondrial localization for all expressed Dic proteins. Full assay details are unavailable in the cached abstract. Retain both NOT dicarboxylate annotations. The sulfate, thiosulfate and phosphate IBAs require reconciliation with the target-specific biochemical result; this is a conflict with experimental evidence, not an objection to the number of PAINT donors.
ThPP was not identified in the accessible abstract as a tested substrate. Therefore the negative carrier assay cannot by itself refute the ProtNLM ThPP claim. No specific molecular function is established sufficiently to populate core_functions. The original prediction is a function paragraph only, reviewed separately in the linked function-description review.
The completed Falcon synthesis was inspected and its key carrier review was fetched as PMID:32842667. The full review describes failed homo- and hetero-exchange assays and possible changes at substrate-contact points CP1/CP2. Its summary table lists subfamily substrates that the prose does not establish for Dic4. These observations strengthen the specific substrate caveat. The primary PMID:21130726 abstract directly reports mitochondrial localization, which is more specific evidence than the provider report’s statement that direct localization was not found. No inspected source identifies a ThPP assay.
Restored broad membrane and transport assertions. Retained the two experimental NOT dicarboxylate rows and unresolved substrate-specific IBAs after critical incorporation of the existing OpenScientist report.
The existing OpenScientist report is reused critically, with exact limitation excerpts in the reviews. Its absence-of-positive-experiment argument is not adopted as loss evidence. General transport is retained as an ancestral carrier inference, without assigning a measured Dic4 substrate. Primary NOT annotations remain negated and accepted.
PAINT: {'family': 'PTHR45618', 'nodes': ['PTN000756618', 'PTN002515516'], 'finding': 'Current specific anion-transport and general carrier assertions were inspected separately; negative tests of one panel do not remove every possible transport function.'}
All 12 rows were assessed, including experimental, electronic, negated and old proposed entries. Source annotation fields and row counts remain unchanged. Remaining questions are recorded in projects/IBA_REVIEW/rereview-2026-09-20/receptor-and-lipid-claims.yaml; coordinated reports will be assessed critically when available.
ProtNLM2 output and the ProtNLM-derived OpenScientist ThPP report were not used as evidence in this pass.
Identity. Q9VVS1 is the 302-residue canonical Dic4 product (FlyBase PA/PC; the 173-residue PB product is a
separate shorter isoform). It carries three Mito_carr repeats (Pfam PF00153 x3, PROSITE SOLCAR x3), so it is the
full-length carrier, not a fragment.
Phylogeny. PANTHER 19.0 places Dic4 in PTHR45618:SF13 MITOCHONDRIAL DICARBOXYLATE CARRIER with Dic1, Dic2,
Dic3, CG16736, CG6893, human SLC25A10, mouse Slc25a10 and yeast DIC1. A new NJ tree
(Dic4-bioinformatics/RESULTS-rereview.md) gives Dic4 the longest branch among the fly Dic paralogs; the order of
Dic1-Dic4 is unresolved (bootstraps <62%). SLC25A10 pairs with Dic1, so it is a co-ortholog of the whole fly
expansion. PAINT: the IBD node PTN000756618 carries dicarboxylate, phosphate, sulfate and thiosulfate transport. The
fly NOT node PTN000640888 removes malate/succinate/oxaloacetate terms from Dic4 (QuickGO IBA rows for Dic4 lack
them; Dic3 lacks only malate). This is a deliberate curator decision, and the phosphate/sulfur-anion terms were
left in place.
Literature. PMID:32842667 (full text) states that DmDic4p "was unable to catalyze either homo- or
hetero-exchanges, after its reconstitution into liposomes" and that its "inability to transport dicarboxylates or
phosphate could arise from a loss of some conserved residues located on CP1 and CP2", in the same assay system in
which Dic1 and Dic3 were active. Proteins were refolded from inclusion bodies, and PMID:21130726 is abstract-only.
Changes.
- Phosphate/sulfate/thiosulfate IBAs (5 rows): UNDECIDED -> MARK_AS_OVER_ANNOTATED. The reason is target-specific
loss evidence (negative reconstitution of phosphate and other exchanges, CP1/CP2 substitutions, fast-evolving
paralog), not the donor count. They were not removed because of the refolding caveat and the untested substrates.
- membrane (IEA): ACCEPT -> MODIFY to mitochondrial inner membrane (rule for correct but general terms).
- transmembrane transport (IEA InterPro): ACCEPT -> UNDECIDED. No substrate is known and the only direct test was
negative.
- NOT dicarboxylate MF/BP, mitochondrion IDA, inner membrane IBA/TAS: ACCEPT, kept.
- Description rewritten; core_functions now gives only the inner-membrane location, with no molecular function
asserted.
- Relative for next review: human SLC25A10 (SF13 co-ortholog; review present with no PENDING rows, status
INITIALIZED). Fly Dic1 is the characterised paralog (review DRAFT).
- Modules: Dic4 is not in any module. The related modules are modules/mitochondrial_dicarboxylate_tricarboxylate_carriers.yaml
and modules/dmel_sulfide_oxidation.yaml (Dic1 only).