FGF12/FHF1 is primarily a sodium-channel regulator. Human FGF12B binding and functional measurements in PMID:22705208(https://pubmed.ncbi.nlm.nih.gov/22705208/) support the channel mechanism independently of the rat FHF1B donor experiments in PMID:12401812(https://pubmed.ncbi.nlm.nih.gov/12401812/). The 2016 variant study links disturbed channel regulation to developmental and epileptic encephalopathy (PMID:27164707(https://pubmed.ncbi.nlm.nih.gov/27164707/); preserved UniProt DISEASE lines 131–137, DEE47/MIM:617166). Neither the human channel nor a human host cell makes the rat donor protein human. Gating direction depends on the tested channel, product and variant.
The review preserves all 49 imported source contexts, including 15 isoform-specific contexts, both NOT assertions, both natural products and all 14 initial reference objects' identifiers, titles and findings. Nineteen references include two additional UniProt sequence sources, two later functional papers and the preserved UniProt file. No NEW annotation or publication quotation is added. Generic interaction annotations remain non-core; the directly functional FGF12B–NaV1.5 interaction is refined to the already represented sodium-channel regulator activity.
Four positive PAINT IBD joins, actual target ancestry, the exact canonical MSA sequence, three ARBA rule matches and the InterPro mapping were checked. The full tree response is incomplete: the received target path is sufficient for these local ancestry joins, not for a whole-family reassessment. Neurogenesis remains a non-core inherited judgment. Expression, seizures and donor counts are not substituted for phylogenetic reasoning.
The two short-product receptor-negative contexts need reconciliation with later direct binding and receptor-signaling results (PMID:12815063(https://pubmed.ncbi.nlm.nih.gov/12815063/), PMID:32357892(https://pubmed.ncbi.nlm.nih.gov/32357892/)). Mitogenesis, receptor activation and survival are distinct endpoints. GO growth-factor activity requires growth or proliferation; receptor-dependent survival alone does not settle it. The old TAS rows retain source-specific uncertainty even where a later study supports a related gene-level signaling role. The inaccessible 1999 full paper is not declared erroneous: official U76381.2 resolves its FGF12b identity.
FGF12-dependent nucleolar protein association is positive evidence, but the productive process it serves remains unresolved. PMID:36411431(https://pubmed.ncbi.nlm.nih.gov/36411431/) and PMID:38468333(https://pubmed.ncbi.nlm.nih.gov/38468333/) support interactions/localization, without isolating an FGF12-dependent ribosome-output step. The 2015 NOLC1/TCOF1 platform work (PMID:26399832(https://pubmed.ncbi.nlm.nih.gov/26399832/)) does not itself identify a FGF12 neural-crest experiment in the inspected material. A scaffold can participate in a process without catalysis; uncertainty here concerns the process served, not a requirement for an enzyme.
The provider deep-research invocation failed before this review. These are manual curation notes, accompanied by the source guide; no provider-branded research file was fabricated. Source caches are unchanged. Unresolved evidence is retained explicitly rather than resolved by prediction.
External review identified seven omitted negative IRD assertions at the FGF12 ancestor PTN000160077. The complete family table was then read, with both gains and losses considered. Growth-factor activity from the electronic family mapping and the fully read 1996 hypothesis is now MARK_AS_OVER_ANNOTATED; the inaccessible 1999 source stays UNDECIDED. Cell-cell signaling moves from non-core to UNDECIDED because the inspected studies do not establish a physiological sender-to-receiver chain. The later direct receptor-binding/signaling measurements and tagged stress-release results remain positive evidence requiring reconciliation with the receptor NOT and extracellular assertions; they are not erased by counting curated records. The IRD export points to the positive ancestor, without identifying an independent negative experiment. See the source guide for exact locators and the evidence package for both positive and negative joins.
After this follow-up the 49 contexts comprise 7 ACCEPT, 25 KEEP_AS_NON_CORE, 1 MODIFY, 3 MARK_AS_OVER_ANNOTATED and 13 UNDECIDED. Core function, products, original reference identities and every imported source field are unchanged. No new annotation or publication quotation is added.