The private review began from 55 authentic human QuickGO records and a genuine 53-row helper forecast. Both schema collisions and all four omitted extensions are preserved in the evidence map. Source-specific uncertainty remains explicit.
PMID:21172611(https://pubmed.ncbi.nlm.nih.gov/21172611/): “All GluA and TARP cDNAs were derived from human”. The following clause excludes rat GluA2; recombinant GluA4 and neuronal preparations remain distinct.
The draft uses primary-source and manually inspected database evidence. No provider research file was generated or impersonated. Native acquisition and genuine history are complete.
Genuine helpers reproduced the 55 incoming records and 53 schema contexts. All twenty reused source records remained byte-identical; four missing publications were acquired normally. Reference metadata and genuine product declarations were reconciled. The existing core quotation was moved to its directly supported experimental row without duplication. Gene, reference, status and history checks passed. The family cross-check covered 951 reviews with zero conflicts; thirteen unrelated adjudications and 232 missing-information cases remain listed in its output. Rendering is regenerated from the final record.
Human GluA4–CACNG8 currents establish receptor activity; auxiliary-subunit effects were measured separately from the study's rodent neuronal localization experiments. PMID:21172611(https://pubmed.ncbi.nlm.nih.gov/21172611/).
The 433-aa P48058-2 ends before the canonical membrane-channel region. Rat GluA1/GluA2 M4 experiments do not directly test this human GluA4 product. PMID:23739980(https://pubmed.ncbi.nlm.nih.gov/23739980/).
Family sequencing identified de novo variants in five unrelated individuals. Structural modeling suggested altered pore or intersubunit behavior; these predictions are not measured variant currents. PMID:29220673(https://pubmed.ncbi.nlm.nih.gov/29220673/).
The vascular study reports GRIA4-dependent protection and rescue. Its abstract supports the contextual phenotype without resolving which receptor-mediated step changes apoptosis. PMID:32151690(https://pubmed.ncbi.nlm.nih.gov/32151690/).
The preserved evidence file is an audit of source records and independent checks; it is no longer cited as biological support for the core. Original-access locators distinguish inspected material from normal-cache availability.
Recombinant channel kinetics do not establish synaptic sidedness. PMID:12670305(https://pubmed.ncbi.nlm.nih.gov/12670305/).
Gria4 knockdown affected cellular binding, but neuronal biochemical assays did not establish GluR4 binding. These conditions do not justify universal negation. PMID:20032460(https://pubmed.ncbi.nlm.nih.gov/20032460/).
The first-publication access limitations remain historical above; current assessment uses the newly recovered originals, whose URLs and locators are preserved in the evidence record. The two original precise-ECO reasons remain historical, with current reviews reached through their pointers.
Gene, reference, status and genuine EDIT-history checks passed. The reference validator checked two snippets and eleven titles, with five snippets skipped; the independent repository-rendering replay verified all seven YAML excerpts, including UniProt continuation-line normalization. Computed status remains COMPLETE with zero validation warnings; publication and review gates remain separate. Four structured source-defect rows were exported privately. All source records, incoming contexts and declared products were preserved. The HTML was regenerated from the final review.
The availability flags describe the immutable normal caches. External original-reading records are separate and carry public URLs, inspected sections and content hashes in the evidence file. The historical journal entries above are retained.
UniProt reports “M4 transmembrane segment mediates tetramerization” by similarity. This does not directly test P48058-2. UniProt P48058, DOMAIN.
Gene, reference, status and genuine EDIT-history checks passed. The repository section helper added two exact record sections and left two ambiguous matches unset. Reference validation checked three snippets and eleven titles, skipped five snippets and found no issues; the independent source-rendering peer verified all eight YAML excerpts. Computed status remains COMPLETE with zero validation warnings. Immutable sources and all incoming contexts, products and actions were preserved. The HTML was regenerated from the final review; publication and new-head gates remain separate.