The review assesses all 35 original annotation contexts and 29 references. The evidence archive preserves the incoming records, source identities, sequence and ancestry checks. Four annotations remain uncertain; two cytoskeletal interactions are retained as non-core.
Three short excerpts appear once each in reference findings. Complete originals remain unavailable for PMID:12767933, PMID:19285977 and PMID:10610709. For PMID:19285977, the abstract preserves its alternative association/assay wording; partner-specific methods remain inaccessible.
The hearing assessment weighs the human evidence and both mouse alleles, rather than using the cloning paper alone. Selected original material was read for Teubner et al. and Boulay et al.. The latter normal cache contains full text; the former remains abstract-only. Human channel composition is supported by the Yum et al. original, including its construct and assay sections; that normal cache is abstract-only. External reading and cache availability are separate.
For PMID:26786210, selected author-manuscript pages and Figure 3B were inspected. Its Gjb6 label is present, but does not resolve the developmental role. For PMID:30280653, the selected publisher astrocyte section was read; underlying experiments were not independently inspected. Its cache is labelled full text but contains introductory material. Clinical records lacking an abstract remain UNVERIFIED.
The ten positive HuRI pairs remain recorded. Three assay projections per pair are not independent physiological validations. Generic binding does not identify a specific GJB6 function. Historical clone sequence equivalence and isoform exclusivity are not established.
The four PAINT assertions were checked against the actual tree, alignment and ancestral placements. Target self-evidence is legitimate. Current human O95452 is a 261-residue product with no declared natural alternative; mutant constructs and fusion proteins are separate experimental entities.
QuickGO’s is_a/part_of ancestor closure for GO:1990349 does not contain GO:0055085. Both existing transport rows are retained: gap-junction transport concerns intercellular exchange, and transmembrane transport does not require an unapposed hemichannel. The existing electrical-coupling process is now explicit in the channel core; no NEW annotation is proposed. Reactome’s retained overview supplies pathway context, without transferring its Cx43-specific claims to GJB6.
All 21 previously frozen normal sources remain unchanged; three missing publications were acquired through the native helper. Existing history is preserved, with a separate MODIFY record for this follow-up. The archive distinguishes prior validation from checks of this revision.