This section supersedes the earlier mechanistic and annotation conclusions above while preserving their provenance. The baseline review compressed several distinct GOA interaction rows and proposed six additional assertions. The present audit restores all 31 deterministic source objects, including individual partners, original references, qualifiers, inference sources and partner isoform suffixes. The raw UniProt/GOA files, both provider research files, and nine original publication caches are unchanged. Six additional normal publication records were recovered in Source31 and imported without overwrite; the import receipt is separate from biological review completion.
PMID:23793029(https://pmc.ncbi.nlm.nih.gov/articles/PMC3786259/) supports assembly of the NEK8-INVS-NPHP3 module. PMID:24998259(https://pubmed.ncbi.nlm.nih.gov/24998259/) uses isolated human SAM domains and shows monomeric ANKS6-SAM interacting with ANKS3-SAM; it does not establish ANKS6 self-polymerization. The original assembly Results and relevant domain/construct Methods were read, without claiming complete original interaction-screen supplements.
The three ANKS3-partner source rows are refined to protein complex scaffold activity using that independent mechanistic evidence together with PMID:29290488(https://pmc.ncbi.nlm.nih.gov/articles/PMC6258031/) and PMID:37733651(https://pmc.ncbi.nlm.nih.gov/articles/PMC10513324/). Their actual recovered recruitment, fractionation/topology, KH pull-down, N-terminal deletion and RNA-recruitment Results were read. ANKS3 mediates ANKS6 recruitment to Bicc1; no direct ANKS6-BICC1 interaction is inferred. The newer clamp geometry is an AlphaFold model tested by biochemical perturbations, not a solved structure of the entire assembly. Human isolated SAM-domain constructs are explicit. Whole-protein ANKS6 expression vectors derive from earlier rat constructs; human HEK293T host cells do not change that provenance.
The four NEK8 and three NEK7 association rows are refined to protein kinase binding. PMID:32707033(https://pubmed.ncbi.nlm.nih.gov/32707033/) explicitly reports the NEK7-ANKS6 association in its main Results. It does not demonstrate NEK7 activation. The DCAF7 pair (P61962) and sixteen HuRI source pairs remain UNDECIDED because the accessed article narratives do not expose their exact pair-level construct and assay records or establish informative replacements. These are access/interpretation limits, not claims that an experiment is false. Original pair identifiers remain intact; no blanket kinase-binding substitution is made for nonkinase partners.
PMID:25599650(https://pubmed.ncbi.nlm.nih.gov/25599650/) explicitly uses rat Anks6 and mouse Nek8 expression constructs in human 293T cells. The actual kinase Results include alpha-casein phosphorylation, kinase-inactive and domain controls, and the unsuccessful attempt to obtain soluble purified proteins. Coexpression/immunoprecipitation supports kinase activation, but it is not purified direct human enzymology. The retained proposed human kinase-activator assertion therefore uses ISS and the rat Anks6 homolog UniProt P0C0T2, rather than treating the rodent assay as direct human activity; the published baseline used TAS. The conserved ANK-repeat interaction architecture supports the ortholog-based interpretation; ANKS6 is not an intrinsic kinase. This evidence-code refinement received independent prospective consultation.
The earlier claim that ANKS6 physically transports NEK8 into cilia is reversed relative to the study: NEK8 helps recruit ANKS6 to the inversin compartment. The actual localization/domain Results and human control/patient-fibroblast context in PMID:26967905(https://pubmed.ncbi.nlm.nih.gov/26967905/) support the location review. Four machine-sourced cytoplasm/cilium/inversin-compartment annotations are retained as ACCEPT. A broader location is not false because a specific compartment is also annotated. The PAINT node and ortholog-source identifiers are preserved; no independent reconstruction of the entire IBD tree is claimed.
The actual PMID:32994509(https://pmc.ncbi.nlm.nih.gov/articles/PMC7524721/) Results, Discussion, Cell lines Methods and complete FLASH Methods were read. The study explicitly expresses hANKS6 with a C-terminal 3xFLAG-HBH tag in inducible Flp-In T-REx 293 cells. UV-C crosslinking, successive His and streptavidin affinity steps, stringent washes and at least eight ANKS6 experiments support RNA binding, with ARL13B/BBS3 negative controls. A contextual NEW GO:0003723 RNA-binding assertion is justified; its physiological targets and molecular contribution remain unresolved, so it is not promoted to a separate core function.
The Results describe ANKS6 and BICC1 AGO2-mRNA binding similarly, whereas the Discussion calls ANKS6 binding weak. No affinity estimate is inferred from that discrepancy. ANKS6 alone did not localize to stress granules under the reported treatment; recruitment to TIA1-positive granules depended on BICC1 coexpression and stress. No constitutive stress-granule location, RNA-processing process or stress-granule assembly function is proposed.
The old molecular-adaptor proposal is redundant with the more informative scaffold refinement, and the separate NEW kinase-binding proposal duplicates the refined source rows. The old Wnt-regulation proposal lacks an established ANKS6-specific step: PMID:24610927(https://pubmed.ncbi.nlm.nih.gov/24610927/) distinguishes human variants/embryonic expression from increased beta-catenin in mutant rat tubules. Its complete actual abstract was read; although Source31 recovered the full body, no new complete-body reading is claimed.
The old positive-Hippo proposal miscited the NEK8 patient study and assigned the wrong regulatory direction. Actual mouse BMEL/Smcc YAP, reporter, endogenous co-IP and nuclear-fraction Results in PMID:32886109(https://pmc.ncbi.nlm.nih.gov/articles/PMC7733532/) connect ANKS6 to increased YAP abundance/output. PMID:37525964(https://pubmed.ncbi.nlm.nih.gov/37525964/) provides human patient bile-duct context; the normal cache is abstract-only and was read completely. These observations do not automatically justify a new opposite-sign process term. Likewise, renal developmental necessity is insufficient by itself to establish the direct process participation required for the old nephron-development proposal. No new BP is asserted, and no comparator or GO-CAM audit is falsely claimed complete.
The revised draft contains 31 preserved source rows plus two evidence-scoped NEW molecular-function assertions: four ACCEPT, ten MODIFY, seventeen UNDECIDED and two NEW. Its two core functions are complex organization and the conserved NEK8-activating role. Unread pair-level supplements and unresolved physiological RNA targets remain explicit. Final independent authored review, source/quote integrity, schema validation, rendering and scaffolded history checks are recorded after this draft.
Final audit checks passed: gene/reference/schema validation, HTML rendering and the scaffolded history validation. The one validation advisory concerns the deliberately unused provider research files: primary papers, with stated read limits, supply the annotation support. All 31 source objects remain intact. Independent authored semantic consultation found no material biological concern; final file hashes and quote checks are recorded separately.
The source review retains all 31 machine-sourced assertions, including their partners and ordinary qualifiers. The 17 unresolved DCAF7/HuRI interactions remain UNDECIDED. The user-supplied action definitions reserve REMOVE for assertions unlikely to be correct on combined evidence, and require uncertainty when relevant evidence is inaccessible. An uninformative term, an uninspected supplementary pair record, or a partner's usual compartment does not establish a false interaction. Partner names are now taken from the unchanged UniProt interaction record. The accessible article narrative is not described as inspection of every pair table.
The three ANKS3 rows no longer carry a scaffold replacement. PMID:24998259 directly establishes human isolated SAM-domain association and is retained as non-core; its two-domain experiment alone does not establish a multiprotein scaffold. The PMID:27173435 and PMID:33961781 screen pairs remain uncertain at their original source-specific assay level, although independent domain experiments make the association plausible. The monomeric isolated ANKS6 SAM result does not exclude full-length oligomerization through other regions or partners.
The published PMID:23793029 Figure 3b legend identifies “V5-tagged full-length rat Anks6”; the adjoining Results and Figure 3c-d describe its coassociation with INVS and NPHP3 and its requirement for the NEK8-NPHP3 co-precipitation signal.
This wording was inspected in the published primary article PDF, page 954, Figure 3 and adjoining Results, DOI:10.1038/ng.2681. The complete published Results and the Figure 3 legend, including the co-precipitation controls, were read as text; the PDF screenshot service failed, so no new pixel-level image assessment is claimed. The cached author manuscript lists both rat and human Anks6 cDNA clones and therefore could not alone establish the panel-specific species. Its bytes remain unchanged. The independent annotation consultant separately read the published Results and Figure 3/4 captions.
One separate NEW GO:0140378 protein complex scaffold activity uses ISS with rat Anks6 UniProtKB:P0C0T2, keeping the assembly result separate from the original human binding rows. The official term is a molecular function under structural molecule activity and requires a complex component that holds the assembly together. ANKS6-dependent recovery of NPHP3 with NEK8 supplies that mechanistic basis; it is not inferred solely from loss-of-function phenotypes or a binary interaction. No NEW biological process is proposed.
The seven NEK7/NEK8 refinements retain their decisions but now distinguish the original source from independent corroboration. PMID:26967905 Figure 3C actually tests NEK8-ANKS6 co-immunoprecipitation and p.T87A-dependent loss of association; its human NEK8 construct does not identify the ANKS6 construct species. PMID:32707033 explicitly reports ANKS6-NEK7 association in its main Results. PMID:26638075 remains abstract-only, while the relevant PMID:27173435 and PMID:33961781 pair tables remain uninspected. Kinase-partner refinement does not assert activation, a purified interface, or that an uninspected screen was mechanistically resolved.
Reference reviews now distinguish verified identity from unresolved pair-level support; the four unresolved screen references are UNVERIFIED for that scientific scope. Descriptions and core summaries state biology, while construct and assay limitations remain in the annotation reasons and these notes. The two previously proposed molecular functions, kinase activation and RNA binding, remain unchanged. Prior publication caches, raw/provider files and the initial history record are preserved.
The follow-up review identified a supported refinement for the PMID:24998259 human ANKS6–ANKS3 interaction. The original human sequences, native-gel association, mapped interface mutations and heterocomplex Results identify the ANKS3 SAM domain as the bound entity. The affinity and structure assays use the ANKS3 I36E variant to suppress homopolymerization while preserving the heterointerface. GO:0032093 SAM domain binding is the specific current molecular-function child of protein domain specific binding, with no listed children. The row is now MODIFY to that term; its original IPI source and partner stay intact. This does not establish universal SAM binding or the oligomeric state of full-length ANKS6. The independently supported scaffold activity stays separate.
The 19 unadjudicated pair-level interactions remain UNDECIDED under the user-supplied ActionEnum and incomplete-evidence rules. The skill’s general preference to remove uninformative protein-binding terms does not override those explicit instructions. No interaction is declared false from genericity or unavailable pair-level support. All other annotation decisions and core functions are unchanged.