Deep research: not run. In this environment falcon times out and perplexity-lite is not installed. The review
uses cached publications (just fetch-gene-pmids human CEP290) and the C. elegans TZ module literature.
This dated section supersedes the action recommendations above while preserving the earlier journal verbatim. The99 original GOA assertions and all qualifiers, partners, products and the explicit NOT minus-end-binding assertion remain unchanged. A genuine isolated normal intake reproduced the99 source objects. The prior genuine Falcon file is preserved; the current Falcon attempt timed out and the explicit Perplexity fallback failed for quota, so this audit is manual rather than a fabricated provider report.
The ClinGen standing binding instruction retains supported generic interactions as non-core. All30 pairs now have target-specific primary association evidence or exact positive interaction records. The bounded source extract carries literal accessions, PMID joins, interaction identifiers, methods, hosts, tags/fragments, expansion and negative fields, with query/source hashes. HuRI and2014 Y2H records use CEP290941–1104; they do not validate endogenous full-length binding. BioPlex and centrosome-network co-complex records are distinguished from proximity and purified binary binding.
The lipid and microtubule cores retain the exact primary anchors already recorded above and short budgeted YAML excerpts. Full PMID:24051377 Methods establish human codon-optimized constructs: purified1–580 binds mixed PS/cholesterol liposomes, whereas1695–1966 binds prepolymerized bovine microtubules. The257–292 helix is predicted, not an independently demonstrated essential interface. Native full-length simultaneous membrane–microtubule bridging remains a model. PMID:23644468’s cellular CEP162 dependence and failure of overexpressed full-length CEP290 to decorate microtubules address a different context from the later purified-domain positive assay; neither entire study is declared false. The rd16 multiresidue deletion is not an exclusive causal dissection of one interface.
Only the older authored NEW microtubule-binding annotation is withdrawn: official positive GO:0005515 ancestry already covers it under the explicit hierarchy rule. Its complete prior object is pinned by immutable source URL and hash in the extract, and its supported activity remains in the core. The NOT GO:0051011 assertion is preserved separately and does not provide positive coverage. The lipid-binding NEW is retained because lipid binding is not in the positive protein-binding hierarchy. No NEW biological process is added.
The mouse donor for protein transport includes PMID:16632484; retinal localization defects and human Rab8a/BBSome assays support the more precise existing refinement to protein localization to cilium. Mouse Tctn1-LAP association/co-fractionation in PMID:21725307 supports contextual MKS-complex membership, while the same source’s nonreproducible tagged co-IP limits an obligate binary/stoichiometric interpretation. The detailed MKS-5/CEP-290 assembly hierarchy in PMID:26982032 is explicitly a C. elegans experiment.
Recovered PMID:16682973 separates human construct reporter/fractionation experiments, bovine retinal co-IP, IMCD3/COS7 imaging and zebrafish developmental perturbation. The ATF4 reporter effect is retained as non-core, not a direct DNA-binding function. PMID:16682973 The developmental transfers retain their original fish/phylogenetic scope. Human centrosomal imaging independently supports the older proteomics locations. PMID:24648492
Both neutrophil locations are now UNDECIDED. Official Reactome6798749 directly includes human O15078 chain1–2479 in specific-granule-lumen and extracellular physical entities. The source PMID:23650620 measures pooled human neutrophil fractions by MS and discusses co-sedimenting contamination; its original target supplement was not recovered, and neither luminal topology nor CEP290 release was verified. An absent signal peptide does not invalidate that positive curated pedigree. The source extract records exact entity/set links and the remaining limitation.
Actual primary access is reported per reference. Canonical caches are preserved even when abstract-only or incomplete; genuine external full text supplied selected Methods/Results for several binding papers. The PMID:26386044 source-specific CEP290 IDA linkage remains unresolved despite the independently supported assembly function. No title-only misattribution claim is made.
Quote accounting distinguishes the exact inherited journal prefix from newly authored YAML and this append. New24051377 excerpts total20 words across the two cores and lipid NEW; the inherited prefix contains22 more. Thus the whole historical file total for that source exceeds25, while newly authored quotations remain at or below25 per source. Other inherited anchors are preserved without repeating them wholesale. Machine identifiers and non-literary structured fields are not publication prose. No machine publication cache, GOA, UniProt, historical provider report or prior history record was edited.
The preceding journal is retained byte-for-byte. This entry supersedes its compressed audit wording where clarified below; it does not revise the historical record or fetched sources. All 99 machine annotation objects, the retained lipid-binding proposal, the complete current reference identities and the empty product set remain unchanged. The only action change is ACCEPT for the explicitly negated minus-end-binding assertion. Nocodazole-resistant centrosomal localization is not a test of minus-end specificity.
The requested new microtubule-binding row remains withdrawn. The published instruction at the reviewed head explicitly rejects a NEW term that is an ancestor or descendant of another proposal or of an existing term. GO:0008017 is a descendant of existing positive GO:0005515. The supported microtubule-binding core remains, with its direct human-domain assay anchor. This follows the explicit authoring instruction; it does not mean generic binding is biologically as informative as microtubule binding. The core-coverage advisory does not override that instruction. The lipid-binding proposal has no such existing hierarchy overlap. No protein-partner row is mechanically reassigned to microtubule binding.
Short primary anchors now accompany the relevant binding, fractionation, ciliogenesis, localization and transition-zone claims. Selected original full Results and legends were read for PMIDs 22441691, 22797915, 24550735, 36606424 and 16682973. The source artifact records URLs, locators and hashes. Figure 1C of PMID:22441691 distinguishes a positive CP110 immunoprecipitate from the negative Neurl4 comparison. PMID:36606424 separates the human HEK293T biochemical interaction from mouse-cell proximity profiling. Challenge pages were not treated as full text. Neither source access nor these assays establish purified binary interfaces.
The thirteen ambiguous optional availability flags (including the analogous PMID:22441691 record) are omitted. The schema describes whether full text is unavailable without specifying cache-only scope; a validator suggestion addresses cache quote matching, and the audit helper only checks stale true flags. Each reference assessment therefore states the selected cache and actual external access separately. External-only quotations use supporting_text_fulltext; the original publication caches are unchanged. Full-text access does not imply every panel or Methods section was inspected.
The repeated Reactome cytosol decisions share a centrosome-entity provenance; they are not independent experiments or evidence that CEP290 performs each named event. Their summaries now identify CEP290's entity context, and the independent fractionation observation appears once in its own annotation rather than being added repeatedly to event citations. The four shared IntAct complex identifiers now have participant-specific accession and source-line anchors, with spoke expansion explicit. PMID:24550735 and PMID:36606424 have MEDIUM relevance; MKS-module context is restored using PMID:28401750 without making CEP290 an obligate stable human MKS subunit. The description is a standalone biological summary, and current prose uses readable spacing. Historical journal spacing is preserved in the protected prefix.
The aggregate quotation limit is mandatory authoring discipline, not a repository validator rule. All current YAML publication quotations, including repeated excerpts and external-only anchors, were counted; the new notes append and source-artifact prose add no publication quotations. The pre-existing journal remains an inherited byte prefix and its older quotations are reported separately rather than reauthored. Current YAML totals (words per source): PMID:16682973 = 18, PMID:18694559 = 21, PMID:22441691 = 12, PMID:22797915 = 15, PMID:22863007 = 10, PMID:24051377 = 20, PMID:24421332 = 14, PMID:24550735 = 5, PMID:24648492 = 8, PMID:26982032 = 5, PMID:28235840 = 7, PMID:28401750 = 8, PMID:36606424 = 11.
This entry supersedes the two earlier rationales for withdrawing the microtubule-binding proposal. Reading the complete published instruction in context shows that its ancestor/descendant restriction belongs to the requirements for a NEW biological-process term. Applying that restriction to this molecular function was an overbroad interpretation. No policy text is changed.
The NEW GO:0008017 microtubule-binding assertion is restored, with IDA and PMID:24051377, alongside the existing lipid-binding proposal. The supporting experiment directly assays purified human CEP290 residues 1695–1966 with prepolymerized bovine microtubules. Cellular CEP162-dependent recruitment concerns a different experimental context; neither result establishes a reconstituted native full-length membrane–microtubule tether. The retained lipid proposal rests on its own purified human N-terminal liposome-binding assay. Both domain activities are supported on experimental grounds, with their distinct substrate and construct limits.
All 99 original machine source objects, the complete existing lipid proposal, other review actions, references, the empty product list and all cores remain unchanged. The source extract preserves the previous withdrawal as historical provenance and records its explicit supersession by this restoration. The four pair-specific IntAct anchors remain linked to their literal source-line hashes; those raw fields and hashes, rather than a composed display label, provide the independent record evidence. No source cache or interaction record is modified.
The entire prior notes file, including its earlier formatting, remains an exact byte prefix. This readable dated correction is the current account of the MF decision; the earlier compressed wording is historical. The restored YAML excerpt adds four words for PMID:24051377, bringing all current YAML quotations for that paper to 24 words including repeated anchors. This append and the source-artifact correction add no publication quotations. The inherited journal contains another 22 words from that paper, so its whole historical YAML-plus-notes total is 46; those inherited excerpts are separately accounted and are not restored into new authored fields.