Issue: Validation failure due to 2 GO terms having obsolete labels.
Root Cause: Two GO terms in the review had labels prefixed with "obsolete" which no longer match the current ontology:
- GO:0098660: had "obsolete inorganic ion transmembrane transport" → corrected to "inorganic ion transmembrane transport"
- GO:0043225: had "obsolete ATPase-coupled inorganic anion transmembrane transporter activity" → corrected to "ATPase-coupled inorganic anion transmembrane transporter activity"
Action Taken: Updated the term labels to remove the "obsolete" prefixes to match the current GO ontology.
Validation Status: After fixing the obsolete labels, the gene now passes validation with only warnings about PENDING annotations.
Note: These terms were likely marked as obsolete at some point but have since been reinstated in the GO ontology, requiring the label correction. CFTR is the cystic fibrosis transmembrane conductance regulator, a chloride channel whose dysfunction causes cystic fibrosis.
This audit preserves all 203 original machine annotation objects and all 107 original reference identifier/title pairs. The old review grouped 448 GOA lines without the newer WITH/qualifier projection. Genuine isolated normal intake now produces 539 source objects from 540 raw data rows (the fetch service printed 541); these are retained as an explicit source comparison, not silently substituted for the historical assertions. Three current UniProt alternative-product descriptors are appended from the genuine normal seed; absence in the old review did not establish absence of isoforms, and no original unscoped assay is reassigned to a product.
The current activity model separates ATP-gated chloride conduction, bicarbonate conduction through the same pore, and intrinsic ATP hydrolysis supporting conformational cycling. The original ATPase report provides PMID:8910473. A strict ATP-turnover-to-gating stoichiometry is not assumed: the Walker-mutant study reports nonidentical catalytic and gating effects (PMID:9931011). Official EC 5.6.1.6 belongs to polypeptide conformational control; therefore the old isomerase mapping is valid at broad EC scope and is refined to measured ATP hydrolysis. By contrast, current GO:0140359 requires primary active transport and does not describe CFTR passive permeation. The old GO:0043225 source ID is retained and its pump-like mechanism is refined to the supported ATP-gated channel. Current official responses also mark GO:0043225 and GO:0098660 obsolete, superseding the earlier note that described them as reinstated.
Human CFTR bicarbonate permeation was measured in BHK inside-out patches (PMID:19019741). The paper reports PMID:19019741. The external-chloride-dependent dynamic switch instead belongs to the oocyte study PMID:15010471. Four original transporter assertions are refined to bicarbonate channel activity; the redundant previously authored bicarbonate NEW is withdrawn with its entire old object preserved in the source artifact. Two further old authored processes are withdrawn: broad pH regulation is already covered by existing pH-elevation assertions, and mucus hydration/clearance does not by itself establish regulated mucus release. The human airway rescue paper PMID:19621064 really measured benzamil-sensitive sodium transport as well as chloride and fluid endpoints; its positive sodium result is acknowledged, while the unresolved performed CFTR-to-ENaC step does not justify retaining the separate authored regulatory-process proposal. No source annotation is deleted by these four authored withdrawals.
The basolateral PAINT location is retained non-core in light of human sweat-duct staining near the basolateral membrane (PMID:1720311), while apical remains the main channel compartment. Full rat Sertoli Methods/Results support a nuclear immunoreactive processed-product context with peptide-blocking, independent-antiserum and nuclear-fraction controls (PMID:29453757). The approximately 85-kDa species is not evidence that intact human CFTR acts as a nuclear channel, and its proposed gene-regulatory activity remains untested. Cytosol annotations are justified by reported retrotranslocation/deglycosylated intermediates and scoped imaging, not merely by intracellular domains. Mouse cholesterol phenotypes are real (PMID:17085523), but the source-specific performed step versus downstream homeostasis remains unresolved; the original transfers are UNDECIDED rather than discarded as evidence-free.
Exact source-specific interaction evidence is in CFTR-source-evidence.json. The old generic groups correspond to 277 raw GOA partner rows. Positive co-IP and fragment/domain assays are retained with host, variant, tag and assay limits; they do not confer partner enzymatic functions on CFTR. Specific PDZ contacts and Sec61-complex association can refine the corresponding grouped assertion where the whole source scope supports it. PMID:12369822's CFTR partner is NHERF2/E3KARP, not the DRA protein foregrounded in its title; PMID:21455491's preserved pair is USP10, not Cif. Indexed original PMID:19465887 Figure 5 Results resolve GFP-F508del CFTR/USP19 co-IP despite incomplete normal-cache Results. The endosomal study PMID:18570918 concerns SNARE proteins, not RME-1/EHD.
Independent target-table review resolves all 31 original PMID:26618866 partners, including treatment-specific F508del rescue networks, and all 94 original PMID:35156780 partners as actual MaMTH positives. The latter includes 32 WT-only, 44 F508del-only and 18 shared detections, with mixed within-paper retest/co-IP confirmation. Neither source is represented as universal native binary binding. PMID:36012204 instead supplies proximity labeling: all 148 matched records (66 of 70 partners) have proximity method/type, and the primary explicitly permits noninteracting nearby proteins. Its grouped binding assertion remains UNDECIDED because the assay cannot resolve physical contact; the four unmatched partners are not negative evidence.
Reactome location rows preserve their own event/entity provenance. They are not separate localization experiments, and recruitment, ubiquitination, lipid hydrolysis, uncoating or motor work performed by other components is not transferred to CFTR cargo. The dated correction above the historical pathway page supersedes its DRA/ENaC citation conflation and reversed ENaC arrow while preserving the old body and genuine provider outputs.
Research provenance: normal fetch and publication retrieval completed in an isolated workspace. A genuine bounded Falcon attempt timed out and the explicit fallback failed authorization/quota; no synthetic provider report was written. Existing provider/citation/raw/cache bytes remain protected. Each reference assessment and the source artifact distinguish cached abstracts from externally accessed primary text and record the applicable assay limits. Screen interactions and channel measurements are assessed separately at their experimental scope. Current quoted text is counted across YAML and this append, with inherited notes reported separately. Unresolved source/assay questions are intentionally visible rather than converted into unsupported certainty.
All 203 historical machine annotation objects, three product descriptors and 113 reference identifier/title pairs remain unchanged. The separate 539-object normal projection is explicitly deferred inside the review's evidence-reference finding; it is not a claim of complete coverage of current GOA. The old findings are preserved in the evidence artifact and twelve assessed statements are restored. Long inherited quotations are not mechanically repeated. Each former bare binding key now quotes an explicitly labeled authored evidence paragraph with its original literal records and assay limits, alongside selected short publication anchors. Such file excerpts are factual paraphrases, not purported primary-paper quotations.
The cholesterol decision now respects the original source qualifier: acts_upstream_of_or_within. Full original source access and that qualifier support non-core upstream effects for the synthesis and transport rows. Neither is an intrinsic sterol-enzyme or cholesterol-pore assertion. The separate broad localization mapping remains unresolved. These conclusions supersede the earlier paragraph's undifferentiated three-row uncertainty.
For PMID:29453757, PMID:26403673, PMID:19019741 and PMID:17085523, the machine caches remain abstract-only while actual external primary text was inspected. The optional ambiguous availability flags are omitted, with explicit URLs, hashes, source titles/section locators and scope in the evidence artifact. A short external-only anchor supports the rat nuclear-band observation. The detailed rat observation remains distinct from an intact human nuclear-channel mechanism. Unavailability is a reason for UNDECIDED, not its exclusive meaning.
The live pathway now has one main title and an accurate diagram. The miscited DRA/ENaC and ClC-3B trafficking claims are corrected inline, and the reversed inhibitory arrow is removed. The previous published page is archived unchanged in the source artifact. Sodium transport rescue remains real contextual evidence; it is not converted into a direct ENaC inhibition mechanism. The distinct ATPase core records a measured chemical reaction, rather than merely repeating ATP dependence of anion flux. The description again names cystic fibrosis and F508del. Spacing and reviewer-specific vocabulary were cleaned in current authored prose; the published notes above remain an exact chronological prefix.
Quotation totals are audited cumulatively across current YAML, the entire 2026 campaign notes and the current pathway, including repeated excerpts and external-only text. Only genuinely pre-campaign inherited quotations are separated; the original twelve findings and old page are preserved as clearly labeled historical provenance. The source-specific 25-word cap is an authoring limit, not a claimed repository requirement.
This section supersedes the cholesterol reclassification and blanket quote restoration in external commit fc8efbc8011a6dfed8ab773319d1f441f4d174f2, an ordinary child of 0b0095b78b7dc5ddffd054c8a68cb355d794293b. Its history record is retained, as are the complete published notes prefix, all source objects, products, reference identities, source artifact records and the corrected pathway. No earlier history is rewritten.
The preserved GOA explicitly uses acts_upstream_of_or_within for the three historical cholesterol/localization transfers. The actual mouse IMP records for cholesterol synthesis and transport use the same relationship. A missing qualifier field in an older YAML serializer is not evidence that the underlying assertion instead means direct participation. The selected full PMID:17085523 Results include measured de novo synthesis and correction of the cholesterol-distribution phenotype by CFTR expression; they are not limited to the abstract's prediction. These observations support the existing upstream scope of synthesis and transport as non-core effects. They neither assign CFTR a sterol-synthetic reaction nor make its pore a cholesterol transporter. Rows 26 and 27 therefore return to KEEP_AS_NON_CORE. Row 28 returns to UNDECIDED because the exact historical broad localization mapping remains unresolved; access to a paper does not resolve every provenance question.
The external commit added 104 publication supporting-text fields. Repeated long excerpts, title/date lines, general topic labels and mismatched target sentences are not interchangeable with source-specific evidence. In particular, the SLC26A9 immunoblot sentence does not establish CFTR localization, and locating USP10 does not by itself locate CFTR. Those added excerpts are removed while the independently supported assertions and source references remain. Useful positive target evidence is retained or replaced by shorter exact assay excerpts: fragment/chaperone and variant/partner association, 14-3-3 engagement, monomer ATPase activity, human beta-cell membrane colocalization, recycling-endosome location, TMEM16A coupling, myosin/Rab11 complex association, pancreatic apical localization, CAL binding and airway-surface-liquid rescue. The durable source artifact records each incoming anchor's disposition without duplicating literature quotations.
Every previously added meaningful file-evidence paragraph, assessed finding and source-access distinction remains intact. The corrected live pathway is unchanged. The quote audit counts all live YAML excerpts, every campaign-authored notes excerpt in the preserved prefix, and current external primary excerpts in the artifact, including repetitions. Pre-campaign inherited notes/findings are identified separately rather than silently recounted as new text. This appendix adds no literature quotation. The resulting bounded anchors preserve checkable positive evidence without exceeding 25 words from any primary source. No molecular function, core mechanism, partner identity or NEW assertion is introduced by this follow-up.
The current review retains bare links to the source-evidence artifact and removes 35 quotations of its own explanatory prose. The adjacent publication excerpts, source records and experimental access limits are unchanged. This supersedes the earlier notes describing authored evidence paragraphs as supporting-text quotations; those paragraphs remain available as interpretations, rather than independent evidence.
The three cholesterol/localization source assertions now explicitly carry qualifier: acts_upstream_of_or_within, matching preserved GOA lines 303, 306 and 309. This fills previously absent metadata; it changes neither the source relationship nor the existing decisions. Cholesterol synthesis and transport remain non-core upstream effects, and the broad localization mapping remains unresolved. The other original source fields, all 203 decisions, three product descriptors and 113 reference identities are retained. The separate current-GOA migration remains deferred as documented in the review.
Spacing and research-provenance wording in the campaign-authored notes are corrected. Previous curation-history records are preserved; a new record documents this edit. No primary quotation, molecular function, process assertion or product-specific activity is added.