The genuine fetch returned 44 original QuickGO contexts and 43 seeded review rows. The POMGNT1-binding contexts 447030759 (UniProt) and 447030760 (IntAct) have the same seeded annotation fields and project to one row. Both assigning-database records remain in the immutable 44-row TSV. The original IntAct record has alternative ER/Golgi extensions that the deterministic TSV/seed format does not represent; the source evidence guide records them explicitly. All 43 seed source objects and their order remain unchanged. Stored decisions are 26 ACCEPT,14 KEEP_AS_NON_CORE,2 TOO_GENERAL and 1 UNDECIDED; original-context decisions are 26 ACCEPT,15 KEEP_AS_NON_CORE,2 TOO_GENERAL and 1 UNDECIDED. No NEW annotations or supporting-text quotations were added.
The core function is the first ribitol-phosphate transfer onto GalNAc C3 of the core M3 O-mannosyl glycan on alpha-dystroglycan. FKTN performs this transfer itself, before the FKRP and RXYLT1 steps; it is not merely required for a downstream phenotype. Original NMR/biochemistry and human loss/rescue support the chemistry PMID:26923585(https://pubmed.ncbi.nlm.nih.gov/26923585/), with independent donor-transfer evidence PMID:27194101(https://pubmed.ncbi.nlm.nih.gov/27194101/). GO:0016780 accurately describes substituted-phosphate transfer. The polymer-specific GO:0047356 was examined but not guessed as a replacement; the specificity question remains explicit. The normal substrate is phosphorylated core M3, while separate GalNAc–GlcNAc–pNp NMR acceptors lack mannose phosphate. No universal artificial-acceptor phosphate requirement is asserted.
Two PAINT IBAs were retained after matching original IBD nodes PTN001034988 (glycoprotein biosynthesis) and PTN000395622 (Golgi membrane) to the actual human lineage in the complete 115-node PTHR15407 tree. Target-self experimental evidence is legitimate grounding. Existing normal family files were preserved; no bulk leaf-GAF was acquired.
Human FKTN pTW322 expressed in HEK293T colocalizes with Giantin PMID:25279699(https://elifesciences.org/articles/03941). The companion mouse Fkrp construct is distinct. The original 26923585 Golgi evidence is mouse-liver fractionation/enzyme activity tracking a Golgi marker, not human immunofluorescence. Myc-fukutin is the cis-Golgi comparator in the scanned localization study PMID:15213246(https://pubmed.ncbi.nlm.nih.gov/15213246/); its endogenous rough-ER observations concern FKRP and are not transferred to FKTN. GO:0000139 part_of GO:0005794 part_of GO:0005737 makes the two cytoplasm rows TOO_GENERAL, without denying their source assignments or inferring soluble cytosolic catalysis.
The ER-specific IDA remains UNDECIDED because the actual immunogold study only tentatively assigns particles to ER and/or Golgi PMID:18808525(https://pmc.ncbi.nlm.nih.gov/articles/PMC2613979/). Independent Golgi evidence supports that organelle, but does not resolve native ER occupancy. Nuclear fractionation/immunogold in human cancer-cell models remains non-core with absorption/RNAi controls and the authors' residual specificity uncertainty. HeLa/ZR-75-1 at 20 nM F697/five days showed increased Ki-67/phospho-c-Jun; 1321N1 at 40 nM/four days showed the opposite proliferation response and no p-JNK/p-c-Jun change PMID:34830034(https://pmc.ncbi.nlm.nih.gov/articles/PMC8622492/). Neither source erases the other by recency, and neither establishes a universal JNK inhibitor or AP-1 coactivator core. Historical secretion and developmental TAS observations remain contextual with original-abstract limits PMID:9690476(https://pubmed.ncbi.nlm.nih.gov/9690476/).
Six stored generic-binding rows cover seven original contexts. They remain non-core under the explicit campaign instruction, overriding the skill's default removal of uninformative binding. No adaptor/enzyme-activator function is inferred from partner identity, co-IP or AP-MS. Retained catalytic-complex membership integrates association with recovered activities PMID:29477842(https://pubmed.ncbi.nlm.nih.gov/29477842/); it does not establish fixed stoichiometry, obligatory assembly, purified pairwise contacts or substrate channeling. Distinct HCT116 and HEK293T glycoforms in PMID:27601598(https://pmc.ncbi.nlm.nih.gov/articles/PMC5098040/) are not merged into one universal glycan structure. CDP-glycerol competition is an in-vitro finding, not a new physiological regulatory annotation PMID:29884773(https://pmc.ncbi.nlm.nih.gov/articles/PMC6078450/).
Both natural products O75072-1 and O75072-2 are preserved exactly as seeded. UniProt's isoform 2 VSP_045961 changes residues 426–461 to NQQGA; no localization or catalytic effect is assigned to that product from the available experiments. Tags, soluble secretion constructs and disease mutants are not natural alternative products.
All 21 normal sources are present. Seventeen exact-main files were hydrated and mode/blob/byte/hash verified before genuine fetch; four qualified absent publications were acquired normally. The known current research-provider credential/quota failures were not retried. Manual primary research is described honestly, and no provider-named research file was authored. Abstract-only caches were not edited to include separately recovered text. Annotation-reviewer and core-function-synthesizer criteria were applied, using independent source consultations and ROOT's full 44-context matrix/chemistry review. Final candidate checks and independent review are recorded separately below when complete.
Schema/reference/ontology validation, strict GOA comparison, report-only status, scaffolded explicit-codex history validation, diff checks and private HTML rendering passed. Six expected generic-binding advisories remain under the campaign instruction, alongside a dependency deprecation warning. The PMID helper found no pathway markdown and is not claimed as notes validation. Private preview text and evidence links were inspected without a pixel claim. All 43 seed objects/order,44 source contexts, two natural products,21 normal sources,51 scratch objects and seven old branch pins remain preserved; no supporting-text quotation was added. The shared ontology/term caches remain unchanged. Final candidate publication awaits independent ROOT review of the committed files and closure.
Re-read the original Methods, Results and Discussion for PMID:18808525(https://pmc.ncbi.nlm.nih.gov/articles/PMC2613979/) and PMID:34830034(https://pmc.ncbi.nlm.nih.gov/articles/PMC8622492/) after PR review. Stored row 41, negative regulation of the JNK cascade, changes to MARK_AS_OVER_ANNOTATED: the real phospho-c-Jun staining response does not isolate regulation of JNK itself. The later 1321N1 phosphorylation null result limits generalization without overturning the earlier observation. The proliferation row stays non-core in its specific HeLa/ZR-75-1 context. Structured reference findings now retain both measured outcomes as CURRENT and record their opposing directions and context limits. The source statement that the earlier study proposed suppression in its tested cells remains true; a hypothetical broader interpretation is not attributed to that paper as a DISPUTED finding, and no superseded_by link is added. Cell type, siRNA dose and endpoint differ, but their causal contributions to the difference have not been isolated.
Current stored decisions are 26 ACCEPT, 13 KEEP_AS_NON_CORE, 2 TOO_GENERAL, 1 MARK_AS_OVER_ANNOTATED and 1 UNDECIDED. The 44 original contexts have 26 ACCEPT, 14 KEEP_AS_NON_CORE, 2 TOO_GENERAL, 1 MARK_AS_OVER_ANNOTATED and 1 UNDECIDED. The earlier journal entry records the initial candidate; this addendum records the revision. All 43 original source objects, both natural products, the biological description and single core remain unchanged. The original history is retained and a separate explicit-codex follow-up record is appended.
Donor-specific experimental claims not independently supported by a cached reference were removed; the official mouse Q8R507 identity and original ISS source field remain, with independent human nuclear evidence. Pathway-level glycosylation annotations remain informative because FKTN catalyzes glycan extension, whereas the cytoplasm term adds no compartment precision to established Golgi localization. The Golgi assembly reason now leads with actual localization support. The tentative native ER claim remains UNDECIDED: routine transit and alternative interaction extensions do not settle the original immunogold assignment. DRAFT status remains explicit.
Six generic-binding rows retain the campaign-authorized non-core treatment. The standing instruction against new supporting-text quotations remains in force independently of the prohibition on speculative NEW annotations. Reference-only support and primary-reading provenance remain explicit; no absent quotation is claimed to have passed a substring check. Authored identifiers and spacing were corrected without editing any normal source file.