The genuine normal fetch produced 11 GOA rows and 11 PENDING source contexts for human EXOC6B/Q9Y2D4. The enriched official QuickGO capture projects byte-exactly to the TSV and preserves every seeded source tuple. No collapse, extension, quotation, core or scientific decision was introduced. The two source-described UniProt products remain unchanged.
Falcon deep research with the configured perplexity-lite fallback was attempted once. Actual diagnostics were HTTP 403 and HTTP 401; the command exited 1 and produced no provider research file. These acquisition notes record that failure and are not an authored provider substitute. Genuine publication acquisition succeeded, including the three ROOT-requested recent references. Cached full-text availability is not evidence that original-paper bodies were read during this intake.
An initial runtime-freeze attempt detected a branch-version race before imports or acquisition. Its complete partial files and failed output remain in attempt-1-runtime-freeze. The successful successor froze all runtime bytes directly from the ROOT-approved fixed commit, instead of reading mutable canonical working-tree source files. No canonical files, Git refs or normal preimages were altered or restored.
The pending source checkpoint is ready for ROOT acceptance and a subsequent scientific assignment. Bibliographic and product acquisition do not adjudicate cytokinesis, membrane fission, transport, binding, cilia or autophagy assertions. EXOC6 and differently named Sec15 products must not be equated with EXOC6B by a shared short name.
Five complex/transport contexts are supported and retained. Actual PAINT tree/MSA/IBD and InterPro target mappings were checked, with independent original reading for yeast complex assembly (PMID:7615633), Dictyostelium discharge (PMID:22323285), and mammalian Sec15B identification (PMID:11406615). The latter used rat/sheep samples and human sequence identification, not a human functional reconstitution. Intracellular protein transport and Golgi-to-plasma-membrane transport retain their distinct cargo/route scopes; neither is falsely described as a current child of the other.
The two review-derived cytokinesis/membrane-fission contexts remain UNDECIDED after recovery of the cited Cell original (PMID:16213214) in a thesis. Human Sec15 midbody and depletion evidence is positive, but the inspected main paper does not resolve Sec15A/EXOC6 versus Sec15B/EXOC6B. The uncertainty is target attribution, not absence of full text or evidence against a role. PMID:15292201 separately demonstrates why the shared Sec15 name must not transfer a Rab11 experiment automatically.
The source guide gives public article links, operative locators, actual image/text access, current-versus-imported PAINT donor differences, and natural-product limits. All eleven machine source objects, two supplied product objects, original reference fields and normal caches remain intact. Only the seven assigned review objects change in this partial proposal; four experimental interaction reviews remain pending. No core, status, NEW annotation or quotation was authored.
The four direct IPI contexts were reviewed individually using original experimental descriptions and target-pair data. EXOC6B–EXOC5 is positive in human HEK293T tandem purification, HuRI and BioPlex; EXOC6B–SHTN1 A0MZ66-4 is positive in HuRI. All four remain KEEP_AS_NON_CORE. This preserves supported generic binding while avoiding an unsupported molecular-activity replacement. No new annotation or publication quotation is proposed.
The strongest construct-level refinement comes from PMID:27173435 Supplementary Data 1: the N-tagged EXOC6B insert matches current residues 2–811, and the C-tagged insert matches residues 1–811. Supplementary Data 2 supplies actual EXOC5/EXOC6B bait observations in both directions; the Data 3 scored pair is independently positive. The shorter natural EXOC6B product is not inferred to share these results. HuRI's historical product labels and BioPlex's protein-group identifiers are kept separate from that exact sequence comparison.
Full operative Boldt and BioPlex experimental sections were read. HuRI's actual Results and archived/current pair records were inspected, while full Methods and supplementary pair-validation tables remain unavailable/unread. The source guide provides public paper, table and dataset locators and distinguishes screen-wide validation from pair-specific evidence. No new source cache was fetched or edited; normal inputs and all seven unassigned review stubs remain unchanged in this lane candidate. Whole-gene integration, synthesis, status and publication are separate phases.
Both accepted source lanes are combined without changing a decision: 11 current contexts, five ACCEPT, four KEEP_AS_NON_CORE and two UNDECIDED. The pending statements above describe the earlier acquisition and individual lane checkpoints. All 11 reviews are now resolved to those decisions; the two UNDECIDED contexts preserve the unresolved historical Sec15 paralog attribution. Thirteen references combine six original identities with seven additions, with no shared-overlay collision and no publication quotations. The supplied description, INITIALIZED status and absent core remain unchanged at this annotation phase. Standalone synthesis and validation status belong to the subsequent phase.
All 17 intake normal sources and six genuine additive normal publications are pinned. Eleven cited publications and three PAINT auxiliary files are selected for later delivery, alongside two unchanged machine sources and three curated files. Uncited acquisition-only records remain private. Publication cache bytes remain immutable; the proposed machine files are the genuine fetched outputs and receive their own fixed-main comparison. A fresh-main comparison belongs to ROOT publication preparation.
The synthesis retains one exocyst contribution, using contributes_to_molecular_function for GO:7770062 vesicle membrane tethering activity. It does not assign autonomous membrane bridging, Rab binding, a nucleotide-exchange activity or membrane-fusion chemistry to EXOC6B. Human EXOC5 association and complex incorporation are combined with mammalian Sec15B identification and the experimentally supported, conserved exocyst role. The individual EXOC6B contact remains a residual mechanistic gap. An unestablished subunit-specific MF is omitted rather than invented.
The current ontology explicitly places exocyst under vesicle tethering complex, whose capable-of relation points to the tethering MF. The three accepted processes describe complementary cargo, route and release facets: intracellular protein transport (GO:0006886), Golgi to plasma membrane transport (GO:0006893) and exocytosis (GO:0006887). The captured is_a/part_of ancestor graph contains no ancestor/descendant relationship between any pair. The route definition precedes exocytosis; the latter includes preparatory tethering and docking. They are retained in one core, not split into redundant activities. No new annotation row or process is proposed.
The human evidence is specifically bounded. PMID:27173435(https://doi.org/10.1038/ncomms11491), Methods and Supplementary Data 1–3, supplies reciprocal EXOC6B–EXOC5 recovery and the verified C-terminally tagged 811-residue insert. The N-terminally tagged insert matches residues 2–811. Affinity purification in HEK293T cells is not purified autonomous tethering. PMID:11406615(https://doi.org/10.1074/jbc.C100320200), Experimental Procedures, Table I and KIAA0919 Results, identifies Sec15B by sheep-brain peptide matching to a historical human sequence, with rat-brain comparisons; it does not determine which subunit binds RalA directly. PMID:7615633(https://doi.org/10.1083/jcb.130.2.299), assembly Results and Figures 2–4, concerns yeast. PMID:22323285(https://doi.org/10.1091/mbc.E11-06-0576), Figures 3–6 and operative Methods, concerns tagged Dictyostelium Sec15/Sec8 at contractile-vacuole contact sites. Its proposed exocyst-associated GEF interpretation is not an EXOC6B GEF assay. No equivalent function of the natural 672-residue product is inferred.
The cytokinesis and membrane-fission UNDECIDED calls are unchanged. The actual human Sec15 evidence in PMID:16213214(https://doi.org/10.1016/j.cell.2005.07.027) lacks a sequence/accession distinction between Sec15 paralogs in the inspected article. That is an attribution limit, not evidence that Sec15 or the exocyst does no work in cytokinesis. The explicit distinction between the Rab11-assayed Sec15 protein and KIAA0919/Sec15B in PMID:15292201(https://doi.org/10.1074/jbc.M402264200) also prevents importing an individual Rab-binding activity by name alone.
Three existing genuine publication caches are promoted to the cited set for recent biological context. PMID:36150098(https://pmc.ncbi.nlm.nih.gov/articles/PMC7615863/), Methods and Figures 3–4, supports biallelic EXOC6B disease association. Patient 1's exon-20-spanning deletion, protein loss and shorter cilia are separate from patient 2's p.Leu134Ter, transcript/RNA-seq and osteogenic-differentiation measurements. Patient 1 cilium frequency is not conclusively reduced across both control comparisons. The original is a patient/cell investigation despite the normal metadata also marking it as a review. The title, identifiers and cache bytes are preserved.
PMID:38305850(https://doi.org/10.1007/s11033-023-09114-9) contributes only the accessed abstract and publisher figure captions from mouse ATDC5 work; its full Methods/Results and figure pixels remain unread. PMID:42068746(https://doi.org/10.1016/j.bbrc.2026.153828) contributes metadata, abstract and clipped publisher Methods/Results from human HEK293T experiments. The available reagent identifier HG23270-UT does not verify its insert sequence or natural-product identity, and the complete autophagic-flux controls and mechanism are not established from those excerpts. These sources motivate a controlled trafficking-versus-turnover question, not a new ciliogenesis or autophagy annotation. Patient necessity and a cell perturbation outcome alone do not identify the step performed by EXOC6B.
All 11 accepted annotation/review objects, two natural-product records and the complete first 13 references remain unchanged. The three additions produce 16 references, with no publication quotations. The preceding annotation checkpoint deliberately retained the seed description and omitted core synthesis; this later section records that separately authorized synthesis. Publication caches remain byte-identical; the two genuine machine inputs are separately listed for main-branch comparison. Current-main reconciliation, canonical status, rendering and history remain ROOT publication work.
Added the curated UniProt FUNCTION/SUBUNIT statements as support and explained why the core uses contributed membrane tethering. PMID:36150098 now supports the exocytosis row and core as human disease context: the authors infer impaired exocytosis, while their documented measurements are protein loss, cilia and other cellular phenotypes, not direct secretion flux. Membrane fission remains UNDECIDED for both unresolved Sec15 paralog identity and the unestablished scission-step contribution. The human EXOC4/EXOC5/EXOC6 cilium comparator is mixed and does not independently justify a NEW annotation. Simplified the standalone summary and several transport rationales; preserved positive interaction evidence and all source objects.