This guide combines the accepted review of 17 database-provenance and 34 literature-provenance contexts. All 51 imported contexts are reviewed; the historical lane checkpoints remain in the notes. The imported terms, evidence codes, reference identities, qualifiers and supporting entities remain unchanged. Recommendations synthesize the source trace with independently read target evidence; they do not convert an electronic or ancestral assertion into a new experimental assertion. Campaign instructions require supported broad terms to be refined with MODIFY and prohibit new publication quotations for this pass.
The database lane independently read the main text of four papers at the Methods, Results and caption passages below. Its original-article scope does not include figure pixels or supplements. The direct lane additionally read other main sections and inspected selected original figures, as separately listed below. Complete availability is not a claim that every paragraph or figure was read.
Current GO:0008409 allows removal of nucleotide residues from the 5′ end. The DNA-specific child GO:0035312 explicitly requires sequential mononucleotides and is therefore not a faithful narrowing of the inspected trimer-product assay. Conversely, GO:0004528 and EC:3.1.4.1 define removal from a 3′-hydroxy terminus. Their alternative names do not override that reaction definition. The imported phosphodiesterase-I context is refined to the experimentally established 5′-to-3′ exonuclease activity.
GO:0017108 remains the imported, experimentally grounded flap-incision facet. Its literal definition includes substrate exclusivity and Okazaki-fragment wording that merit ontology clarification. The revised core uses GO:0004520 DNA endonuclease activity to encompass both established flap incision and independently measured non-flap extrusion incision; it does not add a redundant ancestor core alongside a flap core. The endonuclease and retained 5′ exonuclease core terms are non-ancestral. Quantitative DNA binding and structural contacts support GO:0070336, rather than inferring binding only from cleavage. Human crosslink processing supports GO:0036297; ordinary nucleotide excision repair is not treated as a synonym.
The saved PANTHER PTHR15749 v19 tree and MSA contain human target PTN002496574 with an exact 1017-residue sequence match. It descends from PTN000405526. The five actual IBD rows match nucleus, 5′ exonuclease, flap endonuclease, interstrand-crosslink repair and flap-structured DNA binding, with the literal source donor lists retained and no relevant NOT on the target path. Human FAN1 appearing in its own descendant evidence is valid grounding, not circularity; donor count is not a quality measure. Independently read human assays corroborate all five activities or locations, and the mouse exonuclease source was traced to the actual assay in PMID:24981866. Other plant, worm and yeast donor primary experiments were not newly read here. NO_FAILURE_CORE records the sound inherited biology. The nucleus IBD is now retained as ACCEPT; independently resolved target pools do not manufacture a more specific ancestral IBD.
Current IPR011856 is a broad nuclease-fold superfamily. IPR006642 is the Rad18-type UBZ4 domain, not the SAP region. Its DNA-binding and DNA-repair mappings are preserved as provenance, while the narrower whole-protein recommendation uses separate target assays. Human UBZ recruitment/ubiquitin recognition is not itself a direct DNA-binding assay. IPR033315 and IPR014883 support the FAN1-family and VRR-NUC traces; bacterial substrate preferences are not automatically human functions.
Current ARBA00028773 set 11 matches the human FunFam and vertebrate lineage. The imported mouse Q69ZT1 and ENSMUSP00000130012 identifiers trace to mouse Fan1; the current protein cross-reference includes version .2. Current ARBA00088592 conditions do not match the present human FunFams. The saved rule/target versions document that boundary; they do not prove what matched historically or justify a global rule-failure assertion. The actual target chemistry independently resolves the terminus-specific annotation.
All seven imported nucleus contexts are retained as ACCEPT. Human damage-induced foci and microirradiated tracks independently support GO:0090734 site of DNA damage, a chromosomal region that lacks typed nucleus ancestry. GO:0005654 nucleoplasm is part of the nucleus, but its definition explicitly excludes chromosomes and the nucleolus. HPA supports that distinct pool; damage-focus images alone are not evidence for nucleoplasm. Nucleus has warranted breadth across these nuclear pools. Neither the lateral damage-site term nor nucleoplasm alone preserves the full assertion. This corrects the earlier replacement rationale without denying the damage-associated location, altering the original source objects or weakening valid PAINT inheritance.
The Human Protein Atlas FAN1 cell record and current XML identify HPA048198 as a supported immunofluorescence assay, with an approved overall gene-location assessment. Selected A-431 and HeLa multichannel pixels show nuclear-interior green signal; HeLa also shows cytoplasmic signal. These are not damage-induction or damage-marker experiments. The antigen is residues 921–1010 of the 1017-residue product and is absent from the 533-residue natural product. Nucleoplasm is retained at this resolution without attributing the same localization to both products.
R-HSA-6785732 explicitly recruits human FAN1 entity R-HSA-6785736, Q9Y2M0 residues 1–1017, into the output nuclease-loaded ID2 assembly. The target itself has nucleoplasm as its compartment. R-HSA-6785986 connects catalyst activity 6785992 to physical entity R-HSA-6785734, FAN1 subcomplex R-HSA-6785733 and that same exact target; FAN1 is also released among the outputs. This distinguishes recruited nuclease from damaged-DNA substrate. The graph models a FAN1 dimer and a coordinated multi-nuclease assembly. This lane does not claim a newly read dimer paper, invariant stoichiometry, or a complete experimental reconstitution of the modeled reaction sequence. The source itself qualifies the ordering and redundancy of other nucleases.
The 1017- and 533-residue natural products remain separate literal source objects. The short product substitutes residues 527–533 and lacks 534–1017, including the assayed C-terminal nuclease region. Neither this sequence fact nor full-length evidence establishes the short product's function, inactivity or localization. Engineered 360/364/373–1017 fragments are different molecules. The displayed Chain 1–1017 feature is not evidence of proteolytic maturation.
Reference flags describe the immutable normal cache, while manual notes separately describe actual original access and the passages read. PMID:20603015 remains cache-abstract-only despite separate complete-main access; PMID:20603073 and PMID:24981866 have full-text normal caches. The legacy PMID:25430771 normal contains full text but has no full_text_available key; that metadata absence is preserved, with no invented author flag. Bibliographic identities were checked against saved PubMed metadata and the actual originals. Complete access does not mean every section or figure was read. No source cache was rewritten, no publication quotation was added, and no biological claim is based on a failed retrieval.
The ROOT-authored direct lane read the same four complete main articles more broadly and inspected selected original pixels: Figures 2, 3, 5 and 7 of PMID:20603015; Figures 3, 4 and 7 of PMID:20603073; Figure 1 of PMID:24981866; and Figures 3 and 4 of PMID:25430771. Other figures were read through main Results/captions where recorded. No supplementary files were independently read. The reference notes preserve both reviewers' full scopes, and the private ledgers retain the exact indexed passages and image hashes.
Liu et al., PMID:20671156 was recovered as a genuine official Science main PDF. ROOT read its complete main text and three captions and inspected Figure 1 on PDF page 5. The separate integration peer read the saved main text and captions without a new pixel claim. HeLa MMC-induced foci and chromatin recruitment support damage-site localization; the Figure 3 caption supplies magnesium-containing nuclease conditions. This is not direct measurement of magnesium binding. The normal cache remains abstract-only, distinct from the externally recovered complete main.
Kratz et al., PMID:20603016 remains abstract-only after actual publisher restrictions and the authentic ZORA content endpoint's denial. The old PDF-shaped URL returned HTML, not an original. Its positive human nuclease and UBZ-dependent recruitment statements are retained and separately corroborated by accessible originals. Reference-specific HR and nucleotide-excision-repair experiments remain unresolved; another paper cannot establish what was measured in the inaccessible full text. Shereda et al., PMID:20935496 likewise remains full-text inaccessible, but its verified abstract expressly supports nuclear damage foci, RPA colocalization and UBZ dependence. No unseen assay details are attributed to it.
The human HR reporter in PMID:20603015 is a supported non-core context, with no direct strand-invasion or initiating-resection mechanism assigned. Cellular FANCD2/FANCI/MutL associations remain supported non-core interactions under the campaign exception; they do not automatically establish purified binary interfaces, obligatory stoichiometry or mismatch-repair activity. Current GO:0140036 requires an adaptor that coordinates a ubiquitinated target with another protein. UBZ-dependent recruitment of FAN1 itself does not settle that third-protein requirement. Both imported reader contexts remain UNDECIDED without denying recognition or excluding an unmeasured adaptor role.
PMID:26797144, Lachaud et al. 2016 separates recruitment-dependent fork restraint from crosslink resistance. The complete cached main narrative was read, including Results referring to Figures 1–4 and the Discussion. Human U2OS FAN1 knockout/rescue experiments show that UBZ C44A/C47A restores MMC resistance but fails to restore HU-dependent fork restraint; a nuclease-dead preparation also fails the fork endpoint. Mouse MEF fiber/epistasis experiments and aged-mouse cancer phenotypes are distinct evidence. The proposed fork-reversal mechanism remains untested in this study. No supplementary Methods or figure pixels were read in this feedback pass. The physical cleavage substrate and endonuclease-versus-exonuclease contributions to restraint remain unresolved.
PMID:37549289, Phadte et al. 2023 directly tests full-length human FAN1 and D960A on extrahelical DNA. Actual reading covered Introduction, individual main Results paragraphs, Figures 1–5 captions, Discussion and the main Methods summary; no SI Appendix or figure pixels were inspected. At 125 mM KCl, RFC, PCNA and ATP support incision of nicked circular DNA when the nick and extrusion lie on the same strand. PCNA alone stimulates the corresponding linear substrate, consistent with RFC providing clamp loading; ATP is not required for intrinsic FAN1 catalysis. Random six-base extrusions are also cleaved, so no CAG-exclusive sequence recognition is assigned.
The same study links FAN1 to short-patch extrusion removal through partial depletion/add-back in MLH1/MSH3-deficient HCT116 nuclear extracts. MutSβ competes with FAN1 for extrusion occupancy. MutLα association is a separate interaction: up to fourfold excess did not significantly change FAN1 activity in the tested reaction. Neither result turns FAN1 into a canonical mismatch-repair enzyme. Complete extrusion removal has not been reconstituted from purified components; DNA synthesis and ligation belong to unresolved additional activities, not FAN1.
Biallelic FAN1 deficiency causing karyomegalic interstitial nephritis, human Huntington disease modifier associations and animal repeat-expansion effects are cited background in these papers. This feedback pass did not independently reread the original clinical genetics or animal-repeat studies. They provide disease context, separate from the executed fork and biochemical experiments.
Two coordinated catalytic units describe one enzyme: 5′ exonucleolysis and DNA endonucleolysis. The latter now explicitly covers both flap incision and extrusion incision. The retained crosslink-repair process and GO:0035753 maintenance of DNA trinucleotide repeats are non-ancestral in the retrieved is_a/part_of closure. The repeat role is grounded in FAN1 performing DNA incision plus extract depletion/rescue, rather than solely in a genetic modifier association. Flap-structured DNA recognition remains explicit in substrate positioning and existing accepted binding annotations.
The new process facet underwent a performer/parent/comparator/model check. FAN1 itself performs incision; it is not the consumed DNA substrate or the polymerase/ligase completing repair. The term concerns maintenance of repeat fidelity/copy number under chromosome organization and DNA metabolism. Finite current human queries of the same repair-nuclease role (FEN1, EXO1 and ERCC1) returned no annotations under repeat maintenance; this is not evidence that curators overlooked a universal nuclease function. The FAN1-specific biochemical/extract result supplies the positive basis here. The fixed published GO-CAM index has no FAN1 entry. No NEW annotation row is manufactured, and no ancestor/descendant process is added to the same core unit.
Whole-enzyme fork restraint is explicit in the description and catalytic prose but is not assigned independently to either catalytic mode. GO:2000104 negative regulation of DNA-templated DNA replication was considered; no mode-specific assignment is made. GO:0031297 replication fork processing expressly requires restoration and restart, and the new 2016 restraint experiment does not itself establish those endpoints. Fork reversal is not asserted. Questions ask which cleavage reaction accounts for the observed restraint and which additional activities complete extrusion removal.
Damage-site localization remains the demonstrated crosslink-repair context in the cores. Broader nucleoplasmic detection is retained without assuming that all imaged protein is catalytically active or that every extrusion reaction occupies an induced focus. The two existing HR/NER UNDECIDED contexts remain source-access specific; the two reader/adaptor UNDECIDED contexts remain definition specific. Thirteen original reference objects are unchanged; the GO_REF:0000044 manual assessment has one stale replacement sentence corrected to retain nucleus and the independent damage-site core. Every original identity, title, finding and flag is unchanged. Natural products remain exact, and two correctly identified full-main references are added with separate actual-read scopes. The absence of legacy availability metadata for PMID:25430771 remains exact. No new publication quotations or source-cache edits were made.