FANCD2

Summary: Retain the encompassing nuclear location of FANCD2.

Reason: The actual PANTHER v19 PTHR32086 tree and ungapped MSA place human PTN002528859 (1451 residues, exact displayed UniProt sequence) below PTN000788426; all seven target-path IBDs were checked and none is negated. The nuclear IBD includes human Q9BXW9, mouse Fancd2 and worm fcd-2; the human self-WITH is legitimate experimental grounding. Mouse PMID:24001775 directly stains Fancd2 nuclear foci in embryonic stem cells (Methods immunofluorescence; Results/Figure5). PMID:20075016 is available here as an abstract describing worm FANCD2 focus formation and its dependencies; uninspected full-text details are not claimed. Human nuclear and chromatin localization and HPA nucleoplasm/nuclear bodies corroborate the transfer. Retaining nucleus preserves the breadth of chromosomal and nonchromosomal pools.

Summary: Retain the conserved meiotic pairing role as a germline specialization.

Reason: The actual PANTHER v19 PTHR32086 tree and ungapped MSA place human PTN002528859 (1451 residues, exact displayed UniProt sequence) below PTN000788426; all seven target-path IBDs were checked and none is negated. The pairing IBD cites mouse Fancd2 and Arabidopsis AT4G14970. Mouse PMID:12893777 directly examines Scp3/Rad51-stained spermatocyte spreads and reports excess unsynapsed or mispaired homologs from mid-zygonema (Results/Figure2; chromosome-staining Methods). In contrast, Arabidopsis PMID:29352063 reports normal completed pachytene synapsis (Figure6) despite later univalents and crossover defects. Thus the plant crossover phenotype is not independently recast as defective initial pairing. The positive mouse evidence and conserved placement support retaining this source assertion as non-core; expression in testis alone would not establish it.

Summary: FANCD2 contributes to the intra-S DNA-damage checkpoint in defined experimental contexts.

Reason: The actual PANTHER v19 PTHR32086 tree and ungapped MSA place human PTN002528859 (1451 residues, exact displayed UniProt sequence) below PTN000788426; all seven target-path IBDs were checked and none is negated. The IBD traces to Drosophila Fancd2/FBgn0038827; the accessible PMID:16860002 abstract explicitly reports an irradiation-induced S-phase checkpoint defect. Independent human PMID:14988723 full text shows partial loss of the crosslink-induced slowing of DNA synthesis in FA-D2 cells, restoration by complementation, and parallel CHK1 and NBS1/FANCD2 branches (Results/Figure3). PMID:12086603 provides abstract-level support for ATM/S222-dependent signaling. Mouse PMID:12893777 found an intact IR/PUVA checkpoint in primary Fancd2-null fibroblasts, so necessity is not universal across species, cell states and lesions. This context dependence does not establish loss of the human function or invalidate the PAINT placement.

Summary: Interstrand cross-link repair is a central FANCD2 process.

Reason: The actual PANTHER v19 PTHR32086 tree and ungapped MSA place human PTN002528859 (1451 residues, exact displayed UniProt sequence) below PTN000788426; all seven target-path IBDs were checked and none is negated. The fly-grounded IBD is supported by actual PMID:18258493 experiments with FANCD2 RNAi, crosslink-induced mutagenesis and the lats assay, and by PMID:21865299 Figure6/Methods showing epistasis with MSH2/spel1 under diepoxybutane challenge. These perturbation studies support the inherited process assignment; the protein contribution is independently established by the FANCI-FANCD2 DNA-binding complex and human repair literature. FANCD2 is both a ubiquitination substrate and an active structural repair factor, without thereby acquiring FANCL ligase or nuclease chemistry. The inherited PMID:19965384 quotation concerns Xenopus ID2 and is not relabeled as a human assay.

Summary: Retain meiotic double-strand-break repair as a conserved specialized role.

Reason: The actual PANTHER v19 PTHR32086 tree and ungapped MSA place human PTN002528859 (1451 residues, exact displayed UniProt sequence) below PTN000788426; all seven target-path IBDs were checked and none is negated. The IBD is grounded in Arabidopsis AT4G14970. Actual PMID:29352063 documents crossover reduction, exacerbated atm/fancd2 meiotic chromosome fragmentation, suppression by spo11, and additive crossover effects with msh4 and mus81 (Results/Figures2–7 and Methods). These results place FANCD2 in processing meiotic recombination intermediates; they do not establish its own nuclease activity or prove the same quantitative contribution in humans. The conserved role is retained as a germline specialization alongside the main somatic repair role.

Summary: FANCD2 has a genuine condensed-chromosome pool.

Reason: The actual PANTHER v19 PTHR32086 tree and ungapped MSA place human PTN002528859 (1451 residues, exact displayed UniProt sequence) below PTN000788426; all seven target-path IBDs were checked and none is negated. The GO:0000793 IBD traces to mouse Fancd2 MGI:2448480. Its current IDA donor cites PMID:12913077, whose accessible abstract explicitly includes Fancd2 localization on meiotic chromosomes even though its title foregrounds Fanca. Full publisher text was unavailable, so no antibody or figure detail is inferred. Independent human UniProt evidence describes paired sister-chromatid foci after replication stress. Retain the encompassing chromosome location without conflating meiotic mouse localization with every human mitotic mechanism.

Summary: Retain FANCD2 association with DNA polymerase POLN as an accessory molecular function.

Reason: The actual PANTHER v19 PTHR32086 tree and ungapped MSA place human PTN002528859 (1451 residues, exact displayed UniProt sequence) below PTN000788426; all seven target-path IBDs were checked and none is negated. The GO:0070182 IBD uses human Q9BXW9 itself, correctly recording direct descendant evidence. In PMID:19995904, Flag-POLN coimmunoprecipitates endogenous FANCD2 in human293T cells, the Methods include benzonase, and endogenous POLN/FANCD2/FANCI associations are examined in synchronized HeLa cells. This supports a polymerase association, without proving a purified binary interface or assigning polymerase activity to FANCD2. It is retained as non-core relative to the ID2 DNA-clamp function, consistently with the direct IPI row.

Summary: UniProt subcellular-location mapping supports nuclear FANCD2.

Reason: The inherited GO_REF:0000044 row and current GO_REF:0000120 context describe the same broad nucleus assertion through different source versions. The official SL-0191 mapping is GO:0005634; the current combined context also names mouse Q80V62/ENSMUSP00000045667. Human nuclear biology, HPA and the verified mouse nuclear-focus experiment are concordant. Keep the inherited machine reference intact; a changed database reference is not biological retirement. Neither this mapping nor shared antibody sequence establishes equal localization of all four annotated protein isoforms.

Summary: The InterPro family mapping correctly supports broad DNA repair.

Reason: Actual IPR029448 maps FANCD2 to GO:0006281 and covers residues1–1444 of the displayed1451-residue target. DNA repair is a conservative parent assertion supported by the ID2 complex and direct human repair evidence. An underlying signature name containing nuclease does not demonstrate FANCD2 nuclease catalysis; this mapping is used only for the supplied biological-process term.

Summary: FANCD2 associates with menin in a radiation-responsive complex.

Reason: PMID:12874027 reports human 293-cell Flag-menin purification, a FANCD2 peptide and immunoblot validation, endogenous coimmunoprecipitation controls, and increased association after gamma irradiation. The Men1-null mouse damage phenotype is not a FANCD2 perturbation. The positive interaction remains useful contextual evidence; the inspected assays and specific-binding term search do not establish a more informative FANCD2 molecular activity. Retain supported generic binding as non-core under the project exception, without assigning menin-dependent transcriptional or enzymatic activity to FANCD2.

Summary: FANCD2 binds BRCA2 in human interaction assays.

Reason: PMID:15115758 maps a human FANCD2 N-terminal interaction region using two-hybrid constructs and BRCA2 residues 2350–2545, with human-cell coimmunoprecipitation and mitomycin-C comparisons. The human FANCD2 fragment is residues 1–1102; it is not proof that each natural product was tested. The positive binding is retained, but these assays do not establish a specific adaptor, catalytic or other more informative FANCD2 molecular activity. Supported generic binding is therefore non-core.

Summary: FANCD2 forms the functional FANCI–FANCD2 heterodimer.

Reason: PMID:17460694 selected original Results and complete supplemental Methods support human FANCI–FANCD2 nuclear coassociation, codependence and repair phenotypes. The finer heterodimerization term is supported independently by purified full-length human ID2 and resolved FANCI–FANCD2 interfaces in PMID:32269332. This is a cross-source refinement of the target function, not a claim that the 2007 coimmunoprecipitation alone resolved a binary interface. The natural FANCD2 products remain distinct.

Proposed replacements: protein heterodimerization activity

Summary: FANCD2 associates with BRCA2 in a FANCG-dependent repair complex.

Reason: PMID:18212739 selected original Methods, Results and Figures 1–3 captions establish coimmunoprecipitation in human cells and distinguish FANCG wild type/S7A and FANCD2 K561R contexts, with separate hamster experiments. Association does not require FANCD2 monoubiquitination in the tested setting. The proposed restart-complex interpretation does not itself establish an adaptor or catalytic activity of FANCD2. Retain the positive BRCA2 association as non-core generic binding after the specificity search.

Summary: FANCD2 associates with NBS1 within the MRN interaction context.

Reason: PMID:19609304 reports human-cell MRN association and Supplementary Figure 2A coimmunoprecipitation after Sf9 coinfection with human FANCD2 and NBS1. These are positive physical-association data, not an isolated binary-interface measurement. NBS1 is distinct from the catalytic MRE11 partner. No more informative FANCD2 binding activity was established for this NBS1 context; retain supported generic binding as non-core.

Summary: The imported FSCN1 partner identifier conflicts with the FAN1 nuclease explicitly studied in the cited original.

Reason: The literal imported partner is UniProtKB:Q16658 (FSCN1). However, PMID:20603015 Figure 1A explicitly identifies human KIAA1018/MTMR15/FAN1 as Q9Y2M0, and its Methods and Figure 4 experiments examine that 114-kDa nuclease with FANCD2. The current IntAct export maps several of these assays to Q16658 and includes FAN1 as an FSCN1 alias. This is a checkable source-identifier discrepancy, not evidence that FANCD2 never binds FSCN1. Retain this partner-specific claim as unresolved and request an upstream check; do not silently replace its source identifier or merge it with the separately imported Q9Y2M0 context.

Summary: FANCD2 binds the DNA-binding transcription factor CEBPD in a nuclear-import context.

Reason: PMID:20805509 maps FANCD2 association to CEBPD residues 153–199, with human-cell interaction and nuclear-import assays and separately identified mouse experiments. GO:0140297 specifies binding a DNA-binding transcription factor; it does not require regulation of transcription. CEBPD bridges FANCD2 to importin-4 using distinct interaction regions. FANCD2 is the transported partner here, not the bridging import adaptor or a demonstrated transcriptional regulator.

Proposed replacements: DNA-binding transcription factor binding

Summary: FANCD2 associates with the EGFR protein kinase in human-cell interaction assays.

Reason: PMID:35384245 Dataset EV1A row 2766 identifies EGFR-WT AP-MS with FANCD2/Q9BXW9 (AvgSpec 3, BFDR 0). Methods describe tagged RTKs in human Flp-In 293 cells, stimulation, nuclease-treated affinity purification and background filtering. Independent PMID:35512704 Table S6B row 1353 supports cellular EGFR–FANCD2 BRET. GO:0019901 specifies binding a protein kinase without requiring a regulatory effect. This partner-class refinement preserves cellular association and does not claim an isolated binary interface, a physiological signaling consequence, or FANCD2 kinase activity.

Proposed replacements: protein kinase binding

Summary: FANCD2 associates with EGFR in a quantitative cellular BRET screen.

Reason: PMID:35512704 Table S6B row 1353 reports FANCD2 with both EGFR-WT and EGFR-L858R as positive (dose-response FOC 2.1686 and 2.4414, respectively); the high-stringency classification is comPPI. It is not a mutant-only gain of interaction. Tagged NLuc/Venus constructs in HEK293T cells report proximity consistent with association, with independent EGFR AP-MS support in PMID:35384245. GO:0019901 gives the established enzyme-partner class without claiming an isolated direct interface, signaling regulation or FANCD2 catalysis.

Proposed replacements: protein kinase binding

Summary: FANCD2 is a chromatin-associated component of the Fanconi anemia repair machinery.

Reason: PMID:19965384 directly examines Xenopus FANCI–FANCD2 in egg extracts: depletion and wild-type versus K562R add-back link chromatin recruitment to interstrand-crosslink repair. It is not a direct human FANCD2 perturbation. Independent human FANCD2 chromatin fractionation in PMID:19995904 and PMID:26323318 and human ID2 DNA-clamp reconstitution in PMID:32269332 support the conserved compartment. Chromatin association does not assign nuclease or ubiquitin-transfer activity to FANCD2.

Summary: FANCD2 is a functional component of DNA interstrand-crosslink repair.

Reason: PMID:19965384 demonstrates Xenopus ID2 depletion/add-back effects on replication-coupled crosslink repair, with the FANCD2 ubiquitination-site mutant deficient in chromatin recruitment and repair. Independent human data in PMID:19995904 and human DNA-clamp reconstitution in PMID:32269332 corroborate the conserved structural repair role. FANCD2 participates as part of ID2; the experiments do not assign the incision, DNA synthesis or ubiquitin-transfer reactions to FANCD2.

Summary: FANCD2 is an integral member of the FANCI–FANCD2 DNA-repair complex.

Reason: PMID:32269332 reconstitutes and resolves full-length human FANCI–FANCD2, including its ubiquitinated closed DNA-clamp state and interactions between the two subunits. This directly supports DNA-repair-complex membership. FANCL and UBE2T catalyze ubiquitin transfer in the preparation; FANCD2 contributes the clamp rather than that enzymatic step. Full-length experimental constructs are not equated to every natural FANCD2 product.

Summary: HPA supports nucleoplasmic FANCD2.

Reason: The current HPA record classifies nucleoplasm as a main location, supported by CAB016117 in U-251MG and HPA063742 in A-549, SiHa and U2OS. Actual A-549 and U2OS multichannel images were inspected. The HPA063742 antigen occurs at residue253 in Q9BXW9-1, -2 and -3 and is absent as a complete sequence from the 241-residue -4 product; this overlap prevents isoform-specific assignment. Retain the supported gene-level location without making every nuclear signal a DNA-damage focus.

Summary: Retain a context-dependent nucleolar FANCD2 pool.

Reason: HPA063742 reports nucleoli in A-549, alongside nucleoplasm and cytosol; CAB016117 reports nucleoplasm in U-251MG. The A-549 image shows intranuclear structures compatible with the annotated nucleolar pool. This is a genuine source-reported localization but does not establish a ribosome-biogenesis mechanism or an isoform-specific function. Retain it as non-core relative to the chromatin-associated DNA-repair role.

Summary: HPA supports FANCD2 in nuclear bodies.

Reason: HPA063742 reports nuclear bodies in SiHa and U2OS; the actual U2OS multichannel image shows nuclear puncta. This supports GO:0016604 alongside the inherited UniProt evidence for stress-associated foci. The HPA cells are not demonstrated here to have been subjected to DNA damage, so these particular puncta are not assigned a damage-repair identity from morphology alone. The antigen overlaps three products and cannot resolve individual isoforms.

Summary: FANCD2 is nuclear and accumulates at induced DNA-crosslink sites.

Reason: PMID:30335751 directly follows endogenous or tagged human FANCD2 at nuclear psoralen/UVA damage tracks and examines UHRF1/2-dependent recruitment. Nuclear localization is positive; recruitment to particular lesions does not exhaust the broader nuclear pool. Preserve nucleus rather than replace it laterally with a damage-site term. UHRF-dependent recruitment is distinct from FANCL-catalyzed FANCD2 ubiquitination.

Summary: Reactome supports nucleoplasmic FANCD2 in FANCD2:FANCI complex and UBE2T bind ICL-DNA associated with the FA core complex.

Reason: The exact R-HSA-6785342.2 graph contains Q9BXW9 as an ID2 subunit in the input and DNA-bound output. Independently traced routes include 6785342→420764→419522; 6785342→6785340→420764→419522. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in Monoubiquitination of FANCD2:FANCI.

Reason: The exact R-HSA-6785361.3 graph contains Q9BXW9 as the phosphorylated ubiquitination substrate and monoubiquitinated product. Independently traced routes include 6785361→6788398→6788402→6788481; 6785361→6785367→420757→420751→6785318. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in FANCD2 binds FANCI.

Reason: The exact R-HSA-6785594.2 graph contains Q9BXW9 as the free FANCD2 input and the FANCI-bound output. Independently traced routes include 6785594→419522; 6785594→420764→419522. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in DNA nucleases bind monoubiquitinated ID2 complex.

Reason: The exact R-HSA-6785732.4 graph contains Q9BXW9 as a monoubiquitinated ID2 subunit in nuclease-recruitment assemblies. Independently traced routes include 6785732→6785367→420757→420751→6785318; 6785732→6785734→6785367→420757→420751→6785318. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in DNA nucleases unhook the interstrand crosslink (ICL).

Reason: The exact R-HSA-6785986.2 graph contains Q9BXW9 as an ID2 subunit retained during nuclease-mediated unhooking. Independently traced routes include 6785986→6785734→6785367→420757→420751→6785318; 6785986→6786151→420757→420751→6785318. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in POLN binds ICL-DNA.

Reason: The exact R-HSA-6786155.2 graph contains Q9BXW9 as a monoubiquitinated ID2 component in the POLN-recruitment assembly. Independently traced routes include 6786155→6786151→420757→420751→6785318; 6786155→6786156→6786151→420757→420751→6785318. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in Translesion synthesis across unhooked ICL by POLN.

Reason: The exact R-HSA-6786166.2 graph contains Q9BXW9 as an ID2 component carried through the POLN translesion-synthesis event. Independently traced routes include 6786166→6786156→6786151→420757→420751→6785318; 6786166→420757→420751→6785318. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in FANCD2 deubiquitination by USP1:WDR48.

Reason: The exact R-HSA-6786171.3 graph contains Q9BXW9 as the ubiquitinated substrate and deubiquitinated product of USP1:WDR48. Independently traced routes include 6786171→6785367→420757→420751→6785318; 6786171→6788402→6788481. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in The complex of ATR and ATRIP is recruited to ICL-DNA.

Reason: The exact R-HSA-6788385.2 graph contains Q9BXW9 as an ID2 subunit present during ATR:ATRIP recruitment. Independently traced routes include 6788385→6785340→420764→419522; 6788385→6788386→6785340→420764→419522. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in ATR phosphorylates RPA2, FANCI, FANCD2 and FANCM at ICL-DNA.

Reason: The exact R-HSA-6788392.3 graph contains Q9BXW9 as the FANCD2 phosphorylation substrate and phosphorylated product. Independently traced routes include 6788392→6788386→6785340→420764→419522; 6788392→6788398→6788402→6788481. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. The containing assembly does not transfer its kinase, ubiquitin-ligase, nuclease or polymerase activity to each subunit.

Summary: Reactome supports nucleoplasmic FANCD2 in CDK12 stimulates expression of DNA repair genes.

Reason: The exact R-HSA-6797712.4 graph contains Q9BXW9 as a member of the protein output set produced downstream of CDK12-regulated transcription. Independently traced routes include 6797712→6797713→419522. The reached FANCD2 entities themselves specify nucleoplasm and residues1–1451; this is not inferred merely from the reaction compartment. Retain the cellular location. This event concerns FANCD2 expression, not participation of FANCD2 protein in the transcriptional machinery.

Summary: Human FANCD2 is present in nuclear fractions and damage-associated chromatin.

Reason: PMID:26323318 uses human-cell fractionation and immunoblotting, including HeLa soluble nuclear and chromatin fractions with or without mitomycin C. The chromatin signal under untreated conditions can fall below assay sensitivity; that is not absence from the nucleus. The encompassing nucleus term remains appropriate across soluble and chromatin-associated pools. No natural-product-specific resolution is inferred.

Summary: FANCD2 physically associates with DNA polymerase nu.

Reason: PMID:19995904 reports human POLN association with FANCD2/FANCI, endogenous S-phase coimmunoprecipitation and monoubiquitinated FANCD2 in a chromatin-associated complex after mitomycin C. Methods include benzonase treatment. GO:0070182 is the specific polymerase-binding activity, so no generic-binding substitute is needed. Retain this contextual interaction as non-core; neither an isolated binary interface nor FANCD2 DNA-polymerase activity is demonstrated.

Summary: The FAN1 association has a supported enzyme-partner refinement.

Reason: The complete operative original of PMID:20603016 was not recovered; its verified abstract reports FAN1 interaction with monoubiquitinated FANCD2. Independent actual primary experiments in PMID:20603015 and PMID:20603073 corroborate the same human FAN1–FANCD2 association and its UBZ/ubiquitination dependence. This cross-source judgment refines the supported target claim without inventing access to the Kratz Methods. GO:0019899 names binding an enzyme and does not require FANCD2 to catalyze or activate that enzyme. FAN1 supplies the nuclease chemistry; FANCD2 is a regulated interaction partner. The assays do not establish identical behavior of all four natural FANCD2 products.

Proposed replacements: enzyme binding

Summary: FANCD2 binds the FAN1 nuclease in a ubiquitination-dependent repair context.

Reason: PMID:20603073 reports tagged and endogenous human-cell association, chromatin coimmunoprecipitation and PD20 rescue with FANCD2 wild type versus K561R; FAN1 UBZ mutations impair the interaction and recruitment. GO:0019899 names binding an enzyme and does not require FANCD2 to catalyze or activate that enzyme. FAN1 supplies the nuclease chemistry; FANCD2 is a regulated interaction partner. The assays do not establish identical behavior of all four natural FANCD2 products.

Proposed replacements: enzyme binding

Summary: FANCD2–menin association changes after gamma irradiation.

Reason: GO:0010332 includes a change in state or activity in response to gamma radiation. PMID:12874027 reports a time-dependent increase in human FANCD2–menin association after irradiation, maximal at approximately four hours and returning toward baseline by six hours. That is positive response evidence, although the separate Men1-null mouse damage phenotype is not a FANCD2 loss-of-function test. Retain this contextual response as non-core rather than infer universal checkpoint or survival necessity.

Summary: Purified human FANCD2 preferentially binds single-stranded DNA.

Reason: PMID:19609304 directly compares purified human FANCD2 binding to 100-nucleotide single-stranded DNA and alternative substrates, using two purification approaches and antibody-blocking controls. Supplementary Figure 2B/C supports the binding specificity controls. The cDNA contained exon 44; neither the S/L posttranslational forms nor this tested construct establish equivalence of all natural products. Retain the inherited direct molecular-function proposal; the binding assay does not assign nuclease activity.

Summary: FANCD2 supports homologous-recombination repair, with the cited mutant experiments performed in chicken DT40 cells.

Reason: PMID:15601828 disrupts chicken Fancd2 and tests SCneo/I-SceI repair, gene targeting and immunoglobulin gene conversion with rescue. Corrected I-SceI repair frequency falls approximately 26-fold, whereas spontaneous sister-chromatid exchange increases and RAD51 foci remain detectable. These results support particular HR-repair functions, not universal loss of all recombination. Preserve the inherited ISS inference and its original source identity; do not relabel the chicken perturbation as human IMP or use the broad UniProt quotation as a direct human experiment.

Summary: FANCD2 supports recovery of stalled replication forks, with distinct aphidicolin and hydroxyurea experimental contexts.

Reason: PMID:25659033 directly measures restart after aphidicolin in human PD20 cells and rescue by wild-type or K561R FANCD2; the K561R rescue uses higher expression, which limits physiological equivalence. This supports the restoration/restart scope of GO:0031297. PMID:26797144 separately demonstrates FANCD2–FAN1-dependent restraint of hydroxyurea-stalled fork progression in human/mouse contexts; that is not an interchangeable restart or proven fork-reversal assay. Preserve the inherited proposal and both quotation anchors, without assigning a specific nuclease reaction to FANCD2.

Summary: FANCD2 suppresses R-loop accumulation, but its participation in R-loop disassembly remains unresolved.

Reason: The current GO:0062176 definition specifies R-loop disassembly. Actual PMID:30431240 human-cell S9.6/RNase-H controls, collision assays and RNA-processing-factor associations show suppression of accumulation and propose prevention through RNA processing. PMID:38165804 supports an ID2–SRSF1–NXF1 mRNA-export mechanism that limits R-loop formation. Neither inspected study resolves FANCD2-mediated disassembly of an existing R-loop, and reduced steady-state R-loop levels alone do not distinguish prevention from removal. Preserve the inherited source and quotations while leaving this particular process proposal unresolved; this is not a denial of the positive R-loop/homeostasis phenotype.

Summary: A mitochondrial FANCD2 pool is supported by mouse experiments and retained as a contextual orthology-based proposal.

Reason: PMID:28378742 uses an endogenous C-terminal FLAG/HA Fancd2 knock-in mouse, MEFs and mouse tissues, with functional tag controls, mitochondrial imaging/fractionation and compartment markers. The exact reviewed mouse ortholog is UniProtKB:Q80V62 (Fancd2; MGI:2448480), verified in the saved official UniProt record; the earlier Q80V54 wording was incorrect and is corrected only in this authored reason. Preserve ISO rather than imply direct human localization. This is a non-core inferred pool, not evidence that every human natural product is mitochondrial, and the paper did not find a corresponding increase in mtDNA mutation or copy-number defects. ORGANELLES check (2026-10-25). The organelle claim holds for the mouse ortholog by two methods in one paper, MitoTracker co-staining of the endogenous HA-tagged knock-in protein and co-fractionation with Atp5a1, and mitochondrial Fancd2 is lost on Atad3 knockdown. The fractionation used no ER marker, but FANCD2 is a soluble nuclear protein with no ER route, so the MAM caveat applies less than for membrane proteins. FANCD2 is not in MitoCarta3.0, so this stays a non-core ISO proposal. No sub-compartment row, because the nucleoid association rests on co-immunoprecipitation with Atad3, Tufm and Tfam from one study, with no protease protection or sub-mitochondrial fractionation, and FANCD2 has no presequence or TM segment; that is one line, so the row stays at the root.

Summary: The current mouse-to-human condensed-chromosome transfer is supported.

Reason: The literal Ensembl-Compara donor Q80V62/ENSMUSP00000045667 maps to mouse Fancd2. Current mouse IDA GO:0000793 cites PMID:12913077; its abstract explicitly reports Fancd2 localization on meiotic chromosomes. The publisher PDF request was denied, so individual assay details remain unverified. Established human chromosome-associated FANCD2 provides independent support for this broad location. This is a separate current-source context, not an assertion that an absent legacy row was biologically wrong.

Summary: FANCD2 physically associates with the MRE11 nuclease.

Reason: PMID:19609304 reports MRN association and Supplementary Figure 2A shows human FANCD2–MRE11 coimmunoprecipitation in Sf9 coinfection, with IgG controls. GO:0019899 denotes binding an enzyme and does not require activation of it. The known MRE11 nuclease partner provides a more informative class than generic protein binding. The assay does not resolve an isolated binary interface; the supplemental nuclease assay does not show FANCD2 nuclease activity or MRE11 stimulation under those conditions.

Proposed replacements: enzyme binding

Summary: FANCD2 binds FANCI as the ID2 heterodimer.

Reason: PMID:20603015 observes FANCD2 and FANCI in FAN1-associated human complexes; that source alone does not resolve the FANCD2–FANCI interface. Independent purified human ID2 and its resolved heterodimer interface in PMID:32269332 support GO:0046982 as the target-specific refinement. Preserve the source-specific complex-association scope and distinguish FANCI from the nuclease FAN1.

Proposed replacements: protein heterodimerization activity

Summary: FANCD2 binds the DNA-repair nuclease FAN1.

Reason: PMID:20603015 identifies human FAN1 as Q9Y2M0 and demonstrates DNA-independent coassociation with FANCD2, endogenous controls, and dependence on the FAN1 UBZ domain and FANCD2 monoubiquitination. GO:0019899 names binding an enzyme and does not require FANCD2 to catalyze or activate that enzyme. FAN1 supplies the nuclease chemistry; FANCD2 is a regulated interaction partner. The assays do not establish identical behavior of all four natural FANCD2 products.

Proposed replacements: enzyme binding

Summary: UniProt subcellular-location mapping supports nuclear FANCD2.

Reason: The inherited GO_REF:0000044 row and current GO_REF:0000120 context describe the same broad nucleus assertion through different source versions. The official SL-0191 mapping is GO:0005634; the current combined context also names mouse Q80V62/ENSMUSP00000045667. Human nuclear biology, HPA and the verified mouse nuclear-focus experiment are concordant. Keep the inherited machine reference intact; a changed database reference is not biological retirement. Neither this mapping nor shared antibody sequence establishes equal localization of all four annotated protein isoforms.