FLNB

Summary: PMID:11807098 directly observes stress-fiber association of human FLNB constructs expressed in mouse fibroblast/myoblast hosts, with endogenous mouse protein examined separately.

Reason: PMID:11807098 directly observes stress-fiber association of human FLNB constructs expressed in mouse fibroblast/myoblast hosts, with endogenous mouse protein examined separately. This supports the existing gene-level location and agrees with the imported subcellular mapping. Different splice constructs have different distributions; the result is not universal localization of all nine natural products.

Summary: PMID:22681889 describes photoreactive-nucleotide UV crosslinking, oligo(dT) purification and quantitative proteomics in human embryonic kidney cells.

Reason: PMID:22681889 describes photoreactive-nucleotide UV crosslinking, oligo(dT) purification and quantitative proteomics in human embryonic kidney cells. The FLNB-specific target entry and complete experimental text could not be inspected. This access limitation leaves the RNA-binding assertion unresolved; it does not refute RNA binding or establish a paralog error.

Summary: Actin binding is a defining FLNB activity, independently corroborated by purified human wild-type FLNB actin-binding-domain co-sedimentation with F-actin in the PMID:19505475 abstract.

Reason: Actin binding is a defining FLNB activity, independently corroborated by purified human wild-type FLNB actin-binding-domain co-sedimentation with F-actin in the PMID:19505475 abstract. The original PMID:8327473 concerns a 195-residue TABP clone lacking the actin-binding domain and proposes a signaling model; it is retained as historical provenance, not represented as a direct full-length actin-binding assay. Domain and disease-mutant experiments do not establish identical activity for all nine natural products.

Summary: The exact FLNB-DVL1 experiment in PMID:10644691 could not be recovered.

Reason: The exact FLNB-DVL1 experiment in PMID:10644691 could not be recovered. Official identity and abstract were checked, but the full article was inaccessible and the bounded IntAct query did not recover this target pair. The abstract foregrounding PP2C, Dishevelled and Axin does not show that FLNB was absent from the study. Preserve the original O14640 partner and source, with UNDECIDED pending source-specific evidence.

Summary: The exact FLNB-DVL2 experiment in PMID:10644691 could not be recovered.

Reason: The exact FLNB-DVL2 experiment in PMID:10644691 could not be recovered. Official identity and abstract were checked, but the full article was inaccessible and the bounded IntAct query did not recover this target pair. The abstract foregrounding PP2C, Dishevelled and Axin does not show that FLNB was absent from the study. Preserve the original O14641 partner and source, with UNDECIDED pending source-specific evidence.

Summary: Human FLNB constructs bind beta1-integrin tails with splice-variant and assay dependence.

Reason: The imported beta1-integrin partner P05556 supports refinement to integrin binding, using independent primary PMID:11807098 and the explicit FLNB-beta1A result in the PMID:16076904 abstract. PMID:11807098 reports positive human FLNB fragment yeast assays and stronger biochemical beta1A recovery for the variant lacking the 41-residue repeat19/20 segment; wild-type fragment binding in yeast was not reproduced in its GST-tail pull-down. These differences constrain construct and affinity claims, not the existence of the integrin interaction. The original PMID:10676904 is retained verbatim as machine provenance and remains MISCITED for this claim: its actual bovine embryo-culture study does not supply the binding evidence. No canonical PMID remapping, universal nine-product binding, or purified affinity constant is asserted.

Proposed replacements: integrin binding

Summary: Human FLNB associates with FLNA through assayed filamin fragments and in NT2 cells.

Reason: PMID:12393796 directly assays human FLNA and FLNB carboxy-terminal constructs by yeast two-hybrid and recovers endogenous FLNB in an anti-FLNA immunocomplex from human NT2 cells. Its cytoplasmic colocalization supports the retained occurs_in extension. GO:0031005 specifically covers filamin partners and is more informative than generic protein binding or the broader cytoskeletal-protein class. The paper also supports possible heterodimer formation; the replacement records the demonstrated partner class without assigning an exact stoichiometry or mapping inconsistent fragment endpoints to every current natural product. Colocalization alone is not treated as the interaction assay.

Proposed replacements: filamin binding

Summary: FLNB is recovered in a GRB2 SH2-bait complex without a resolved pairwise interface.

Reason: PMID:12577067 uses GST-GRB2 SH2 affinity purification of EGF-responsive complexes; the exact FLNB-GRB2 IntAct record EBI-350217 is spoke-expanded association. Retain this contextual co-complex assertion without inferring a directly measured binary interface. The separate PMID:17474147 FLNB-peptide/GRB2-SH3 assay does not identify the interface for this SH2-bait complex and therefore does not justify an SH3-binding replacement for this source context.

Summary: The cited study explicitly includes FLNB binding to FATZ1/MYOZ1.

Reason: The PMID:16076904 abstract explicitly reports that FATZ1 binds FLNB and its historical variant-1 as well as the other filamins. The imported Q9NP98 partner is MYOZ1, also named FATZ1/calsarcin2; official GO:0051373 covers this family and is a child of cytoskeletal protein binding. Refine to FATZ binding without transferring FLNC-specific binding-site mapping, muscle mechanism or construct endpoints to FLNB. The complete paper and exact FLNB assay panel have not been recovered in this consultation, so precise assay and product assignments remain unverified; the target and partner interaction itself are explicitly identified in the accessible original abstract.

Proposed replacements: FATZ binding

Summary: The cited study explicitly includes FLNB binding to the cytoskeletal Z-disc protein myotilin.

Reason: The PMID:16076904 abstract explicitly reports binding of myotilin to FLNB and its historical variant-1. Myotilin is an actin-cytoskeletal/Z-disc assembly component, established in the source framing and official Q9UBF9 metadata, so cytoskeletal protein binding is an evidence-backed partner class. This does not require transferring the paper's FLNC-specific interface mapping to FLNB. The complete FLNB assay details and mapping to the nine current natural products remain unverified. No muscle developmental or catalytic function is added from the partner's biology. Although cytoskeletal protein binding is an ancestor of the actin-binding terms retained elsewhere, this original PMID:16076904 assertion concerns MYOT rather than actin. Its supported partner class refines this generic row without claiming additional independent core coverage.

Proposed replacements: cytoskeletal protein binding

Summary: An FLNB peptide binds three assayed human NCK1 SH3 domains.

Reason: Exact positive IntAct records for PMID:17474147 pair human FLNB residues240-253 with NCK1 domain fragments2-61,115-165 and190-252 in peptide-array assays. These domain-ligand measurements support SH3 domain binding. The accessible original abstract describes the domain-screen design; its full-length validation rate concerns a PLCG1 subset and is not validation of these FLNB-NCK1 pairs. Keep the original IPI source and partner, with peptide/domain scope and no inferred endogenous or universal product interaction.

Proposed replacements: SH3 domain binding

Summary: An FLNB peptide binds an assayed human GRB2 SH3 domain.

Reason: Exact positive IntAct record EBI-1963515 for PMID:17474147 pairs human FLNB residues240-253 with GRB2 residues156-215 in a peptide-array assay. This supports SH3 domain binding with explicit domain/peptide scope. It does not establish full-length endogenous binding or a physiological interface in the separate PMID:12577067 SH2-bait complex. The paper's abstract-level full-length validation statistic belongs to the PLCG1 subset and is not transferred to this pair.

Proposed replacements: SH3 domain binding

Summary: Covalent FLNB ISGylation is established; independent noncovalent ISG15 binding remains unresolved.

Reason: PMID:19270716 establishes FLNB as an ISG15 conjugation substrate: endogenous FLNB is modified in IFN-treated A549 cells, the expressed conjugate requires ISG15 C-terminal Gly-Gly, UBP43 reverses it, denaturing NTA recovery uses 8 M urea, and K2467R prevents the mapped conjugation in fragment and full-length experiments. These are positive covalent-modification results, whereas GO:0005515 inherits the noncovalent interaction meaning of GO:0005488. They do not independently demonstrate a noncovalent FLNB-ISG15 recognition activity. The exact original IntAct assay record and supplementary information were not recovered, so an additional curator-seen binding experiment cannot be excluded. Preserve the imported IPI assertion, P05161 partner and PMID, but mark its binding interpretation UNDECIDED pending that evidence. This does not reject the ISGylation result or assert that FLNB cannot bind ISG15.

Summary: FLNB associates with RAC1 in the assayed interferon-responsive scaffold.

Reason: Actual PMID:19270716 Results show RAC1 association with the expressed carboxy-terminal FLNB scaffold and with complemented full-length FLNB in M2 cells; ISG15 conjugation and the K2467R substitution alter this association. RAC1 is the assayed small monomeric GTPase, making small GTPase binding the informative replacement. This records the cellular association and regulated scaffold scope; it does not claim purified binary affinity, a nucleotide-state preference, or FLNB GEF/GAP/GTPase catalysis. The IFN-dependent time course and specific constructs are not generalized to all tissues or natural products.

Proposed replacements: small GTPase binding

Summary: FLNB associates with MEKK1/MAP3K1 in the assayed signaling scaffold, supporting MAP kinase kinase kinase binding.

Reason: PMID:19270716 directly reports MEKK1 association with expressed carboxy-terminal FLNB and with complemented full-length FLNB in the IFN-response experiments, with reduced recovery after FLNB ISGylation. Its kinase assay measures MEKK1-mediated MKK4 phosphorylation. This identifies the partner as a MAP kinase kinase kinase and supports GO:0031435, whose definition specifically covers binding to a protein that phosphorylates a MAP kinase kinase. This is a more informative partner-class replacement; no formal child relationship to the earlier GO:0120283 replacement is asserted. The observed cellular scaffold association is not a purified binary affinity measurement, and neither kinase nor E3-ligase activity is assigned to FLNB. The Q13233 context remains distinct from the unresolved imported Q9Y6R4 context.

Proposed replacements: mitogen-activated protein kinase kinase kinase binding

Summary: The imported partner Q9Y6R4 is MAP3K4/MEKK4.

Reason: The imported partner Q9Y6R4 is MAP3K4/MEKK4. Actual PMID:19270716 Results and Figure 3E instead identify Myc-MKK4, a different kinase, in the inspected experiment. Current UniProt retains MAP3K4, and the exact original pair record and supplementary information were not recovered. This discrepancy leaves the exact binding assertion UNDECIDED: preserve Q9Y6R4 and do not silently substitute MAP2K4 or assert that no additional MAP3K4 experiment exists.

Summary: The broad cytoplasm location is entailed by independently supported cytosol and cortex localization.

Reason: The broad cytoplasm location is entailed by independently supported cytosol and cortex localization. Official QuickGO ancestry verifies the cytosol-to-cytoplasm relation. The individual FLNB target images from PMID:16791210 were not recovered, so this is a specificity judgment rather than an accusation based on its nuclear-protein-focused title or a rejection of the original experimental assertion.

Summary: The actual Human Protein Atlas subcellular report supports FLNB cytosol localization with antibody and cell-line scope.

Reason: The actual Human Protein Atlas subcellular report supports FLNB cytosol localization with antibody and cell-line scope. This agrees with the cytoplasmic human NT2 observations in PMID:12393796 and cytosolic scaffold context in Reactome:R-HSA-1169398. Text/table evidence was inspected; no independent image-pixel or antibody-specificity revalidation is claimed.

Summary: The exact Reactome:R-HSA-1169398 summary places FLNB scaffold ISGylation in the cytosol, consistent with actual human-cell experiments in PMID:19270716 and the Human Protein Atlas cytosol report.

Reason: The exact Reactome:R-HSA-1169398 summary places FLNB scaffold ISGylation in the cytosol, consistent with actual human-cell experiments in PMID:19270716 and the Human Protein Atlas cytosol report. FLNB is the modified scaffold, not the enzyme catalyzing ISG15 conjugation. The existing location is supported without adding a conjugating activity or process.

Summary: Actin cytoskeleton is an independently supported, more specific descendant of cytoskeleton, confirmed by official QuickGO ancestry.

Reason: Actin cytoskeleton is an independently supported, more specific descendant of cytoskeleton, confirmed by official QuickGO ancestry. Human FLNB structural and cellular evidence supports the narrower location. The broad imported mapping remains intact and is marked TOO_GENERAL rather than biologically incorrect.

Summary: The actual Human Protein Atlas subcellular report supports plasma-membrane localization in its antibody and cell-line contexts.

Reason: The actual Human Protein Atlas subcellular report supports plasma-membrane localization in its antibody and cell-line contexts. Human NT2 cell-periphery observations in PMID:12393796 and integrin-associated FLNB constructs in PMID:11807098 provide additional context. Peripheral membrane association is compatible with this location and does not imply a transmembrane FLNB topology or universal product localization.

Summary: The actual mouse donor Q80X90 carries IDA PMID:22114352.

Reason: The actual mouse donor Q80X90 carries IDA PMID:22114352. Original Supplement 2 Table S1 explicitly identifies Filamin B (gi|38257404), with 14 matching peptides and 8.3% sequence coverage in isolated wild-type brush borders; FLNA is a separate entry. Supplement 1 describes mouse-database searches. This supports contextual donor fraction localization and the existing transfer, not direct human localization or a FLNB-specific knockout/compensation phenotype. The table reports the best experiment and does not establish gene-unique peptides or natural-product resolution.

Summary: Original PMID:21423176 Results explicitly identify filamin B in isolated focal adhesions from human HFF1 fibroblasts and report reduced abundance after myosin-II inhibition.

Reason: Original PMID:21423176 Results explicitly identify filamin B in isolated focal adhesions from human HFF1 fibroblasts and report reduced abundance after myosin-II inhibition. Retain the fibroblast extension. Fractionation controls and acknowledged residual cytoskeletal contamination limit spatial resolution; this does not assign myosin motor activity or direct adhesion-assembly catalysis to FLNB.

Summary: Original PMID:11807098 shows focal-contact enrichment of human FLNB constructs in mouse hosts, particularly the combined var-1/hinge-1 deletion; it is not a universal full-length localization result.

Reason: Original PMID:11807098 shows focal-contact enrichment of human FLNB constructs in mouse hosts, particularly the combined var-1/hinge-1 deletion; it is not a universal full-length localization result. Independent human HFF1 focal-adhesion proteomics in PMID:21423176 explicitly identifies filamin B and supports the gene-level location. The imported mouse donor context is preserved, without mapping historical construct endpoints to all nine current products.

Summary: Cell-cortex localization agrees with the membrane-associated actin scaffold role and original PMID:12393796 human NT2 cell-periphery colocalization.

Reason: Cell-cortex localization agrees with the membrane-associated actin scaffold role and original PMID:12393796 human NT2 cell-periphery colocalization. The subcellular mapping is retained with that cellular scope; peripheral association does not imply membrane-spanning topology or equal distribution of every splice product.

Summary: PMID:19270716 directly supports regulated FLNB scaffold participation in IFN-responsive RAC1/MEKK1 signaling, distinct from the kinases and conjugating enzymes acting on the scaffold.

Reason: PMID:19270716 directly supports regulated FLNB scaffold participation in IFN-responsive RAC1/MEKK1 signaling, distinct from the kinases and conjugating enzymes acting on the scaffold. Retain broad signal transduction as contextual and non-core. The original PMID:8327473 partial TABP clone and TSH-related hypothesis are historical provenance, not a verified modern full-length signaling assay.

Summary: The full-length human beta-filamin architecture described in PMID:9651345 agrees with actin-cytoskeleton localization, independently supported by the actual human FLNB construct experiments in PMID:11807098.

Reason: The full-length human beta-filamin architecture described in PMID:9651345 agrees with actin-cytoskeleton localization, independently supported by the actual human FLNB construct experiments in PMID:11807098. Accept this core location without portraying the PMID:9651345 abstract as a complete experimental-localization replay or assigning the result indiscriminately to all products.

Summary: Supported plasma-membrane localization is more specific than this broad membrane term, with official QuickGO ancestry confirming the relation.

Reason: Supported plasma-membrane localization is more specific than this broad membrane term, with official QuickGO ancestry confirming the relation. The original PMID:8327473 partial TABP clone and hydrophobic-region predictions are not evidence for modern full-length FLNB membrane topology. Mark the broad location TOO_GENERAL while retaining the original NAS source context.

Summary: PMID:11807098 observes GFP-tagged human filamin B aligned with Z-lines and M-bands in some well-differentiated mouse myotubes.

Reason: PMID:11807098 observes GFP-tagged human filamin B aligned with Z-lines and M-bands in some well-differentiated mouse myotubes. This is positive contextual localization and supports retention as non-core; it is not a universal tissue distribution or an all-product assignment. Endogenous mouse protein and human tagged constructs remain distinct experimental evidence.

Summary: InterPro IPR044801 maps to actin cytoskeleton organization.

Reason: InterPro IPR044801 maps to actin cytoskeleton organization. Human FLNB construct localization and structural experiments in PMID:11807098, together with direct actin-binding evidence in PMID:19505475, support this core role. The separate ARBA rule antecedents were not completely audited; acceptance rests on the demonstrated biology and verified InterPro mapping, not an invented rule mechanism.

Summary: Actin cytoskeleton organization is independently supported by later human FLNB construct and structural evidence in PMID:11807098, full-length architecture in PMID:9651345, and the actin-binding assay summarized in PMID:19505475.

Reason: Actin cytoskeleton organization is independently supported by later human FLNB construct and structural evidence in PMID:11807098, full-length architecture in PMID:9651345, and the actin-binding assay summarized in PMID:19505475. Retain the original TAS citation, PMID:8327473, as historical provenance; its partial TABP clone is not presented as a direct actin-organization experiment.

Summary: Human FLNB carboxy-terminal fragment self-association is supported by the original PMID:12393796 experiments, with separate fragment crosslinking in PMID:11807098 supporting dimerization.

Reason: Human FLNB carboxy-terminal fragment self-association is supported by the original PMID:12393796 experiments, with separate fragment crosslinking in PMID:11807098 supporting dimerization. This is consistent with the structural role of characterized full-length filamin B. The tested fragments and historical splice variants are not equated to every current natural product, particularly the truncated products.

Summary: Selected original PMID:25468996 Results, Methods and Discussion describe E-cadherin-BirA* proximity labeling in human MKN28 cells with parental controls and replicate proteomics.

Reason: Selected original PMID:25468996 Results, Methods and Discussion describe E-cadherin-BirA* proximity labeling in human MKN28 cells with parental controls and replicate proteomics. Proximity alone does not demonstrate binary cadherin binding. Main Table 1 lists FLNA, but the larger FLNB-specific supplementary entry was not recovered, so neither FLNB absence nor a wrong-gene annotation follows. Keep this HDA assertion UNDECIDED pending target-level evidence. The normal cache omits the relevant Results/Methods; the selected primary text was read separately through the indexed official PMC page.

Summary: Official mouse GOA traces donor Q80X90 to IDA PMID:19144319.

Reason: Official mouse GOA traces donor Q80X90 to IDA PMID:19144319. Its authentic abstract describes quantitative phagosomal proteomics in resting and IFN-gamma-activated macrophages, but the FLNB target entry, peptide discrimination and detailed preparation were inaccessible. Keep the inherited localization UNDECIDED pending source verification; this is not evidence against phagosomal localization or a curator/paralog-confusion claim.

Summary: The actual InterPro IPR044801 mapping to actin filament binding agrees with independent human FLNB actin-binding-domain co-sedimentation reported in PMID:19505475.

Reason: The actual InterPro IPR044801 mapping to actin filament binding agrees with independent human FLNB actin-binding-domain co-sedimentation reported in PMID:19505475. It also fits the structural and cellular FLNB evidence in PMID:11807098 and PMID:9651345. Accept the defining activity without extrapolating full-length or mutant behavior to every natural product.

Summary: The original authors' urinary-exosome database explicitly lists FLNB/NP_001448 with one peptide and links the entry to PMID:19056867.

Reason: The original authors' urinary-exosome database explicitly lists FLNB/NP_001448 with one peptide and links the entry to PMID:19056867. This supports contextual cargo localization while preserving the original urine extension. One reported peptide does not resolve natural products or establish regulated secretion, vesicle biogenesis, or an extracellular FLNB function.

Summary: The original PMID:23533145 main text describes urinary/prostatic vesicle preparations, but its exact supplementary FLNB target row was not recovered.

Reason: The original PMID:23533145 main text describes urinary/prostatic vesicle preparations, but its exact supplementary FLNB target row was not recovered. Retain this broad existing location with explicit curator trust and independent FLNB urinary-vesicle detection associated with PMID:19056867. This does not claim source-specific target replay, natural-product resolution, or a functional exosome role.