CASK

Summary: CASK restrains keratinocyte adhesion to collagen in the studied culture system.

Reason: CASK-directed depletion accelerated HaCaT adhesion and spreading on collagen. This is a supported context-dependent epithelial phenotype, distinct from the principal neuronal and membrane-scaffolding functions. Evidence beyond the cached abstract comes from the independently read official publisher Results/Fig. 9; the publication cache remains abstract-only.

Summary: A GUK-like domain does not confer GMP kinase activity on CASK.

Reason: GO:0004385 specifies ATP-dependent GMP-to-GDP catalysis. The C-terminal CASK-SH3GK construct lacked detectable guanylate kinase activity and GMP binding in the later biochemical study, with yeast guanylate kinase as a positive control. CASK protein phosphorylation uses its separate N-terminal kinase domain; it cannot rescue this electronic GMP-kinase assignment. The CASK-specific negative activity result and yeast-GK positive control were read in the official publisher Results/Fig. 1B of PMID:12482754, beyond its abstract-only cache; PSD95 mutational rescue and ligand-binding controls are not relabeled CASK experiments.

Summary: The historical source did not demonstrate GMP kinase catalysis.

Reason: The complete 1998 paper explicitly states that guanylate kinase activity had not been tested; its discussion is a sequence-based proposal. Later direct CASK-SH3GK assays found no detectable GMP kinase activity. This decision concerns the actual catalytic reaction, not the validity of the independently active N-terminal protein kinase. The CASK-specific negative activity result and yeast-GK positive control were read in the official publisher Results/Fig. 1B of PMID:12482754, beyond its abstract-only cache; PSD95 mutational rescue and ligand-binding controls are not relabeled CASK experiments.

Summary: Refine the assayed protein-kinase activity to demonstrated serine phosphotransfer.

Reason: PMID:18423203 maps phosphorylated CASK and neurexin peptides containing serine acceptors, including neurexin residues 422-433, alongside direct ATP-dependent biochemical assays and the S24D/V26L kinase hypomorph. GO:0106310 captures the demonstrated serine phosphotransfer and matches the catalytic core. This refinement does not assert that every other acceptor residue is experimentally excluded. Magnesium inhibits rather than activates the measured wild-type enzyme; the broad EC-derived electronic row remains a valid less-specific summary.

Proposed replacements: protein serine kinase activity

Summary: The N-terminal CASK domain catalyzes protein phosphorylation.

Reason: The 2008 study demonstrates ATP-dependent autophosphorylation and phosphorylation of neurexin in biochemical assays, supported by kinase-hypomorph controls and neuronal experiments. Activity occurs without magnesium and is inhibited, rather than stimulated, by added magnesium. The protein name does not establish calcium/calmodulin dependence.

Summary: Receptor binding is part of the conserved CASK scaffold function.

Reason: The PAINT assertion is a phylogenetic judgment rather than a donor-count vote. Direct neurexin and syndecan interaction evidence agrees with the inherited receptor-binding role; no target-specific loss is established. The original ancestral node and all supporting entities remain unchanged. Receptor recognition is a mechanistic component of the existing adaptor core, not an additional core process.

Summary: CASK GUK-region association with the rabphilin3a C2B domain is experimentally reported.

Reason: The complete PMID:11377421 abstract identifies the two domains and explicitly reports yeast two-hybrid, GST pull-down and coimmunoprecipitation confirmation. The exact O14936/Q9Y2J0 UniProt record corroborates the partner identity. These are sufficient grounds to retain the qualitative association while deferring to the experimental curator; complete construct endpoints, species and controls are not asserted as read. The proposed rabphilin-CASK-neurexin vesicle-targeting model is not treated as a demonstrated exocytosis mechanism or GMP-kinase activity.

Summary: The CASK-PRKD2 association has exact curated corroboration.

Reason: The Smad-network abstract does not expose the CASK-PRKD2 experiment. The exact curated O14936/Q9BZL6 record nevertheless corroborates the reported association. No kinase-substrate relationship, phosphorylation event or Smad-regulatory function is inferred. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: CASK association with ID1 is supported by several explicitly reported assay types.

Reason: The complete PMID:15694377 abstract reports yeast two-hybrid screening, endogenous coimmunoprecipitation from ECV304 lysates and mammalian two-hybrid assays. The exact O14936/P41134 record corroborates the partner. Retain the association without requiring all unpublished controls to be reconstructed. The differential binding concerns ID1 splice variants, not CASK products; the proposed link to growth regulation remains mechanistic speculation rather than a new molecular activity.

Summary: CASK organizes LIN7-containing multiprotein assemblies.

Reason: The 2006 paper maps LIN7 and EPS8 to distinct CASK/LIN2 L27 regions, uses human CASK bait with mammalian lysate, and recovers the three proteins in mammalian complexes. Together with the curated LIN7-CASK-APBA1 assembly, this supports a molecular adaptor mechanism beyond undifferentiated binding. Worm developmental experiments and canine cell localization are not described as human experiments.

Proposed replacements: molecular adaptor activity

Summary: EPS8 association is supported within the CASK/LIN7 scaffold system.

Reason: Human CASK GST bait recovered mammalian EPS8; mammalian coimmunoprecipitation and lateral localization corroborate the complex. The physiological first-L27 targeting experiment was performed in worm. Retain the supported interaction without assigning human EGF-receptor signaling or a direct interface from coimmunoprecipitation alone.

Summary: Neurexin is a specifically identified CASK binding partner and recruited kinase substrate.

Reason: Full-length CASK binds the neurexin cytoplasmic tail and supports efficient phosphotransfer; PDZ-mediated recruitment is mechanistically distinct from kinase catalysis. The source partner is rat Nrxn1 and remains exactly preserved.

Proposed replacements: neurexin family protein binding

Summary: The CASK-HPV E6 association has exact curated corroboration.

Reason: The original MCC/Scrib-focused abstract does not expose the CASK-E6 assay. The exact O14936/P06463 record explicitly identifies a cross-species interaction. Retain a viral association, without assuming a normal human core function or direct recognition of an E6 PDZ motif. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: APBA1-derived peptide binding is directly supported.

Reason: The human CASK kinase-domain study identifies a shared hydrophobic docking pocket for Mint1/APBA1 and CASKIN1 sequences. This is an interaction surface, not proof that APBA1 is a kinase substrate; the supported binding is retained without turning the isolated peptide assay into a complete cellular assembly experiment.

Summary: A TIAM1-derived docking peptide binds the CASK kinase domain.

Reason: The study directly tests the TIAM1 motif at the shared CASK docking pocket, but proposes rather than establishes a physiological CASK-TIAM1 signaling complex. Retain bounded biochemical binding; do not infer guanine-nucleotide exchange activity, its regulation, or TIAM1 phosphorylation.

Summary: CASKIN1 docking to the CASK kinase domain is experimentally supported.

Reason: Purified human CASK residues 1–310 bind the CASKIN1 interaction region. Pocket substitutions disrupt binding with folding controls. These results establish a docking interaction, not substrate phosphorylation or universal behavior of all six CASK products.

Summary: The CASK SH3-GK region recognizes a phosphorylated whirlin peptide.

Reason: The 2011 study tests a whirlin sequence with and without phosphorylation and reports binding to CASK SH3-GK only for the phosphorylated peptide. Phosphoprotein binding captures this demonstrated specificity; it does not assign guanylate kinase catalysis or claim that the whole endogenous complex was reconstructed.

Proposed replacements: phosphoprotein binding

Summary: The CASK-LIN7A association has exact curated corroboration.

Reason: The Hippo-network target table was not independently authenticated. The exact O14936/O14910 association is corroborated by UniProt and by the independently read CASK-LIN7 scaffold experiments in PMID:16688213. Those independent experiments do not verify the Hippo dataset or make every detected complex an active Hippo regulator. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-LIN7A association has exact curated corroboration.

Reason: The proteome-scale target construct and controls remain uninspected. The exact O14936/O14910 record and the independently read CASK-LIN7 experiments support retaining the association. No particular screen control or Hippo/other pathway mechanism is inferred from this retention. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-SH2D4A association has exact curated corroboration.

Reason: The proteome-scale study provides an experimental IPI assertion, and the exact O14936/Q9H788 association recurs in the reviewed UniProt interaction record. The individual screen construct and controls are not claimed as read; SH2-domain-mediated docking is not inferred from the partner name. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-SNTB2 association has exact curated corroboration.

Reason: The TRAIL-network physical-association row is corroborated by the exact O14936/Q13425 UniProt record. The source-specific capture controls remain uninspected; an association does not establish CASK participation in death-receptor signaling. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: CASK copurifies with MAS1 in the reported receptor-interactome experiment.

Reason: The complete target-specific Results and Methods describe human receptor tandem-affinity purification and mass spectrometry. This supports a contextual physical association; it does not identify a direct receptor-CASK interface or establish that CASK regulates the receptor pharmacology. Retain qualified non-core binding rather than assign a more specific direct-binding mechanism.

Summary: CASK copurifies with ADRA1D in the reported receptor-interactome experiment.

Reason: The complete target-specific Results and Methods describe human receptor tandem-affinity purification and mass spectrometry. This supports a contextual physical association; it does not identify a direct receptor-CASK interface or establish that CASK regulates the receptor pharmacology. Retain qualified non-core binding rather than assign a more specific direct-binding mechanism.

Summary: CASK copurifies with CYSLTR2 in the reported receptor-interactome experiment.

Reason: The complete target-specific Results and Methods describe human receptor tandem-affinity purification and mass spectrometry. This supports a contextual physical association; it does not identify a direct receptor-CASK interface or establish that CASK regulates the receptor pharmacology. Retain qualified non-core binding rather than assign a more specific direct-binding mechanism.

Summary: The CASK-DUSP10 association has exact curated corroboration.

Reason: The phosphatase-network IPI is corroborated by the exact O14936/Q9Y6W6 reviewed interaction record. No direct dephosphorylation of CASK, enzyme-substrate mechanism or MAPK regulation is inferred from association with a phosphatase. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-SH2D4A association has exact curated corroboration.

Reason: The developmental-variant study reports an IPI edge also recorded for O14936/Q9H788 in UniProt. This corroborates the broad partner association. The precise tested mutation and effect size remain unresolved; no normal-versus-mutant comparison, disease mechanism or altered affinity is asserted. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-ARHGEF26 association has exact curated corroboration.

Reason: The RhoGEF-network IPI is corroborated by the exact O14936/Q96DR7 reviewed interaction record. The target-specific table and capture controls remain uninspected; no guanine-nucleotide exchange activity or direct Rho regulation is attributed to CASK. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-LIN7A association has exact curated corroboration.

Reason: The kinase-network edge agrees with the exact O14936/O14910 record and with independently read CASK-LIN7 scaffold evidence. A kinase-centered experimental design does not make LIN7A a CASK substrate or verify this individual capture experiment. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-APC association has exact curated corroboration.

Reason: The kinase-network IPI agrees with the exact O14936/P25054 UniProt record; independent PMID:39009827 also reports a bounded CASK-APC motif interaction. The motif evidence does not authenticate the kinase-network table or establish APC phosphorylation or a physiological Wnt-regulatory complex. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-TIAM1 association has exact curated corroboration.

Reason: The kinase-network IPI agrees with the exact O14936/Q13009 record and with a directly tested TIAM1-derived docking peptide in PMID:21763699. The peptide result does not verify this network capture, a native signaling complex or TIAM1 phosphorylation. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-SNTB2 association has exact curated corroboration.

Reason: The kinase-network association is corroborated by the exact O14936/Q13425 record. Retain contextual physical association, without interpreting the network as a demonstrated phosphorylation reaction or direct binary interface. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-ARHGEF26 association has exact curated corroboration.

Reason: The kinase-network association is corroborated by the exact O14936/Q96DR7 record. The uninspected target-level assay does not justify a phosphorylation, exchange-factor regulation or direct-binding claim. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-KCNA4 association has exact curated corroboration.

Reason: The cell-specific BioPlex association agrees with the exact O14936/P22459 reviewed record. The exact cell-specific table and bait direction were not independently read; channel association alone does not establish gating, transport or channel-regulatory activity by CASK. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-APBA1 association has exact curated corroboration.

Reason: The BioPlex association agrees with the exact O14936/Q02410 record and with direct APBA1-derived peptide binding in PMID:21763699. This corroboration does not reconstruct the BioPlex cellular complex, authenticate its individual controls or establish APBA1 phosphorylation. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-SNTB2 association has exact curated corroboration.

Reason: The cell-specific BioPlex IPI is corroborated by the exact O14936/Q13425 reviewed record. The individual AP-MS controls and cell specificity remain uninspected; no direct interface or additional molecular activity follows. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-DLGAP4 association has exact curated corroboration.

Reason: The BioPlex IPI is corroborated by the exact O14936/Q9Y2H0 reviewed interaction record. Do not import the whirlin phosphorylation-dependent assay or PSD95-DLGAP mechanisms into this unresolved CASK-DLGAP4 interface. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-EPHA7 association has exact curated corroboration.

Reason: The receptor-kinase-network IPI is corroborated by the exact O14936/Q15375 reviewed record. No EPHA7 activation, receptor phosphorylation, downstream signaling or direct receptor-binding interface is established by this qualitative association alone. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-TIAM1 association has exact curated corroboration.

Reason: The chromosome-21-network association agrees with the exact O14936/Q13009 record and the independently read TIAM1 docking-peptide evidence. The specific network experiment remains uninspected; neither a chromosome-dosage mechanism nor a complete physiological signaling complex is inferred. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The CASK-CFTR association has exact curated corroboration.

Reason: The proximity-labeling source is interpreted with its assay limit: proximity alone does not demonstrate direct binding. The exact O14936/P13569 reviewed interaction record separately corroborates the broad curated partner association. No direct contact, CFTR gating or transport-regulatory function is inferred, and the original proximity experiment is not described as independently validated. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The PMID:36115835 association with UniProtKB:O15427 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P03126 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P03409 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P04201 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P06463 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P13569 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P22459 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P25054 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P25100 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P34741 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:P36809 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q02410 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q13009 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q15375 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q15418 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q81964 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q96DR7 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q9BZL6 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q9IR59 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q9NS75 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q9P202 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q9Y2H0 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The PMID:36115835 association with UniProtKB:Q9Y6W6 needs an exact assay-to-annotation join.

Reason: The paper distinguishes isolated PDZ–motif affinity from full-length prey copurification. It specifically reports CASK capture by viral baits despite undetectable binding by the isolated CASK PDZ domain, and warns that subthreshold affinity is not proof of noninteraction. The annotated pair is retained while its holdup, AP-MS or literature-benchmark provenance is resolved; no direct-binding-positive or removal conclusion is inferred from the aggregate database.

Summary: The CASK-CFTR association has exact curated corroboration.

Reason: The mutation-dependent motif study does not expose the exact CFTR result in the inspected narrative, but the exact O14936/P13569 reviewed record supports the broad association. Retain only that association. The CFTR mutation, wild-type comparison and full-length cellular validation remain unresolved; the demonstrated MAPT neointeraction is not transferred to CFTR. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: The 2024 paper provides target-specific evidence for an APC motif interaction.

Reason: The Discussion explicitly reports reduced CASK binding after APC W2658L in the peptide-selection dataset. Retain a bounded motif interaction, without assuming a physiological APC-CASK complex, kinase substrate relationship or cellular validation of this pair.

Summary: The CASK-SLC16A3 association has exact curated corroboration.

Reason: The exact O14936/O15427 reviewed interaction record corroborates this proteomics IPI, even though the inspected main-text CASK experiments concern SLC43A2. Those experiments are not relabeled as SLC16A3 validation. The source-specific supplement remains uninspected; no transporter activity or transport regulation is assigned to CASK. The retained qualitative association relies on the experimental annotation together with the exact reviewed interaction record, not on treating NbExp as independently verified replicates. No contradictory result for this exact broad association was established.

Summary: Human syndecan-2 cytoplasmic-tail binding is directly supported.

Reason: Human CASK PDZ bound the syndecan-2 cytoplasmic tail in two-hybrid and affinity assays with sequence/competition controls. The same paper separately maps protein 4.1 binding but describes the full syndecan-CASK-cytoskeleton linkage as a model; it does not demonstrate endogenous syndecan-CASK coimmunoprecipitation.

Summary: ATP binding is experimentally established for the CASK kinase domain.

Reason: Nucleotide-bound structures, ATP competition and phosphotransfer assays support this domain-derived electronic annotation. ATP binding by the N-terminal domain is distinct from the unsubstantiated GMP-kinase assignment to the C-terminal GUK-like domain. ATP binding is the cofactor-recognition component of the existing catalytic core; no redundant standalone core is added.

Summary: Basement-membrane localization remains unresolved at the actual extracellular boundary.

Reason: GO:0005604 is an extracellular matrix compartment, distinct from the basal keratinocyte layer or its basolateral plasma membrane. The official publisher Results/Fig. 1 describe intracellular CASK and mark the basement-membrane zone, but the primary image-level attribution of CASK to that matrix has not been established here. The attempted full-size figure access failed; captions and basal position alone do not settle the localization. Retain UNDECIDED for the experimentally curated compartment assertion, without interpreting general intracellular localization as proof that an extracellular or peripheral association is impossible. The same anatomical assessment applies to both preserved source rows.

Summary: Basement-membrane localization remains unresolved at the actual extracellular boundary.

Reason: GO:0005604 is an extracellular matrix compartment, distinct from the basal keratinocyte layer or its basolateral plasma membrane. The official publisher Results/Fig. 1 describe intracellular CASK and mark the basement-membrane zone, but the primary image-level attribution of CASK to that matrix has not been established here. The attempted full-size figure access failed; captions and basal position alone do not settle the localization. Retain UNDECIDED for the experimentally curated compartment assertion, without interpreting general intracellular localization as proof that an extracellular or peripheral association is impossible. The same anatomical assessment applies to both preserved source rows.

Summary: Extracellular basement-membrane localization overextends the conserved intracellular epithelial localization evidence.

Reason: GO:0005604 denotes extracellular matrix, not the basal epidermal cell layer or a plasma membrane. GO_REF:0000107 transfers experimentally curated donor annotations through Ensembl orthology; its mouse provenance is retained, not dismissed as uncurated. The directly read publisher Results/Fig. 1 of PMID:18664494 place adult human CASK in cytoplasm and cell-cell junctions within the basal epidermis, while the reviewed protein record describes nuclear, cytoplasmic and peripheral cell-membrane pools. These positive localization observations support an intracellular epithelial scaffold but do not justify extending that conserved distribution to the extracellular matrix compartment. Mark this automated compartment transfer as over-annotated on that biological distinction. This is not an experimentally demonstrated absence, a claim that the donor annotation is wrong, or an inference from failed image access. The two human IDA rows remain UNDECIDED because their exact matrix-level attribution has not been independently resolved. Do not replace this row with basolateral plasma membrane: that distinct location already has separately supported annotations.

Summary: CASK can occupy nuclear compartments in a context-dependent manner.

Reason: Human keratinocyte experiments directly corroborate nuclear CASK and its differentiation-dependent redistribution. The original electronic or orthology evidence is retained; nuclear localization does not by itself establish a transcriptional, RNA-processing or chromatin enzymatic activity.

Summary: CASK can occupy nuclear compartments in a context-dependent manner.

Reason: Human keratinocyte experiments directly corroborate nuclear CASK and its differentiation-dependent redistribution. The original electronic or orthology evidence is retained; nuclear localization does not by itself establish a transcriptional, RNA-processing or chromatin enzymatic activity.

Summary: A nuclear-lamina-associated CASK pool is supported in keratinocytes.

Reason: The source reports retained nuclear-lamina staining after extraction, supported by native and GFP-CASK experiments. This is a contextual subcellular pool rather than a constitutive location for every tissue or product. Evidence beyond the cached abstract comes from the independently read official publisher Results/Fig. 5 and nuclear-extraction Methods; the publication cache remains abstract-only.

Summary: A nuclear-lamina-associated CASK pool is supported in keratinocytes.

Reason: The source reports retained nuclear-lamina staining after extraction, supported by native and GFP-CASK experiments. This is a contextual subcellular pool rather than a constitutive location for every tissue or product. Evidence beyond the cached abstract comes from the independently read official publisher Results/Fig. 5 and nuclear-extraction Methods; the publication cache remains abstract-only.

Summary: The HPA nucleoplasm annotation needs its actual antibody and image record.

Reason: The human keratinocyte paper supports nuclear compartments but does not authenticate the independent HPA experiment or its nucleoplasmic resolution. Preserve the IDA assertion pending that source-level inspection.

Summary: Nucleolar CASK localization is supported in the tested keratinocyte conditions.

Reason: The complete study describes nucleolin-associated nucleolar staining and retention of CASK in extracted nuclear preparations, with GFP-CASK corroboration. This does not establish ribosome biogenesis or an RNA-binding molecular function. Evidence beyond the cached abstract comes from the independently read official publisher Results/Fig. 5D; the publication cache remains abstract-only.

Summary: Nucleolar CASK localization is supported in the tested keratinocyte conditions.

Reason: The complete study describes nucleolin-associated nucleolar staining and retention of CASK in extracted nuclear preparations, with GFP-CASK corroboration. This does not establish ribosome biogenesis or an RNA-binding molecular function. Evidence beyond the cached abstract comes from the independently read official publisher Results/Fig. 5D; the publication cache remains abstract-only.

Summary: A cytoplasmic CASK pool is consistent with its intracellular scaffold and kinase activities.

Reason: Direct human keratinocyte localization and the biochemical/cellular kinase study corroborate a cytoplasmic pool. CASK can redistribute between membrane and nuclear compartments; this annotation does not impose an exclusive or uniform localization. Evidence beyond the cached abstract comes from the independently read official publisher Results/Figs. 1D and 2; the publication cache remains abstract-only.

Summary: A cytoplasmic CASK pool is consistent with its intracellular scaffold and kinase activities.

Reason: Direct human keratinocyte localization and the biochemical/cellular kinase study corroborate a cytoplasmic pool. CASK can redistribute between membrane and nuclear compartments; this annotation does not impose an exclusive or uniform localization.

Summary: A cytoplasmic CASK pool is consistent with its intracellular scaffold and kinase activities.

Reason: Direct human keratinocyte localization and the biochemical/cellular kinase study corroborate a cytoplasmic pool. CASK can redistribute between membrane and nuclear compartments; this annotation does not impose an exclusive or uniform localization.

Summary: Cytosolic CASK is compatible with its soluble precursor pool before membrane recruitment.

Reason: The kinase study describes recruitment of expressed CASK from cytosol to plasma membrane by neurexin. This supports the general location without asserting that all CASK complexes remain freely soluble.

Summary: The curated Reactome assemblies provide a contextual cytosolic location.

Reason: The cached event summaries describe CASK-containing LIN7/APBA1, protein 4.1 and neurexin assemblies. Retain the curator-assigned cytosolic context, while distinguishing it from the peripheral membrane anchoring of the assembled scaffold; the summary alone is not an independent localization experiment.

Summary: The curated Reactome assemblies provide a contextual cytosolic location.

Reason: The cached event summaries describe CASK-containing LIN7/APBA1, protein 4.1 and neurexin assemblies. Retain the curator-assigned cytosolic context, while distinguishing it from the peripheral membrane anchoring of the assembled scaffold; the summary alone is not an independent localization experiment.

Summary: The curated Reactome assemblies provide a contextual cytosolic location.

Reason: The cached event summaries describe CASK-containing LIN7/APBA1, protein 4.1 and neurexin assemblies. Retain the curator-assigned cytosolic context, while distinguishing it from the peripheral membrane anchoring of the assembled scaffold; the summary alone is not an independent localization experiment.

Summary: Peripheral plasma-membrane localization is central to CASK scaffolding.

Reason: Human keratinocyte localization and recruitment by neurexin corroborate the conserved membrane pool. The historical syndecan paper combines human binding experiments with predominantly rat tissue localization; these organism contexts remain distinct.

Summary: Peripheral plasma-membrane localization is central to CASK scaffolding.

Reason: Human keratinocyte localization and recruitment by neurexin corroborate the conserved membrane pool. The historical syndecan paper combines human binding experiments with predominantly rat tissue localization; these organism contexts remain distinct.

Summary: Peripheral plasma-membrane localization is central to CASK scaffolding.

Reason: Human keratinocyte localization and recruitment by neurexin corroborate the conserved membrane pool. The historical syndecan paper combines human binding experiments with predominantly rat tissue localization; these organism contexts remain distinct.

Summary: Peripheral plasma-membrane localization is central to CASK scaffolding.

Reason: Human keratinocyte localization and recruitment by neurexin corroborate the conserved membrane pool. The historical syndecan paper combines human binding experiments with predominantly rat tissue localization; these organism contexts remain distinct.

Summary: CASK acts at cell junctions through membrane-associated protein assemblies.

Reason: The direct human keratinocyte experiment supports the cell-junction location and agrees with conserved CASK scaffold activity. The human target appearing among PAINT supporting entities is legitimate descendant evidence, not circular propagation.

Summary: Human keratinocyte observations support CASK at cell-cell junctions.

Reason: The official publisher Results and Fig. 1D explicitly report adult human skin, with CASK in cytoplasm and cell-cell junctions of the epidermal basal layer. Figs. 2 and associated Results separately examine human HaCaT keratinocytes. These observations are distinct from the newborn-mouse skin and rat hair-follicle experiments foregrounded in the abstract. This is external full-prose/caption evidence; the canonical cache remains abstract-only.

Summary: The transferred cell-cell-junction location agrees with human observations.

Reason: The original mouse-derived electronic assertion is retained and is independently corroborated by direct human skin and keratinocyte junctional localization. No reassignment of its original evidence code or donor is made.

Summary: Focal-adhesion enrichment requires the exact high-throughput CASK record.

Reason: The article describes a myosin-II-responsive focal-adhesion proteome, but the CASK-specific detection and enrichment record has not been resolved. Increased focal-adhesion staining after CASK knockdown in another study is a phenotype, not localization of CASK itself.

Summary: CASK can modulate adhesion through membrane–cytoskeletal assemblies.

Reason: The historical TAS source proposes an adhesion-related scaffold mechanism from binding and localization. The later human keratinocyte perturbation study directly demonstrates altered adhesion. Retain broad contextual involvement without treating CASK as an extracellular adhesion ligand or asserting a universal direction of effect.

Summary: Positioning partner proteins is a conserved CASK scaffold function.

Reason: The inherited PAINT judgment agrees with the LIN2/CASK-LIN7 assembly and with domain-specific receptor-localization experiments. The organism-specific worm and canine experiments support a conserved mechanism; no claim of a newly reconstructed PAINT tree or target-specific loss is made.

Summary: CASK restrains proliferation in the tested human keratinocyte system.

Reason: CASK knockdown increased HaCaT growth and cell-cycle markers, with organotypic culture corroboration and two siRNAs. This context-specific regulatory phenotype is supported, but it is not the defining molecular function of CASK. Evidence beyond the cached abstract comes from the independently read official publisher Results/Figs. 6-8; the publication cache remains abstract-only.

Summary: CASK restrains proliferation in the tested human keratinocyte system.

Reason: CASK knockdown increased HaCaT growth and cell-cycle markers, with organotypic culture corroboration and two siRNAs. This context-specific regulatory phenotype is supported, but it is not the defining molecular function of CASK. Evidence beyond the cached abstract comes from the independently read official publisher Results/Figs. 6-8; the publication cache remains abstract-only.

Summary: The NAS protein-transport assertion needs the complete structural-paper context.

Reason: The cached L27-complex paper is abstract-only. Selected official full-text Introduction and Methods identify mouse CASK/LIN7 L27 constructs and cite earlier trafficking work, but the complete target transport evidence was not recovered. Assembly of a structural binding module is compatible with trafficking and does not independently establish the cited transport process; retain uncertainty without inferring motor activity.

Summary: CASK associates with a membrane-linked actin cytoskeletal scaffold.

Reason: The historical human binding experiments identify protein 4.1 association, and curated later neurexin-CASK-protein 4.1 events describe coupling to actin/spectrin assemblies. Retain scaffold-associated localization; neither direct actin binding nor actin-polymerization catalysis is asserted.

Summary: General membrane association is supported by more specific CASK localization evidence.

Reason: The orthology transfer is consistent with directly observed human cell-junction/plasma-membrane CASK. The inherited machine row remains intact; the core summary uses the more informative plasma-membrane location.

Summary: Membrane association is independently well established for CASK.

Reason: The original NAS structural-paper attribution is not fully verified, but direct human localization and neurexin recruitment independently establish this broad compartment. Retain the biologically secure location while recording the original source-access limitation in the reference assessment.

Summary: Basolateral membrane localization fits the conserved epithelial scaffold role.

Reason: Historical tissue localization and mammalian CASK/LIN7/EPS8 colocalization support a basolateral or lateral membrane pool. This is compatible with the PAINT node and orthology inference; it is not an extracellular basement-membrane assertion.

Summary: Basolateral membrane localization fits the conserved epithelial scaffold role.

Reason: Historical tissue localization and mammalian CASK/LIN7/EPS8 colocalization support a basolateral or lateral membrane pool. This is compatible with the PAINT node and orthology inference; it is not an extracellular basement-membrane assertion.

Summary: An extraction-resistant nuclear CASK pool is reported in keratinocytes.

Reason: The complete study operationally assigns retained CASK to nuclear-matrix/nucleoskeletal preparations. Retain this context-dependent localization without expanding it to a universal structural constituent of the nucleus. Evidence beyond the cached abstract comes from the independently read official publisher Results/Fig. 5 and nuclear-extraction Methods; the publication cache remains abstract-only.

Summary: The electronic cell-junction assignment agrees with direct human localization.

Reason: Human adult skin and HaCaT observations support a junctional CASK pool. The ARBA source remains unchanged; independent primary evidence supports the location without assigning all epithelial phenotypes to a single molecular mechanism.

Summary: The vesicle transfer needs the actual mouse donor experiment.

Reason: CASK-containing trafficking assemblies make this plausible, but a broad trafficking role or receptor complex does not locate CASK to a particular vesicle population. The exact donor localization was not inspected.

Summary: Neurexin binding is a directly supported CASK activity.

Reason: Biochemical and cellular experiments establish binding and recruitment of neurexin by full-length CASK. This independently supports the transferred annotation and is distinct from neurexin phosphorylation. Neurexin recognition is represented in the existing scaffold and substrate-recruitment synthesis; a separate duplicate core is unnecessary.

Summary: Presynaptic membrane association is consistent with neurexin-anchored CASK.

Reason: The source is a curated rat-to-human similarity transfer. Neurexin-associated recruitment and the conserved presynaptic scaffold mechanism agree with this localization; no assertion of exclusive presynaptic expression is made.

Summary: Synaptic CASK localization fits its established neuronal scaffold function.

Reason: The general synaptic annotation agrees with neurexin-associated complexes and neuronal experiments. It does not establish every specialized synapse compartment or all product-specific locations.

Summary: These nucleotide-metabolism transfers depend on an unsupported GMP-kinase reaction.

Reason: The GO-derived inference is explicitly sourced from GMP kinase activity, which is contradicted by the CASK-SH3GK biochemical assessment and was not tested in the historical TAS paper. Protein-substrate phosphorylation by the separate N-terminal kinase does not place CASK in GMP-to-GDP metabolism. The CASK-specific negative activity result and yeast-GK positive control were read in the official publisher Results/Fig. 1B of PMID:12482754, beyond its abstract-only cache; PSD95 mutational rescue and ligand-binding controls are not relabeled CASK experiments.

Summary: These nucleotide-metabolism transfers depend on an unsupported GMP-kinase reaction.

Reason: The GO-derived inference is explicitly sourced from GMP kinase activity, which is contradicted by the CASK-SH3GK biochemical assessment and was not tested in the historical TAS paper. Protein-substrate phosphorylation by the separate N-terminal kinase does not place CASK in GMP-to-GDP metabolism. The CASK-specific negative activity result and yeast-GK positive control were read in the official publisher Results/Fig. 1B of PMID:12482754, beyond its abstract-only cache; PSD95 mutational rescue and ligand-binding controls are not relabeled CASK experiments.

Summary: Regulation of neurotransmitter secretion is a conserved CASK-associated neuronal role.

Reason: The PAINT inference integrates descendant experimental evidence and is coherent with CASK organization of presynaptic partner assemblies. No target-specific loss or phylogenetic misplacement has been demonstrated; retain the sign-neutral inherited process without inventing a universal stimulatory or inhibitory direction.

Summary: Ciliary-membrane localization needs the rat source experiment.

Reason: General cell-surface CASK localization does not establish a ciliary compartment. The donor-specific cilium evidence could not be verified in the available source material or direct donor query; preserve this curated similarity assertion for follow-up.

Summary: The wound-healing direction is broader than the directly tested outcomes.

Reason: The full paper shows reduced CASK near a mouse wound and increased human keratinocyte proliferation/adhesion after knockdown. These support candidate healing-related mechanisms, but do not resolve a decrease in the rate or extent of tissue repair caused by CASK. Neither a CASK-perturbed repair outcome nor a directionally established repair-step mechanism in the wounded tissue is demonstrated. Retain the narrower experimental epithelial processes separately; this does not require one particular assay or deny a possible repair role. Evidence beyond the cached abstract comes from the independently read official publisher Results/Fig. 9; the publication cache remains abstract-only.

Summary: CASK restrains growth-factor responsiveness in the tested keratinocytes.

Reason: Knockdown enhanced the measured response to keratinocyte growth factor and TGF-alpha in human cells. Retain the contextual regulatory process without assigning direct receptor inhibition or extending it to all growth factors. Evidence beyond the cached abstract comes from the independently read official publisher Results/Fig. 7B; the publication cache remains abstract-only.

Summary: The specialized Schaffer collateral–CA1 location needs its mouse donor record.

Reason: General hippocampal/synaptic CASK evidence does not independently establish this anatomical synapse compartment. The exact transferred mouse experiment has not been inspected; preserve it without claiming a contradiction.

Summary: CASK directly phosphorylates serine-containing protein substrates.

Reason: The primary study maps serine-containing phosphorylated CASK and neurexin peptides and tests phosphotransfer with biochemical and cellular controls. This supports the experimental annotation and Rhea-derived mapping; only this more specific activity is used for the kinase core summary.

Summary: CASK directly phosphorylates serine-containing protein substrates.

Reason: The primary study maps serine-containing phosphorylated CASK and neurexin peptides and tests phosphotransfer with biochemical and cellular controls. This supports the experimental annotation and Rhea-derived mapping; only this more specific activity is used for the kinase core summary.

Summary: Synaptic-vesicle exocytosis regulation is consistent with conserved CASK assemblies.

Reason: The mouse-derived process is compatible with the curated presynaptic CASK-LIN7-APBA1 organization. Retain the sign-neutral contextual transfer without asserting that CASK is itself a membrane-fusion catalyst or that every CASK-containing complex regulates release identically.