CDC14A

Summary: Retain the ancestral microtubule-organization inference as non-core.

Reason: The PAINT ancestral assertion is retained, with human centrosomal and spindle localization and deregulation phenotypes compatible with a regulatory role. The inspected microtubule-regrowth assay in PMID:12134069, however, used transient GFP-hCDC14A expression: high expression abolished centrosomal nucleation, while low expression did not. This does not establish microtubule organization as a principal normal human function. The drebrin/Arp2 mechanism in PMID:30467237 concerns actin and cilium regulation, not a demonstrated microtubule-organization step. Reported COS-7 colocalization in PMID:29293958 used transfected GFP-tagged mouse Cdc14a and does not supply an endogenous human organization assay. Normal mitotic timing in RPE1 CDC14A/CDC14B knockouts limits an indispensable-role claim without disproving every regulatory contribution. Preserve the original PAINT node and term but remove this process from the core synthesis.

Summary: A cell-cycle-dependent spindle-pole pool is supported.

Reason: Human CDC14A localization was detected at spindle poles in anaphase. Retain this phase-specific compartment without treating the protein as an obligate structural spindle component.

Summary: Protein serine/threonine phosphatase activity is the defining enzymatic function.

Reason: Direct human catalytic-domain measurements support strong phosphoserine preference within this broader catalytic class. Multiple human protein substrates corroborate the conserved PAINT activity. Low phosphothreonine efficiency does not invalidate the class.

Summary: The serine/threonine phosphatase mapping is supported by direct catalysis.

Reason: Retain the original Rhea and EC mappings. CDC14A reverses phosphorylation of human protein substrates, with marked preference for proline-directed phosphoserine sites.

Summary: CDC14A dephosphorylates MAPK6/ERK3.

Reason: The Reactome reaction is supported by the primary binding and dephosphorylation studies. This activity does not imply that its depicted nuclear compartment was experimentally established.

Summary: Exact human protein-phosphotyrosine catalysis remains unresolved.

Reason: The original recombinant-enzyme abstract describes dual-specificity kinetics, but its precise human protein-phosphotyrosine substrate assays were not accessible. Later human catalytic-domain assays show strong phosphoserine preference; the inspected tyrosine-peptide comparison used yeast enzyme. In PMID:41308992, phosphatase-dead mouse CDC14A residues 1-345 bound a pSer-Pro peptide in a hydrophobic proline-recognition pocket, while active truncation kinetics used pNPP and ApSPRRR. That mouse structural/biochemical context supports selectivity but is neither a direct assay of the seeded human product nor a protein-phosphotyrosine assay. Preference and pNPP hydrolysis do not establish absence or presence of human protein-Tyr catalysis; retain UNDECIDED rather than declaring this EC/RHEA inference false.

Summary: CDC14A associates with UXT in the reported assays.

Reason: The primary study reports a human-library two-hybrid hit and reciprocal tagged-protein co-immunoprecipitation in HEK293T cells. Endogenous co-immunoprecipitation was not established. Retain the supported generic interaction under the project instruction; these experiments do not identify UXT as a phosphatase substrate.

Summary: The exact HuRI partner is supported by curated assay records.

Reason: IntAct record IM-25472-156817 links CDC14A to Q8N5M1 under PMID:32296183 and Supplementary Table 9, with corresponding array and pooled-screen records. These human-protein two-hybrid observations support a low-resolution interaction. Repeated assay records are not independent physiological replication, and no partner activity or substrate relationship is transferred.

Summary: The exact HuRI partner is supported by curated assay records.

Reason: IntAct record IM-25472-161486 links CDC14A to Q9NWQ9 under PMID:32296183 and Supplementary Table 9, with corresponding array and pooled-screen records. These human-protein two-hybrid observations support a low-resolution interaction. Repeated assay records are not independent physiological replication, and no partner activity or substrate relationship is transferred.

Summary: A nuclear CDC14A pool is supported.

Reason: The original primary abstract explicitly reports nuclear localization. Retain this broad compartment while recognizing cell-state and construct-dependent distribution; it does not establish nucleolar activity.

Summary: The broad nuclear mapping is biologically supported.

Reason: Independent nuclear observations support this UniProt localization mapping. Export and C-terminal targeting affect the balance between nuclear and cytoplasmic pools.

Summary: Human Protein Atlas supports nucleoplasmic localization.

Reason: The current CDC14A Human Protein Atlas record reports supported nucleoplasmic staining with antibody HPA023783. This is localization evidence; it does not identify a nuclear substrate or establish nucleolar activity.

Summary: Nucleoplasmic localization is retained with a source-compartment limitation.

Reason: Independent Human Protein Atlas staining supports the location. Reactome explicitly states that the MAPK6 reaction location is undetermined despite depicting it in nucleoplasm; the reaction diagram is not direct evidence of compartment-specific catalysis. Preserve the original TAS assertion and reference.

Summary: Nucleoplasmic localization is retained with a source-compartment limitation.

Reason: Independent Human Protein Atlas staining supports the location. Reactome explicitly states that the MAPK6 reaction location is undetermined despite depicting it in nucleoplasm; the reaction diagram is not direct evidence of compartment-specific catalysis. Preserve the original TAS assertion and reference.

Summary: Physiological nucleolar activity of wild-type CDC14A is unresolved.

Reason: The qualifier is is_active_in, which requires activity in the nucleolus rather than possible nuclear access. Selected human experiments distinguish centrosomal CDC14A from nucleolar CDC14B; nucleolar accumulation of an export-defective CDC14A mutant does not establish wild-type activity there. These target observations weigh against a constitutive wild-type nucleolar role, but do not exclude every cell-state-specific nucleolar function. UNDECIDED therefore records unresolved biology, not a rule that biological evidence cannot challenge an IBA without a tree reconstruction. The complete PAINT node reconstruction was not inspected, and no erroneous-node or paralog-misattribution claim is made.

Summary: CDC14A functions in cytoplasmic pools.

Reason: Human substrate dephosphorylation and cell-state-dependent cytoplasmic distribution support the active-in assertion. The PAINT donor list is an ancestral-function judgment rather than a count of independent experiments.

Summary: The centrosome is a supported site of CDC14A function.

Reason: Human localization and phosphatase-dependent cytoskeletal regulation corroborate this PAINT active-in assertion. The target in WITH/FROM is legitimate descendant experimental grounding, not circular evidence.

Summary: Centrosomal localization is independently well established.

Reason: The precise CDC14A entry in the cited proteomics supplements was not inspected. Nevertheless, independent human localization experiments directly support this location; retain the experimental curator assertion without claiming its original target-level proteomics result was reverified.

Summary: The centrosome vocabulary mapping is supported.

Reason: Human localization studies independently establish an interphase centrosomal pool, consistent with the electronic mapping.

Summary: The broad organizing-center location can be specified as centrosome.

Reason: Human experiments support the more informative centrosome compartment. Preserve the original electronic source while proposing this location refinement.

Proposed replacements: centrosome

Summary: Spindle localization is supported during mitosis.

Reason: Selected human localization experiments place CDC14A at the central spindle and spindle poles. This is a temporally restricted pool rather than its only constitutive location.

Summary: A regulatory contribution to mitotic exit is plausible but not universally required.

Reason: Human CDC14A dephosphorylates CDH1 and stimulates APC(CDH1) in the original biochemical study. Conversely, human RPE1 single and double CDC14A/CDC14B knockouts do not alter mitotic timing. Retain the weaker regulatory PAINT assertion as non-core; it does not assert the indispensable yeast mitotic-exit role.

Summary: Normal hearing depends on CDC14A function.

Reason: Human genetic and ortholog functional evidence supports hearing dependence; retain it as a non-core physiological outcome rather than a sensory-receptor activity.

Summary: Normal hearing depends on CDC14A function.

Reason: Human genetic and ortholog functional evidence supports hearing dependence; retain it as a non-core physiological outcome rather than a sensory-receptor activity.

Summary: Hair-cell stereociliary localization is supported by ortholog evidence.

Reason: Mouse hair-cell staining supports the inherited stereociliary compartment; no direct human localization assay is claimed.

Summary: Stereocilium-tip enrichment is supported in mouse hair cells.

Reason: Mouse staining and tagged-protein localization support tip enrichment. Preserve this orthology transfer without presenting it as a human IDA.

Summary: Stereocilium-tip enrichment is supported in mouse hair cells.

Reason: Mouse staining and tagged-protein localization support tip enrichment. Preserve this orthology transfer without presenting it as a human IDA.

Summary: The positive cytokinesis role needs reconciliation across human perturbation models.

Reason: Older human RNAi and misexpression studies report cytokinesis defects, whereas later human single and double knockouts do not impair cytokinesis. These results do not establish absence of every regulatory role, but a direct positive physiological mechanism and the context of the discrepancy remain unresolved. Preserve the PAINT assertion rather than removing it on nonessentiality alone.

Summary: Kinociliary localization is a supported vertebrate tissue context.

Reason: Retain mouse-supported kinociliary localization without inferring that all human products have the same distribution.

Summary: Kinociliary localization is a supported vertebrate tissue context.

Reason: Retain mouse-supported kinociliary localization without inferring that all human products have the same distribution.

Summary: CDC14A regulates cilium growth through protein dephosphorylation.

Reason: Human RPE1 experiments connect direct DBN1 dephosphorylation to control of ciliary growth. A regulation-of-assembly term captures this mechanism better than an undifferentiated assembly claim. GO:1902018 negative regulation of cilium assembly describes the tested RPE1 direction: loss of phosphatase activity lengthens cilia and overexpression reduces ciliation. It is not selected as the general replacement because zebrafish morpholino depletion instead shortens kinocilia, whereas germline mouse and zebrafish mutants can have normal kinocilium lengths. The parent GO:1902017 preserves regulatory participation without generalizing one direction across those cell and perturbation contexts. The original PAINT node is preserved.

Proposed replacements: regulation of cilium assembly

Summary: A mitotic spindle-associated CDC14A pool is supported.

Reason: Retain the PAINT active-in inference, corroborated by human MKlp2/KIF20A-dependent recruitment and separate in-vitro INCENP dephosphorylation. No direct in-cell spindle catalytic-turnover assay was inspected. This is an ancestral activity inference supported by target evidence, not a recoding of localization as direct activity evidence.

Summary: Kinociliary basal-body localization is supported by mouse experiments.

Reason: Preserve the exact mouse supporting entity and basal-body compartment; product-specific human localization is not established.

Summary: Kinociliary basal-body localization is supported by mouse experiments.

Reason: Preserve the exact mouse supporting entity and basal-body compartment; product-specific human localization is not established.