CDK13

Summary: CDK13 is the catalytic subunit of a cyclin-dependent kinase complex.

Reason: The broad ARBA complex assignment is supported independently by the isolated human CDK13/cyclin K complex and its kinase activity. The specific cyclin K-CDK13 complex is represented in the core; the rule chain itself was not reconstructed.

Summary: Conserved CDK13 contributes to alternative splice-site selection.

Reason: The source Q69ZA1 is mouse Cdk13, whose curated experimental splicing and nuclear-complex annotations cite PMID:17261272. Its reporter study uses a HEK293T host; host species is not taken as proof of a human construct. Independent human CDC2L5/CDK13 experiments in PMID:16721827 support the target role. Preserve the ortholog-based source and the broader splicing process without claiming that CDK13 catalyzes RNA cleavage.

Summary: Conserved CDK13 contributes to alternative splice-site selection.

Reason: The source Q69ZA1 is mouse Cdk13, whose curated experimental splicing and nuclear-complex annotations cite PMID:17261272. Its reporter study uses a HEK293T host; host species is not taken as proof of a human construct. Independent human CDC2L5/CDK13 experiments in PMID:16721827 support the target role. Preserve the ortholog-based source and the broader splicing process without claiming that CDK13 catalyzes RNA cleavage.

Summary: Human CDK13 changes alternative splice-site selection.

Reason: The original human CDC2L5 study reports overexpressed constructs affecting constitutive splicing and alternative splice choice, together with nuclear-speckle localization. This supports the existing splicing role. A separate source links CDK13 to ASF/SF2 phosphorylation; the original paper alone is not described as a purified kinase-substrate assay.

Summary: CDK13 forms a distinct cyclin K complex.

Reason: Full Results/Figure 2 in human 293 cells show reciprocal cyclin K/CDK13 association and distinguish this holoenzyme from cyclin K/CDK12. Preserve the part_of assertion; CDK12 DNA-damage transcription experiments are not transferred to CDK13.

Summary: Structural and biochemical evidence establishes the human CDK13/cyclin K complex.

Reason: The source examines recombinant human kinase/cyclin domains expressed in insect cells and separately full-length human complexes recovered from HCT116 cells. Crystal structure and kinase assays establish the complex, with fragment/host scope distinguished from endogenous full-length assays.

Summary: CDK13 was recovered in a human RNA-capture screen.

Reason: The source uses photoreactive-nucleotide UV crosslinking and oligo(dT) purification in HEK293 cells. The first author’s institutional deposition of this project contains the exact CDK13 class II Table S1 hit, enriched in two of three experiments. Table S3 records failed immunoprecipitation and no CLIP result, not a negative RNA-binding test. Retain broad RNA binding at screen resolution without a sequence-specific RNA-recognition core or an unsupported mRNA-only refinement.

Summary: Protein kinase activity is independently established.

Reason: The kinase-domain inference is corroborated by human CDK13 CTD phosphorylation and a distinct ZC3H14 serine-phosphorylation reaction. Retain the correct broad kinase activity alongside the more specific existing activities.

Summary: The historical kinase assignment is supported by later direct biochemistry.

Reason: The original CHED paper identifies a kinase-related human sequence; it is not presented as the definitive purified-enzyme assay. Subsequent human CDK13/cyclin K CTD kinase experiments establish the broad molecular function independently, while the original TAS provenance is retained.

Summary: CDK13 is a cyclin-dependent protein serine/threonine kinase.

Reason: The EC-based inference agrees with ATP-dependent human CDK13/cyclin K phosphorylation of CTD serines and the cyclin-associated enzyme structure. The molecular activity does not by itself specify which biological process a given substrate controls.

Summary: CDK13 associates with the splicing-related protein C1QBP/p32.

Reason: The authentic human CDC2L5 abstract explicitly reports direct p32 association; UniProt links the interaction to tested CDK13 isoform 2. Preserve the gene-level source assertion without claiming that the interaction is unique to that isoform or assigning p32’s functions to CDK13. The supported generic interaction is retained under the explicit ClinGen project instruction.

Summary: CDC37 associates with human CDK13 in a chaperone screen.

Reason: The exact target-pair IntAct record EBI-9483122 links human CDC37 and CDK13 to the cited luminescence-based mammalian interaction assay in HEK293T cells. This supports association in the tested tagged context, not a purified binary interface or chaperone activity of CDK13. Retain the generic binding as non-core under the standing project instruction.

Summary: Refine the cyclin K interaction to cyclin binding.

Reason: Human CDK13 and cyclin K are linked by the exact crystallographic and size-exclusion records; the protein fragments and heterologous expression host are preserved in the evidence extract. The partner is the established regulatory cyclin K; GO:0030332 describes CDK13 binding that cyclin, rather than incorrectly transferring the kinase’s catalytic function to its partner. The independent holoenzyme structure corroborates this informative partner-class refinement.

Proposed replacements: cyclin binding

Summary: Refine the cyclin K interaction to cyclin binding.

Reason: The exact human CDK13/cyclin K pull-down record EBI-28939636 is linked to the kinase interactome in HEK293 Flp-In T-REx cells. The partner is the established regulatory cyclin K; GO:0030332 describes CDK13 binding that cyclin, rather than incorrectly transferring the kinase’s catalytic function to its partner. The independent holoenzyme structure corroborates this informative partner-class refinement.

Proposed replacements: cyclin binding

Summary: CDC37 is recovered with CDK13 in the kinase interactome.

Reason: The exact human pair is recorded under EBI-28939636, PMID:32707033, in the tagged CDK13 pull-down experiment. It supports contextual chaperone association without an inferred catalytic function or direct isolated binding interface. Retain as non-core under the explicit project binding instruction.

Summary: Refine the cyclin K interaction to cyclin binding.

Reason: Exact human CDK13/cyclin K anti-tag co-IP records are present in BioPlex. Repeated release representations are not counted as independent experiments. The partner is the established regulatory cyclin K; GO:0030332 describes CDK13 binding that cyclin, rather than incorrectly transferring the kinase’s catalytic function to its partner. The independent holoenzyme structure corroborates this informative partner-class refinement.

Proposed replacements: cyclin binding

Summary: Refine the cyclin K interaction to cyclin binding.

Reason: The genuine OpenCell downloadable interaction table contains CDK13 as target and CCNK as interactor; library metadata identifies Q14004 with a C-terminal endogenous tag. This is a human-cell affinity-purification association, not a purified-interface measurement. The partner is the established regulatory cyclin K; GO:0030332 describes CDK13 binding that cyclin, rather than incorrectly transferring the kinase’s catalytic function to its partner. The independent holoenzyme structure corroborates this informative partner-class refinement.

Proposed replacements: cyclin binding

Summary: ATP binding supports CDK13 phosphotransfer.

Reason: The domain-based ATP-binding assertion is corroborated by the ATP-bound human CDK13 kinase structure and ATP-dependent phosphorylation assays. It is an integral property of the catalytic core.

Summary: CDK13 was detected in pooled human tear fluid.

Reason: The authentic source XML Table II contains CDK13_HUMAN, Cyclin-dependent kinase 13, Homo sapiens and CDK13 in the same row; the accession is independently linked to Q14004. Samples were pooled tears from 25 breast-cancer patients and 25 controls. Retain the HDA detection at a non-core extracellular location without inferring active secretion or extracellular kinase function.

Summary: The Reactome target record supports a contextual granule-release location.

Reason: The exact event R-HSA-6800434 connects the ficolin-rich granule set R-HSA-6800431 to extracellular set R-HSA-6806481; both contain a specific human CDK13 entity. Retain the curated extracellular region assignment as non-core. This is curated proteomic/event context, not an independent demonstration of CDK13 secretion chemistry or an extracellular catalytic role.

Summary: The nucleus is a functional compartment of CDK13.

Reason: The PAINT is_active_in assertion agrees with direct nuclear CDK13 enzyme isolation, transcription-associated activity and nuclear RNA-processing substrates. The target’s presence among descendant sources is legitimate experimental grounding, not circularity; no full ancestral-node re-curation is claimed.

Summary: CDK13 functions in the nucleoplasm.

Reason: The Reactome transcription event’s nucleoplasmic location is independently supported by the human CDC2L5 localization study and nuclear kinase/substrate work. Retain it as a core location without treating the curated pathway as a new experiment.

Summary: The inherited transcriptional cyclin/CDK complex is biologically supported.

Reason: GO:0008024 describes a cyclin/CDK complex supporting productive RNA-polymerase-II elongation through CTD phosphorylation; it is not restricted by its name to CDK9. Human CDK13/cyclin K directly phosphorylates CTD and regulates transcript output. Preserve the PAINT complex assertion and use the specific CDK13 holoenzyme in the core; no donor-count argument or guessed tree placement is introduced.

Summary: The PAINT CTD-kinase assertion has direct human target support.

Reason: Human CDK13 biochemical activity corroborates the conserved ancestral-node inference. Self Q14004 among the experimental descendants is valid PAINT provenance. CDK12-specific DNA-repair results are not needed for or transferred into this catalytic assertion.

Summary: Human CDK13 directly phosphorylates the RNA-polymerase-II CTD.

Reason: Selected full primary Results/Figure 4C and purification Methods include human CDK13 isolated from HeLa nuclear material and tested on GST-CTD. The CDK12-focused title does not negate that target experiment. Later human CDK13/cyclin K assays independently establish Ser2/Ser5 phosphorylation; no significant global CTD-loss conclusion is invented for the original knockdown experiment.

Summary: The combined catalytic mapping correctly assigns CTD kinase activity.

Reason: The Rhea/EC/ARBA mapping agrees with direct human CDK13/cyclin K kinase biochemistry. Preferred pre-existing Ser7 phosphorylation is a substrate-recognition feature, not evidence that CDK13 catalyzes Ser7 phosphorylation.

Summary: Signaling-category enrichment extends beyond the measured transcriptional response.

Reason: In PMID:26748711, human HCT116 CDK13 knockdown was followed by expression microarrays and qRT-PCR; RNAi-resistant rescue restored three of four tested transcript levels. Figure 7C assigns enriched biological-process categories to the affected genes. These experiments establish transcript-output regulation, but do not directly test altered signal transmission or a defined signaling response. Assigning regulation of signal transduction from this enrichment extends the measured result to a downstream process. Mark the source-specific process assertion as over-annotated while retaining CDK13 kinase and transcriptional functions; this is not a wrong-gene attribution or a denial of every possible signaling role.

Summary: Nuclear speckles are a functional localization of CDK13.

Reason: The original human CDC2L5 localization study places the protein in nucleoplasm with enrichment in speckles and tests dependence on the N-terminal RS region. This independently supports the immunofluorescence or automated source while retaining its exact provenance; no RNA-cleavage activity is inferred from localization.

Summary: Nuclear speckles are a functional localization of CDK13.

Reason: The original human CDC2L5 localization study places the protein in nucleoplasm with enrichment in speckles and tests dependence on the N-terminal RS region. This independently supports the immunofluorescence or automated source while retaining its exact provenance; no RNA-cleavage activity is inferred from localization.

Summary: Nuclear speckles are a functional localization of CDK13.

Reason: The original human CDC2L5 localization study places the protein in nucleoplasm with enrichment in speckles and tests dependence on the N-terminal RS region. This independently supports the immunofluorescence or automated source while retaining its exact provenance; no RNA-cleavage activity is inferred from localization.

Summary: Kinase association has independent human target evidence.

Reason: Mouse donor Q69ZA1 has a kinase-binding IDA citing PMID:22547058. Selected full Figure 5 establishes cyclin K-associated kinase recovery, which alone does not settle a binary CDK13–CDK12 interaction. Independently, exact human BioPlex records recover CDK12 with CDK13 (EBI-54717052 and EBI-55012677). Retain protein kinase binding at this non-core association scope, without claiming a direct interface, common stable cyclin K holoenzyme, or reciprocal kinase activation. The separate holoenzymes demonstrated in PMID:22012619 remain distinct.

Summary: CDK13 belongs to a nuclear cyclin-dependent kinase holoenzyme.

Reason: The mouse ortholog Q69ZA1 is the genuine source, with experimental nuclear-complex evidence linked to PMID:17261272. Independent human cyclin K/CDK13 isolation and nuclear kinase function support the transferred broad complex membership. This does not equate mouse L-type cyclin constructs with the exact human cyclin K holoenzyme. Treat the corresponding conserved association at the asserted broad level.

Summary: CDK13 belongs to a nuclear cyclin-dependent kinase holoenzyme.

Reason: The mouse ortholog Q69ZA1 is the genuine source, with experimental nuclear-complex evidence linked to PMID:17261272. Independent human cyclin K/CDK13 isolation and nuclear kinase function support the transferred broad complex membership. This does not equate mouse L-type cyclin constructs with the exact human cyclin K holoenzyme. Treat the corresponding conserved association at the asserted broad level.

Summary: The historical perturbation supports a hematopoietic differentiation context.

Reason: The source cloned human CHED but its antisense phenotype was selective inhibition of megakaryocyte development in murine bone-marrow cultures. Retain the existing broad non-core hematopoietic association with that organism and assay boundary; do not describe it as a direct human marrow experiment or infer an intrinsic differentiation enzyme.

Summary: Cyclin binding is integral to the CDK13 catalytic complex.

Reason: The regulatory cyclin K physically associates with human CDK13 in reciprocal immunoprecipitation and the purified complex structure. This independently supports cyclin binding while preserving the original electronic or IPI assertion. Broader L-type cyclin evidence has its separate mouse/construct scope.

Summary: Cyclin binding is integral to the CDK13 catalytic complex.

Reason: The regulatory cyclin K physically associates with human CDK13 in reciprocal immunoprecipitation and the purified complex structure. This independently supports cyclin binding while preserving the original electronic or IPI assertion. Broader L-type cyclin evidence has its separate mouse/construct scope.

Summary: CDK13 supports the transcription-elongation program through CTD phosphorylation.

Reason: The conserved transcriptional-CDK role is supported by human CDK13 CTD phosphorylation and by target-specific knockdown/rescue effects on transcript output. The original studies do not uniformly measure elongation velocity; their phosphoregulation and output assays, rather than an invented direct rate measurement, ground this existing process. Retain the original evidence and the nuclear kinase core.

Summary: CDK13 supports the transcription-elongation program through CTD phosphorylation.

Reason: The conserved transcriptional-CDK role is supported by human CDK13 CTD phosphorylation and by target-specific knockdown/rescue effects on transcript output. The original studies do not uniformly measure elongation velocity; their phosphoregulation and output assays, rather than an invented direct rate measurement, ground this existing process. Retain the original evidence and the nuclear kinase core.

Summary: CDK13 supports the transcription-elongation program through CTD phosphorylation.

Reason: The conserved transcriptional-CDK role is supported by human CDK13 CTD phosphorylation and by target-specific knockdown/rescue effects on transcript output. The original studies do not uniformly measure elongation velocity; their phosphoregulation and output assays, rather than an invented direct rate measurement, ground this existing process. Retain the original evidence and the nuclear kinase core.

Summary: CDK13 supports RNA-polymerase-II transcriptional output.

Reason: The original paper demonstrates human CTD kinase activity while explicitly limiting its in-vivo knockdown conclusion. Later human CDK13 knockdown and RNAi-resistant rescue support transcript-output regulation. Retain the broad existing positive-transcription process on the combined evidence, without converting CDK12-specific in-vivo experiments into CDK13 assays.

Summary: CDK13 perturbation changes cell-cycle phase distributions in human cells.

Reason: The existing GO_REF:0000108 assertion follows the actual ontology path from cyclin-dependent kinase activity through a part_of relationship to cell-cycle regulation; it is not a malformed propagation. Independent human 786-O experiments in PMID:41680470 Figure 2d compare CDK13 knockdown and overexpression with their controls and demonstrate altered endpoint phase distributions. These findings support a non-core regulatory contribution without assigning a direct checkpoint substrate, measuring transition rates, or relying on the disputed S-phase direction. The reference finding records the Results-versus-figure discrepancy explicitly, and the source-evidence artifact gives the publisher PDF locator and displayed values. Neither a doubling-time phenotype nor a nonselective inhibitor supplies this conclusion.

Summary: Protein-serine phosphorylation is directly established.

Reason: The Rhea serine-kinase assertion is supported by human CDK13 CTD serine phosphorylation and independent CDK13/cyclin T1 phosphorylation of ZC3H14 S475. The latter experiment compares full-length wild-type and S475A substrate in vitro and links the site to RNA-surveillance-complex recruitment in human melanoma cells. CDK13 is the kinase, not the RNA-degrading enzyme.

Summary: The Reactome target record supports a contextual granule-release location.

Reason: The exact event R-HSA-6800434 connects the ficolin-rich granule set R-HSA-6800431 to extracellular set R-HSA-6806481; both contain a specific human CDK13 entity. Retain the curated ficolin-1-rich granule lumen assignment as non-core. This is curated proteomic/event context, not an independent demonstration of CDK13 secretion chemistry or an extracellular catalytic role.