DES

Summary: Inherited intermediate-filament structural activity.

Reason: Human DES lies beneath the PTN000580114 IBD node in PTHR45652. The structural assignment agrees with direct desmin assembly and the target filament network; the mixed descendant evidence is support for an ancestral structural role, not a donor-count test.

Summary: Desmin supplies the intermediate-filament scaffold.

Reason: The full original combines human patient pathology with assembly and cell-expression experiments using hamster desmin carrying the corresponding deletion. That homologous mechanistic evidence, together with direct human desmin assembly in PMID:28470624, supports structural constituent activity without treating the hamster constructs as human protein.

Summary: Desmin binds the cytoskeletal protein synemin/DMN.

Reason: The original uses full-length human desmin and translated DMN rod constructs in cell-free co-immunoprecipitation, with nonbinding DMN fragments as controls. Synemin is a cytoskeletal intermediate-filament protein, supporting the partner-class refinement. The assay does not establish binding specifically to an assembled filament.

Proposed replacements: cytoskeletal protein binding

Summary: Desmin binds epiplakin fragments.

Reason: The scanned original tests human epiplakin GST fragments against human recombinant desmin in slot-blot assays, with GST and BSA controls. Epiplakin is independently characterized as controlling intermediate-filament reorganization. Thus binding to a cytoskeletal regulatory protein is supported; the immobilized assay and weaker desmin signal do not establish a native stoichiometric complex or make DES itself a regulator.

Proposed replacements: cytoskeletal regulatory protein binding

Summary: Desmin associates with MLH1.

Reason: The available abstract explicitly reports the interaction, and exact human DES–MLH1 IntAct records include co-immunoprecipitation in HeLa and HEK293T cells in addition to a two-hybrid record. Full original access failed, so unseen assay detail is not asserted. The independent curated pair records support retaining the physical association. MLH1 participation in a repair complex does not assign the partner PMS2 endonuclease activity to MLH1 or a repair function to DES.

Summary: Desmin binds myotubularin, which modulates desmin filament assembly.

Reason: The original reports human MTM1 bait/human muscle-library DES prey and additional biochemical and cell assays. Myotubularin alters desmin assembly, establishing a cytoskeletal regulatory partner. Desmin did not alter the assayed MTM1 phosphatase activity; the refinement concerns the bound partner class, not DES phosphatase regulation.

Proposed replacements: cytoskeletal regulatory protein binding

Summary: Desmin binds nebulin cytoskeletal modules.

Reason: Human DES constructs and human nebulin M160–164 were used in purified binding assays, including co-sedimentation and ELISA with protein-alone/BSA controls. The interaction supports cytoskeletal protein binding. Chick, rat and canine localization experiments are separate evidence, and nebulin binding does not establish direct actin binding by DES.

Proposed replacements: cytoskeletal protein binding

Summary: Plectin–desmin binding is retained with an explicit access limit.

Reason: The publisher-indexed body explicitly reports desmin binding to plectin fragments as data not shown, and the official DES record identifies the pair. The complete original and supplement were not recovered, so unseen assay details and exact fragment boundaries are not asserted. Independently, the official plectin record establishes a structural cytoskeletal component that interlinks filaments. This supports cytoskeletal protein binding as a partner-class refinement while preserving curator deference for the interaction; a more specific cytoskeletal-regulatory claim is not required.

Proposed replacements: cytoskeletal protein binding

Summary: Desmin binds the catalytic protein EHHADH.

Reason: The exact EHHADH–DES record is present in the 2014 map with array, pooled and validated two-hybrid records. The original specifies pairwise quadruplicate retesting and sequence confirmation. EHHADH is a bifunctional metabolic enzyme, supporting enzyme binding as a partner-class refinement; no peroxisomal localization or fatty-acid catalytic activity is assigned to DES. This term describes binding to the established partner class; it does not require or assert that DES activates, inhibits or performs the partner's catalytic or regulatory activity.

Proposed replacements: enzyme binding

Summary: Desmin binds the actin regulator phostensin.

Reason: The 2014 map has the exact PPP1R18–DES pair with validated two-hybrid evidence. Official phostensin functional annotation describes actin-filament pointed-end capping and modulation of actin dynamics. This supports binding to a cytoskeletal regulatory protein without assigning actin capping or phosphatase catalysis to DES.

Proposed replacements: cytoskeletal regulatory protein binding

Summary: Desmin binds the recorded UBE2I product Q7KZS0.

Reason: The exact positive, validated two-hybrid record maps to unreviewed 184-residue Q7KZS0, not the 158-residue reviewed canonical UBE2I sequence. The sequences share their first 137 residues but diverge at the C terminus. Binding is supported for the recorded product; predicted UBC-domain/active-site features alone do not establish that this tested product has the complete catalytic properties of canonical UBE2I, so no enzyme-class refinement is imposed.

Summary: Desmin binds the DUX1 double-homeodomain protein.

Reason: The original identifies human DES fragments in the DUX1 screen and confirms interaction of full human GST-DES with translated DUX1; GST and luciferase controls are reported. The DES 260–470 fragment can bind one DUX1 homeodomain. DUX1 lacked the strong transactivation observed for DUX4 in the tested yeast context; homeodomain binding alone does not establish a defined transcriptional function for this DES association.

Summary: Desmin binds the transcription factor DUX4.

Reason: The original confirms binding of human GST-DES to translated DUX4, with GST/luciferase controls and a negative DUX4-tail construct. The mouse-library screen and mouse-cell co-immunoprecipitation are distinguished from this human-protein assay. DUX4 is an established transcription factor, allowing a partner-class refinement; this does not assign transcriptional regulation to DES or infer every DUX4 isoform behaves identically. The narrower DNA-binding transcription factor binding class is supported by the DUX4 double-homeodomain protein and its demonstrated transcriptional activity; this binding result does not establish transcriptional regulation by DES.

Proposed replacements: DNA-binding transcription factor binding

Summary: Desmin binds a neurofilament-light construct.

Reason: The isoform interaction study has exact positive DES–NEFL array/pooling/retest records. The NEFL participant feature is a sufficient binding region at residues 1–377, so the record is not represented as a full-length native heterofilament. NEFL is a cytoskeletal protein, supporting that binding class without asserting binding to an assembled intermediate filament.

Proposed replacements: cytoskeletal protein binding

Summary: Desmin binds neurofilament light in a complementation assay.

Reason: The exact PMID:27107012 pair record is split-luciferase complementation in human cells. This is not automatically described as the BFG two-hybrid method used for other pairs in the paper. NEFL identity supports cytoskeletal protein binding, while complementation does not establish native filament stoichiometry.

Proposed replacements: cytoskeletal protein binding

Summary: Desmin binds vimentin.

Reason: The exact pair is supported by barcode-fusion genetics and validated two-hybrid records. Vimentin is a cytoskeletal intermediate-filament protein; the supported refinement describes binding to that protein, without claiming that the assayed entity was an assembled filament.

Proposed replacements: cytoskeletal protein binding

Summary: Desmin binds the cytoskeletal protein nebulette.

Reason: The original reports purified desmin–nebulette binding, domain assays, and distinct chick/rat and human-cell contexts. The human RD cell co-immunoprecipitation has limited bait detection and is not treated as the sole proof. Nebulette is a sarcomeric cytoskeletal protein; the proposed actin-coordination model does not establish direct actin binding by DES.

Proposed replacements: cytoskeletal protein binding

Summary: Desmin binds alpha-B-crystallin during filament assembly.

Reason: The full original directly tests purified human desmin and human CRYAB by co-sedimentation, electron microscopy, timed addition and viscometry. CRYAB is the protein-folding chaperone partner; the interaction does not make DES a chaperone. No exact desmin binding-site boundary is asserted because the original text and figure-caption descriptions differ.

Proposed replacements: protein-folding chaperone binding

Summary: Desmin co-purifies with vimentin in BioPlex 2.0.

Reason: The exact pair is a spoke-expanded anti-tag affinity-purification association. The matched evidence is in HEK293T cells. vimentin is an established cytoskeletal protein, so cytoskeletal protein binding refines the already-curated physical association without adding a claim of direct contact, assembled-filament recognition or native stoichiometry. Gene-level scoring and tagged-bait assays do not establish every isoform combination. Independent DES–VIM binary assays in PMID:27107012 and PMID:32296183 also support the pair.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–GOLGA6L9 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The recorded partner fragment spans residues 159–432. Its golgin-family designation does not establish a specific scaffold or cytoskeletal regulatory function for the bound fragment. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–periplakin interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. Periplakin is a cornified-envelope component with proposed linkage to intermediate filaments. This pair assay does not establish a defined adaptor function or the higher-order filament substrate. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–PIAS2-alpha interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The participant is O75928-2, the 572-residue PIAS2-alpha product, not PIAS2-beta. Official isoform-specific function describes SUMO ligase activity for alpha, supporting ubiquitin-like protein ligase binding; DES sumoylation is not inferred. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed. This term describes binding to the established partner class; it does not require or assert that DES activates, inhibits or performs the partner's catalytic or regulatory activity.

Proposed replacements: ubiquitin-like protein ligase binding

Summary: HuRI supports a physical DES–KRT37 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. KRT37 is a keratin cytoskeletal protein. The interaction supports binding to that protein, not necessarily an assembled keratin filament. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–KRT75 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. KRT75 is a keratin cytoskeletal protein. A physical pair does not itself establish a native heterofilament or shared expression in every muscle cell. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–NEFL interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. NEFL is a cytoskeletal intermediate-filament protein. This binding-class refinement does not assert a native neurofilament assembly in muscle. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–YES1 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. YES1 is an established protein kinase. Binding to this partner supports protein kinase binding without asserting that DES is a kinase or a demonstrated YES1 substrate. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed. This term describes binding to the established partner class; it does not require or assert that DES activates, inhibits or performs the partner's catalytic or regulatory activity.

Proposed replacements: protein kinase binding

Summary: HuRI supports a physical DES–vimentin interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. Vimentin is a cytoskeletal intermediate-filament protein; the tested pair supports protein-component binding, not a claim about assembled filament recognition. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–KRT13 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. KRT13 is a keratin cytoskeletal protein. The refinement describes the assayed protein class without assigning epithelial differentiation to DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–GFAP interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. GFAP is a cytoskeletal intermediate-filament protein. The binding result does not assign an astrocyte-specific biological role to DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–KRT15 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. KRT15 is a keratin cytoskeletal protein. Native tissue coexistence and filament stoichiometry were not established by this reference-map assay. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–DDX6 interaction.

Reason: The exact HuRI DES–DDX6 pair includes validated two-hybrid evidence. The current GO:0051117 definition covers binding to an enzyme that hydrolyzes ATP, and the official DDX6 record explicitly assigns ATP + water to ADP + phosphate with curated evidence ECO:0000305. That established ATPase partner class supports a narrower refinement than enzyme binding. This does not assert a new DES catalytic activity, RNA unwinding by DES, modulation of DDX6 ATP turnover, or a native P-body function. The official catalytic assignment is not represented as an independently reread purified-enzyme experiment. This term describes binding to the established partner class; it does not require or assert that DES activates, inhibits or performs the partner's catalytic or regulatory activity. DDX6 also has a curated DEAD-box family assignment and PROSITE-supported helicase domains; the class assessment is not inferred from a GO label or protein name alone.

Proposed replacements: ATPase binding

Summary: HuRI supports a physical DES–KRT20 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. KRT20 is a keratin cytoskeletal protein. Binding to the protein is supported, while assembled keratin-filament recognition was not tested. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–peripherin interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. Peripherin is a cytoskeletal intermediate-filament protein. This is a binding-class statement, not an assignment of neuronal differentiation to DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–LBX1 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. LBX1 is an established transcription factor. The partner class supports transcription factor binding without assigning transcriptional activity or developmental target regulation to DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed. LBX1 is a homeobox DNA-binding transcription factor, supported by its curated function and PROSITE homeobox domain. GO:0140297 therefore specifies the partner class without assigning transcriptional regulation to DES.

Proposed replacements: DNA-binding transcription factor binding

Summary: HuRI supports a physical DES–LMO4 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. LMO4 is a transcriptional regulatory factor. The current broad transcription factor binding definition includes proteins required to regulate transcription; this does not assert DNA binding by LMO4 or DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed. This term describes binding to the established partner class; it does not require or assert that DES activates, inhibits or performs the partner's catalytic or regulatory activity.

Proposed replacements: transcription factor binding

Summary: HuRI supports a physical DES–POLR2G/RPB7 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. RPB7 is a noncatalytic RNA-polymerase subunit. Membership in that enzyme complex does not make this isolated partner the polymerase catalytic unit, so enzyme binding is not inferred. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–RNF4 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. RNF4 is an E3 ubiquitin-protein ligase. The partner class supports ubiquitin protein ligase binding without asserting DES ubiquitination, degradation or ligase regulation. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed. This term describes binding to the established partner class; it does not require or assert that DES activates, inhibits or performs the partner's catalytic or regulatory activity.

Proposed replacements: ubiquitin protein ligase binding

Summary: HuRI supports a physical DES–EHHADH interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. EHHADH is a catalytic bifunctional metabolic protein. The refinement does not assign its catalytic reactions, metabolic pathway or peroxisomal compartment to DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed. This term describes binding to the established partner class; it does not require or assert that DES activates, inhibits or performs the partner's catalytic or regulatory activity.

Proposed replacements: enzyme binding

Summary: HuRI supports a physical DES–KRT33B interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. KRT33B is a keratin cytoskeletal protein. The result supports binding to the protein, while assembled filament recognition and native tissue stoichiometry remain unestablished. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal protein binding

Summary: HuRI supports a physical DES–TCEA2 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. TCEA2 regulates RNA-polymerase elongation. The current broad transcription factor binding definition covers transcription-regulating proteins; this does not assign polymerase catalysis or transcriptional activation to DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed. This term describes binding to the established partner class; it does not require or assert that DES activates, inhibits or performs the partner's catalytic or regulatory activity.

Proposed replacements: transcription factor binding

Summary: HuRI supports a physical DES–CWF19L2 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The recorded CWF19L2 feature spans residues 155–893. Its spliceosome role does not establish catalytic RNA activity or a specific adaptor role for the DES interaction. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–LONRF3 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The partner name and domain architecture do not by themselves establish the catalytic or regulatory properties of this interaction. No ligase or protease activity is assigned to DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–phostensin/PPP1R18 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. Official phostensin function describes regulation of actin-filament dynamics through pointed-end capping. Binding to that regulatory protein is supported; DES is not thereby assigned actin-capping or phosphatase activity. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Proposed replacements: cytoskeletal regulatory protein binding

Summary: HuRI supports a physical DES–C1orf216 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The official record does not establish a specific functional partner class that would justify a narrower binding term. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–M1AP interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. A role for M1AP in meiosis does not establish a specific molecular binding class or participation of DES in meiosis. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–CT55 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The positive pair is recorded in both bait/prey orientations. CT55 disease or spermatogenesis associations do not define a specific binding function for DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–PIH1D2 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The PIH1 domain-containing partner name alone does not establish a chaperone, enzyme or adaptor role for the tested association. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–LENG1 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The available partner functional information does not justify assigning a narrower molecular function to the physical association. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–NEK6 isoform 4 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The participant is Q9HC98-4 and the recorded binding region is residues 26–338. Reconstructing the 25-residue N-terminal extension shows this span retains the canonical kinase domain; protein kinase binding is justified without assigning DES phosphorylation or NEK6 regulation. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed. This term describes binding to the established partner class; it does not require or assert that DES activates, inhibits or performs the partner's catalytic or regulatory activity.

Proposed replacements: protein kinase binding

Summary: HuRI supports a physical DES–TFIP11 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. TFIP11 spliceosome association does not establish an intrinsic RNA-catalytic activity or a specific molecular adaptor function for the DES pair. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–ZMYM5 isoform 2 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. The exact participant Q9UJ78-2 is 208 residues and differs from the displayed 669-residue product. The general full-gene transcriptional-regulator description cannot establish that activity for this short tested product, so no transcription-factor-class refinement is imposed. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: HuRI supports a physical DES–BAG5 interaction.

Reason: The exact partner/reference records include validated two-hybrid evidence, and the original describes quadruplicate retesting and sequence confirmation. BAG5 is a molecular-chaperone regulator. Binding a co-chaperone does not itself establish binding to a protein-folding chaperone or a chaperone activity of DES. This is a biophysical assay result; native cell coexistence and stoichiometry are not assumed.

Summary: Desmin binds GFAP in the neurodegeneration interaction map.

Reason: The exact pair has positive array, pooled and validated two-hybrid records. The original describes autoactivation controls and yeast interaction assays; subset mammalian validation is not assigned to this pair without a matching record. GFAP is a cytoskeletal intermediate-filament protein, supporting the protein-binding class without claiming recognition of an assembled filament.

Proposed replacements: cytoskeletal protein binding

Summary: Desmin binds peripherin in the neurodegeneration interaction map.

Reason: The exact pair has positive array, pooled and validated two-hybrid records. The original describes autoactivation controls and yeast interaction assays; subset mammalian validation is not assigned to this pair without a matching record. peripherin is a cytoskeletal intermediate-filament protein, supporting the protein-binding class without claiming recognition of an assembled filament.

Proposed replacements: cytoskeletal protein binding

Summary: Desmin binds tested huntingtin fragments.

Reason: The exact records support HTT regions 1–88, 1–511 and 1932–2642 by array/pooling and validated two-hybrid assays. A checksum identifying the reference HTT sequence does not mean the entire full-length protein was tested. The fragment results support physical association without assigning huntingtin transport/scaffold functions or a disease-modifier role to DES.

Summary: Desmin binds junctophilin-3 in the interaction map.

Reason: The exact JPH3 pair has validated two-hybrid evidence. Junctophilin forms membrane junctions, but that does not automatically classify this interaction as cytoskeletal regulation or give DES a calcium-channel regulatory function. Native membrane geometry was not established by the yeast assay.

Summary: BioPlex 3.0 records DES association with NEFL.

Reason: The exact pair is a spoke-expanded anti-tag affinity-purification association. The matched evidence is in HEK293T cells. NEFL is an established cytoskeletal protein, so cytoskeletal protein binding refines the already-curated physical association without adding a claim of direct contact, assembled-filament recognition or native stoichiometry. Gene-level scoring and tagged-bait assays do not establish every isoform combination.

Proposed replacements: cytoskeletal protein binding

Summary: BioPlex 3.0 records DES association with VIM.

Reason: The exact pair is a spoke-expanded anti-tag affinity-purification association. The matched evidence includes HEK293T and HCT116 contexts. VIM is an established cytoskeletal protein, so cytoskeletal protein binding refines the already-curated physical association without adding a claim of direct contact, assembled-filament recognition or native stoichiometry. Gene-level scoring and tagged-bait assays do not establish every isoform combination. Independent DES–VIM binary assays in PMID:27107012 and PMID:32296183 also support the pair.

Proposed replacements: cytoskeletal protein binding

Summary: BioPlex 3.0 records DES association with GFAP.

Reason: The exact pair is a spoke-expanded anti-tag affinity-purification association. The matched evidence is in HEK293T cells. GFAP is an established cytoskeletal protein, so cytoskeletal protein binding refines the already-curated physical association without adding a claim of direct contact, assembled-filament recognition or native stoichiometry. Gene-level scoring and tagged-bait assays do not establish every isoform combination.

Proposed replacements: cytoskeletal protein binding

Summary: BioPlex 3.0 records DES association with PRPH.

Reason: The exact pair is a spoke-expanded anti-tag affinity-purification association. The matched evidence is in HEK293T cells. PRPH is an established cytoskeletal protein, so cytoskeletal protein binding refines the already-curated physical association without adding a claim of direct contact, assembled-filament recognition or native stoichiometry. Gene-level scoring and tagged-bait assays do not establish every isoform combination.

Proposed replacements: cytoskeletal protein binding

Summary: Context-dependent nuclear localization is retained.

Reason: The source is mouse desmin P31001. The full PMID:26787680 original detects endogenous and tagged murine desmin in a minority of differentiating cardiac progenitors and immature cardiomyocytes, using desmin-null antibody controls and purified nuclei. Mature cardiomyocyte nuclei lack that signal. Transfer to human is plausible but remains developmental orthology evidence, not a constitutive nuclear core location.

Summary: Context-dependent nuclear localization is retained.

Reason: The source is mouse desmin P31001. The full PMID:26787680 original detects endogenous and tagged murine desmin in a minority of differentiating cardiac progenitors and immature cardiomyocytes, using desmin-null antibody controls and purified nuclei. Mature cardiomyocyte nuclei lack that signal. Transfer to human is plausible but remains developmental orthology evidence, not a constitutive nuclear core location.

Summary: Desmin anchors at the nuclear envelope in cardiac cells.

Reason: The mouse donor has direct localization evidence in PMID:35959657: super-resolution imaging measures desmin proximity to lamin-B1-defined nuclear envelopes in controls and PKP2-deficient cardiomyocytes. Independent rat desmin-depletion experiments in PMID:31822208 support the structural connection. This is envelope-associated filament anchoring, not an integral nuclear-membrane protein assignment.

Summary: Desmin anchors at the nuclear envelope in cardiac cells.

Reason: The mouse donor has direct localization evidence in PMID:35959657: super-resolution imaging measures desmin proximity to lamin-B1-defined nuclear envelopes in controls and PKP2-deficient cardiomyocytes. Independent rat desmin-depletion experiments in PMID:31822208 support the structural connection. This is envelope-associated filament anchoring, not an integral nuclear-membrane protein assignment.

Summary: Desmin has a cytoplasmic filament network and disease-associated aggregates.

Reason: The original directly compares human R350P muscle biopsy material with R349P mouse models, using mutation-specific and pan-desmin antibodies. Human subsarcolemmal and sarcoplasmic mutant aggregates support this broad location; pathological aggregation is distinguished from the normal filament network. Cytoplasm is retained as a broad location rather than a separate functional core.

Summary: Broad cytoplasmic localization is consistent with the desmin network.

Reason: The mouse donor and official target subcellular record support cytoplasmic desmin. The current ARBA00026971 record emits this term, but no complete target-matching clause was reconstructed from the available InterPro/taxonomy inputs; that rule limitation does not contradict independently established cytoplasmic localization.

Summary: Reactome places DES-containing sarcomeric assemblies in cytosol.

Reason: The cited reaction contains a sarcomeric complex with a typed hasComponent path through R-HSA-390588 (Z Disk Complex) to R-HSA-390570, whose reference entity is human P17661 residues 1–470 in cytosol. The location is consistent with desmin. Reaction names describing myosin ATP turnover or troponin calcium binding do not assign those catalytic/binding activities to DES.

Summary: Reactome places DES-containing sarcomeric assemblies in cytosol.

Reason: The cited reaction contains a sarcomeric complex with a typed hasComponent path through R-HSA-390588 (Z Disk Complex) to R-HSA-390570, whose reference entity is human P17661 residues 1–470 in cytosol. The location is consistent with desmin. Reaction names describing myosin ATP turnover or troponin calcium binding do not assign those catalytic/binding activities to DES.

Summary: Reactome places DES-containing sarcomeric assemblies in cytosol.

Reason: The cited reaction contains a sarcomeric complex with a typed hasComponent path through R-HSA-390588 (Z Disk Complex) to R-HSA-390570, whose reference entity is human P17661 residues 1–470 in cytosol. The location is consistent with desmin. Reaction names describing myosin ATP turnover or troponin calcium binding do not assign those catalytic/binding activities to DES.

Summary: Reactome places DES-containing sarcomeric assemblies in cytosol.

Reason: The cited reaction contains a sarcomeric complex with a typed hasComponent path through R-HSA-390588 (Z Disk Complex) to R-HSA-390570, whose reference entity is human P17661 residues 1–470 in cytosol. The location is consistent with desmin. Reaction names describing myosin ATP turnover or troponin calcium binding do not assign those catalytic/binding activities to DES.

Summary: Cytoskeleton is a broad parent location.

Reason: The mouse-derived cytoskeleton assignment is biologically correct. Direct target assembly and the HPA structural location identify the intermediate filament cytoskeleton more specifically; its current definition and direct is_a parent GO:0005856 match the network scope. Refine the existing source assertion, even though the specific term also occurs elsewhere; no donor error or NEW annotation is implied.

Proposed replacements: intermediate filament cytoskeleton

Summary: Desmin belongs to the intermediate-filament structure.

Reason: The target is a descendant of the PTN000580114 IBD node. The conserved filament location agrees with direct human desmin assembly and the endogenous muscle network; the inference is evaluated by ancestry and target biology, not by donor number.

Summary: Orthology and intermediate-filament domains support the filament location.

Reason: Mouse desmin and the target InterPro intermediate-filament match agree with direct human assembly evidence. The sequence/domain transfer preserves the structural role and there is no target-specific loss evidence.

Summary: Intermediate filaments are the defining desmin structure.

Reason: The cited original studies human tissue and homologous hamster desmin assembly/expression. The species distinction limits the construct claim, while independent purified human desmin assembly directly confirms the same filament location.

Summary: Cell-cell junction is a broad location for cardiac desmin.

Reason: Human DES is within the PTN002901991 clade and its own junctional evidence among IBD descendants is legitimate grounding. Existing human intercalated-disc evidence identifies the precise muscle cell-cell junction. Current typed ancestry places intercalated disc beneath cell-cell junction through cell-cell contact zone. Refine the existing assertion; no duplicate NEW is added.

Proposed replacements: intercalated disc

Summary: Cell-cell junction is a broad location for cardiac desmin.

Reason: Mouse desmin is an appropriate structural donor and independent human evidence establishes localization to cardiac intercalated discs. That term specifies the relevant muscle cell-cell junction and has a checked is_a path to GO:0005911. Refine this existing source assertion without alleging lineage divergence or adding a NEW annotation.

Proposed replacements: intercalated disc

Summary: Fascia-adherens association is retained as a junctional context.

Reason: The source is mouse desmin P31001 with an IDA linked to PMID:11732910. Its available abstract explicitly identifies desmin in an enriched mouse cardiac intercalated-disc/fascia-adherens fraction. The complete original was not recovered; the existing transfer is retained in this bounded structural context rather than asserted as an independently repeated human assay.

Summary: Desmin supports muscle contraction through muscle structure.

Reason: The full PMID:9697706 original establishes human genetics and muscle/cardiac disease, not a direct contraction assay. Established filament assembly and muscle architecture make structural support of contraction plausible, but the phenotype is not treated as motor or ATPase activity or as a separate core mechanism.

Summary: Desmin contributes to nuclear-envelope organization through mechanical anchoring.

Reason: The current mouse donor has an IMP assignment from the PKP2 study, which directly observes desmin anchoring but perturbs PKP2. Independent full-original rat experiments in PMID:31822208 directly deplete desmin and show nuclear-envelope/lamina infolding within a preserved cell frame, with microtubule-dynamicity controls. This supports conserved structural participation rather than a generic downstream phenotype, while human transfer and cardiac context remain explicit.

Summary: Desmin contributes to nuclear-envelope organization through mechanical anchoring.

Reason: The current mouse donor has an IMP assignment from the PKP2 study, which directly observes desmin anchoring but perturbs PKP2. Independent full-original rat experiments in PMID:31822208 directly deplete desmin and show nuclear-envelope/lamina infolding within a preserved cell frame, with microtubule-dynamicity controls. This supports conserved structural participation rather than a generic downstream phenotype, while human transfer and cardiac context remain explicit.

Summary: Intermediate-filament organization is the more informative existing process.

Reason: The original addresses disruption of desmin assembly and the muscle filament network, using hamster constructs and human tissue with those contexts distinguished. Intermediate filament organization describes this structural process more precisely than cytoskeleton organization and has a checked is_a path through intermediate filament cytoskeleton organization. Refine this source assertion even though the term is already present on other source rows.

Proposed replacements: intermediate filament organization

Summary: Cardiac contraction effects are retained as a physiological consequence.

Reason: The cited original is a human genetics/segregation report, not an assay of a distinct heart-contraction regulator. DES supports cardiac architecture and its disruption can affect cardiac performance. Retain this inherited TAS statement as a non-core physiological association without turning it into a separate regulatory molecular mechanism.

Summary: Desmin binds nebulin, a cytoskeletal partner.

Reason: The original screens human skeletal-muscle cDNA with rabbit nebulin constructs and repeatedly identifies DES; the annotated partner Q04020 is rabbit nebulin. The follow-up fragment screen supports this pair. Myopalladin is a different interaction in the same paper, and its GST validation is not reassigned to DES.

Summary: Desmin localizes at cardiac intercalated discs.

Reason: The existing IDA agrees with the established junctional filament network. The available PMID:24200904 abstract explicitly describes intercalated-disc staining changes while Z-band localization is preserved. Independent structural context supports curator deference; unseen imaging methods, constructs or controls are not invented.

Summary: Desmin localizes at cardiac intercalated discs.

Reason: The existing IDA agrees with the established junctional filament network. The PMID:26724190 abstract describes desmin distribution and filament-assembly defects, but the full original was not recovered. Independent structural context supports curator deference; unseen imaging methods, constructs or controls are not invented.

Summary: Inherited Z-disc localization matches the human muscle network.

Reason: Human DES descends from PTN002901991 and direct human muscle electron microscopy in PMID:11353857 places desmin around Z-lines. Target evidence legitimately contributes to the IBD; it is not circular propagation.

Summary: Desmin is localized around muscle Z discs.

Reason: This existing IDA is consistent with independently read human muscle electron microscopy in PMID:11353857. The PMID:24200904 abstract also explicitly describes preserved Z-band localization. The reliable, independently established location is retained by curator deference; a title emphasizing another protein does not establish misattribution.

Summary: Desmin is localized around muscle Z discs.

Reason: This existing IDA is consistent with independently read human muscle electron microscopy in PMID:11353857. The cited full original was not recovered, so no unseen source-specific imaging details are asserted. The reliable, independently established location is retained by curator deference; a title emphasizing another protein does not establish misattribution.

Summary: Desmin is localized around muscle Z discs.

Reason: This existing IDA is consistent with independently read human muscle electron microscopy in PMID:11353857. The cited full original was not recovered, so no unseen source-specific imaging details are asserted. The reliable, independently established location is retained by curator deference; a title emphasizing another protein does not establish misattribution.

Summary: Mouse orthology supports the conserved Z-disc network.

Reason: The mouse P31001 record and target subcellular annotation agree with direct human muscle ultrastructural localization around Z-lines. This is a conserved structural site, not an inference based solely on sequence similarity.

Summary: Neuromuscular-junction localization remains contextual.

Reason: The mouse donor has IDA support from PMID:11694502. The available abstract reports syncoilin–desmin interaction and colocalization at the mouse neuromuscular junction; it also distinguishes that interaction from heterofilament formation in COS-7 cells. The full original was not recovered. Retain the established donor location as bounded orthology evidence, not a new DES junction-assembly mechanism.

Summary: Sarcolemmal association is part of the desmin structural network.

Reason: The target lies within PTN002901991, and the IBD includes human DES experimental grounding. Sarcolemma-associated desmin links the peripheral filament network to muscle-cell architecture; the assignment does not mean desmin is a transmembrane protein. Independent PMID:21135508 reports normal skeletal-muscle MTM1–desmin colocalization at the sarcolemma, with mouse and human muscle distinguished. The human supplemental images were not independently reanalysed. This corroborates the peripheral structural site without turning the R350P aggregate experiment into normal topology evidence. The mouse descendant has a separate sarcolemma IDA from PMID:11694502; its original remains unavailable, so no unseen methods are claimed.

Summary: Sarcolemmal association is a core desmin location; this source examines R350P muscle.

Reason: The full original distinguishes human R350P biopsy material from mouse R349P models and uses a mutation-specific antibody. Human mutant desmin is prominent in subsarcolemmal aggregates. Accept the established gene-level peripheral sarcolemmal location, consistent with the other sarcolemma rows and the structural core, while retaining this experiment’s disease/variant scope. Normal skeletal-muscle localization is independently corroborated by PMID:21135508. The R350P experiment is not recast as an assay of normal filament distribution or membrane insertion.

Summary: Sarcolemmal association is conserved between mouse and human desmin.

Reason: The mouse donor and official target subcellular record are consistent with the peripheral muscle intermediate-filament network. The transfer describes structural association at the sarcolemma, not integral membrane topology. Independent PMID:21135508 reports normal skeletal-muscle MTM1–desmin colocalization at the sarcolemma, with mouse and human muscle distinguished. The human supplemental images were not independently reanalysed. This corroborates the peripheral structural site without turning the R350P aggregate experiment into normal topology evidence.

Summary: Desmin self-association supports filament assembly.

Reason: The original assays desmin assembly and its modulation by myotubularin. Homomeric association is a direct component of the filament-building mechanism, not merely an arbitrary partner interaction.

Summary: HuRI directly supports DES self-association.

Reason: The exact DES–DES pair has positive array/pooling and validated two-hybrid records, with both products matching canonical P17661. Independent filament assembly makes self-association mechanistically coherent. The yeast assay itself does not resolve filament architecture or stoichiometry.

Summary: Desmin is associated with the contractile muscle-fiber structure.

Reason: The mouse-derived location agrees with the muscle cytoskeletal network. It is a broad anatomical component assignment and does not imply that desmin is the force-generating motor.

Summary: Desmin directly builds and organizes intermediate filaments.

Reason: The target descends from the PTN000580114 IBD node, and human DES experimental evidence is among legitimate descendant sources. Direct assembly and target muscle pathology corroborate the structural participation; this is more than a necessity-only phenotype.

Summary: Desmin variants perturb intermediate-filament organization.

Reason: The available cited abstract reports defective desmin filament formation or assembly. The full original was not recovered, so assay details are not invented. Independent purified human desmin assembly in PMID:28470624 supports retaining the existing IMP process assertion by curator deference.

Summary: Desmin variants perturb intermediate-filament organization.

Reason: The available cited abstract reports defective desmin filament formation or assembly. The full original was not recovered, so assay details are not invented. Independent purified human desmin assembly in PMID:28470624 supports retaining the existing IMP process assertion by curator deference.

Summary: Human Protein Atlas localizes DES to intermediate filaments.

Reason: The current official HPA record for ENSG00000175084/P17661 reports ICC/IF localization to intermediate filaments with supported reliability and GO:0045111. This independently corroborates the defining structural location; the label is an HPA location term, not a separate molecular activity.

Summary: The target-matching ARBA rule supports intermediate-filament cytoskeleton.

Reason: ARBA00028480 has a clause requiring IPR039008, IPR050405 and Primates. All are present in the current official human DES record, and the rule emits GO:0045111. Direct HPA and assembly evidence agree with this specific structural location.

Summary: Cell-tip anchoring is supported in cardiac muscle context.

Reason: The mouse P31001 donor has direct localization support in PMID:35959657. Super-resolution imaging compares desmin clusters with N-cadherin-marked cardiomyocyte ends in controls and PKP2-deficient cells. This supports a bounded terminal anchoring location; it does not imply universal cell-tip localization in all DES-expressing cell types.

Summary: Cell-tip anchoring is supported in cardiac muscle context.

Reason: The mouse P31001 donor has direct localization support in PMID:35959657. Super-resolution imaging compares desmin clusters with N-cadherin-marked cardiomyocyte ends in controls and PKP2-deficient cells. This supports a bounded terminal anchoring location; it does not imply universal cell-tip localization in all DES-expressing cell types.

Summary: Desmin contributes to skeletal-muscle development through the structural network.

Reason: The PTN002901991 IBD is supported by zebrafish descendant evidence and human DES is within the inherited clade. No target-specific loss is established. Retain the developmental process as a conserved contextual role rather than a second molecular-function core; the number of donor records is not a weakness.

Summary: Desmin is reported in an extracellular-vesicle proteomic preparation.

Reason: The full original describes differential-ultracentrifugation enrichment of pooled human expressed-prostatic-secretions urine and duplicate LC-MS with protein/peptide FDR below 1%. The existing HDA identification is retained, but this does not establish active DES secretion, intravesicular topology or a core extracellular function. The paper explicitly discusses co-isolated material and fluid-background limitations.

Summary: Desmin is associated with human cardiac myofibrillar preparations.

Reason: The full original uses permeabilized human cardiomyocytes and measures myofilament proteins including desmin during troponin-exchange experiments. This supports the location in a cardiac myofibrillar structure. The troponin mutagenesis and force-regulation results are not attributed to desmin.

Summary: Supramolecular fiber is a true but broad description of desmin filaments.

Reason: The current ARBA00026673 rule emits supramolecular fiber, but its complete target-matching clause was not reconstructed from captured domain/taxonomy inputs. Independent human desmin assembly establishes intermediate filaments, which have a checked is_a path to supramolecular fiber through polymeric cytoskeletal fiber. Use the supported specific structure to refine this existing broad assertion; unresolved rule conditions are not described as a false transfer.

Proposed replacements: intermediate filament