COG6

Summary: COG6 associates with Golgi membranes.

Reason: The UniProt location mapping agrees with endogenous COG6 staining and its interaction-dependent association with Golgi membranes. COG6 is a peripheral component, not a transmembrane SNARE.

Summary: COG6 associates with Golgi membranes.

Reason: The source describes COG6 in the Golgi-associated COG complex. Independent endogenous staining and SNARE-dependent membrane association support this compartment.

Summary: COG6 associates with Golgi membranes.

Reason: The Reactome event places the COG complex at the Golgi membrane; its physical-entity chain explicitly contains COG6/Q9Y2V7. This is a location assertion within the assembled machinery, not a claim that COG6 alone catalyzes the named event.

Summary: COG6 associates with Golgi membranes.

Reason: The Reactome event places the COG complex at the Golgi membrane; its physical-entity chain explicitly contains COG6/Q9Y2V7. This is a location assertion within the assembled machinery, not a claim that COG6 alone catalyzes the named event.

Summary: COG6 associates with Golgi membranes.

Reason: The Reactome event places the COG complex at the Golgi membrane; its physical-entity chain explicitly contains COG6/Q9Y2V7. This is a location assertion within the assembled machinery, not a claim that COG6 alone catalyzes the named event.

Summary: COG6 associates with Golgi membranes.

Reason: The Reactome event places the COG complex at the Golgi membrane; its physical-entity chain explicitly contains COG6/Q9Y2V7. This is a location assertion within the assembled machinery, not a claim that COG6 alone catalyzes the named event.

Summary: COG6 associates with Golgi membranes.

Reason: The Reactome event places the COG complex at the Golgi membrane; its physical-entity chain explicitly contains COG6/Q9Y2V7. This is a location assertion within the assembled machinery, not a claim that COG6 alone catalyzes the named event.

Summary: COG6 associates with Golgi membranes.

Reason: The Reactome event places the COG complex at the Golgi membrane; its physical-entity chain explicitly contains COG6/Q9Y2V7. This is a location assertion within the assembled machinery, not a claim that COG6 alone catalyzes the named event.

Summary: COG6 associates with Golgi membranes.

Reason: The Reactome event places the COG complex at the Golgi membrane; its physical-entity chain explicitly contains COG6/Q9Y2V7. This is a location assertion within the assembled machinery, not a claim that COG6 alone catalyzes the named event.

Summary: Conserved COG6 participation in retrograde Golgi traffic.

Reason: The inherited intra-Golgi recycling process is retained on the integrated COG6/COG trafficking evidence, including the target experimental premises, Golgi SNARE recognition and curated intra-Golgi complex events. PMID:21807881 directly tests endosome-to-TGN cargo retrieval; its BFA assay instead concerns Golgi-to-ER redistribution. Neither readout is relabeled as intra-Golgi flux. The historical PAINT node placement remains uninspected, distinct from the positive current target evidence.

Summary: COG6 deficiency compromises retrograde trafficking.

Reason: The original patient study reports reduced COG6, altered COG-complex fractionation and rescue by wild-type COG6. Its BFA/GM130 assay measures Golgi-to-ER redistribution, not an isolated intra-Golgi COPI step. Retain the curator’s intra-Golgi assertion in the integrated COG6/SNARE context while keeping that actual assay boundary explicit.

Summary: COG6 supports the COG-dependent recycling machinery.

Reason: The human knockout source directly tests COG6-dependent complex stability and Golgi architecture. Its toxin assays use representative COG knockouts, so an uninspected SubAB panel is not assigned to COG6. Retain the curator's intra-Golgi recycling assertion in the integrated COG machinery context. Independent PMID:21807881 establishes endosome-to-TGN retrieval, while PMID:23057818 supports COG6 recognition of Golgi SNAREs; these are complementary mechanisms, not an intra-Golgi cargo-flux measurement.

Summary: Retain the positive BCL10 interaction.

Reason: The exact COG6–O95999 pair is positive in BFG-Y2H (2 matched records). BCL10 is an immune-signaling adaptor; this pair does not establish that COG6 bridges its signaling complex.

Summary: Refine the positive ZBTB16 association by partner class.

Reason: The exact COG6–Q05516 pair is positive in BFG-Y2H (2 matched records). ZBTB16 is a DNA-binding transcriptional repressor; its verified full-length mapped product supports this partner-class refinement.

Proposed replacements: DNA-binding transcription factor binding

Summary: Retain the positive ARHGEF5 interaction.

Reason: The exact COG6–Q12774 pair is positive in BFG-Y2H (2 matched records). The ARHGEF5 record maps a 1079–1597 region of a guanine-nucleotide exchange factor. Binding a regulator does not by itself establish enzyme binding or GTPase regulation by COG6.

Summary: Retain the positive FAM90A1 interaction.

Reason: The exact COG6–Q86YD7 pair is positive in BFG-Y2H (2 matched records). FAM90A1 has no securely established functional partner class that would make this pair more informative.

Summary: Retain the positive CCHCR1 interaction.

Reason: The exact COG6–Q8TD31-3 pair is positive in BFG-Y2H (2 matched records). The mapped CCHCR1 isoform 3 interaction region is 90–835. Its coiled-coil and cell-localization information do not identify a specific binding activity here.

Summary: Refine the positive NTAQ1 association by partner class.

Reason: The exact COG6–Q96HA8 pair is positive in BFG-Y2H (2 matched records). NTAQ1 is a catalytic N-terminal glutamine amidase. The mapped product supports enzyme binding; an interaction does not establish that COG6 regulates deamidation.

Proposed replacements: enzyme binding

Summary: Retain the positive SCNM1 interaction.

Reason: The exact COG6–Q9BWG6 pair is positive in BFG-Y2H (2 matched records). SCNM1 is a minor-spliceosome component; the pair does not establish binding to an assembled spliceosome or RNA by COG6.

Summary: Refine the positive RAB6B association by partner class.

Reason: The exact COG6–Q9NRW1 pair is positive in BFG-Y2H (2 matched records). RAB6B is a small monomeric Rab GTPase. This refines the bound partner class without assigning a nucleotide state or GTPase-regulatory activity to COG6. The added ortholog corroboration is assessed separately in PMID:18256213; it does not change the original human screen scope.

Proposed replacements: small GTPase binding

Summary: Retain the positive WASHC3 interaction.

Reason: The exact COG6–Q9Y3C0 pair is positive in BioPlex 2.0 AP-MS (1 matched records). WASHC3 is a WASH regulatory-complex component. The positive AP-MS association does not resolve a direct interface or establish that the assayed protein itself is a protein component of the cytoskeleton.

Summary: Retain the positive ZFC3H1 interaction.

Reason: The exact COG6–G3V1X1 pair is positive in HuRI binary Y2H (3 matched records). The mapped ZFC3H1 accession is a 358-residue product. Preserve that exact source product rather than transferring a complete nuclear RNA-surveillance mechanism.

Summary: Retain the positive DTNB interaction.

Reason: The exact COG6–O60941-5 pair is positive in HuRI binary Y2H (3 matched records). The exact 560-residue DTNB isoform 5 is authenticated. Retained scaffold regions and dystrophin-complex association do not by themselves settle the assayed product’s cytoskeletal-component class.

Summary: Refine the positive SNAP29 association by partner class.

Reason: The exact COG6–O95721 pair is positive in HuRI binary Y2H (3 matched records). SNAP29 is a SNARE, and independent COG6 experiments show directed Y2H and HeLa co-immunoprecipitation with SNAP29. The mapped COG6 SNARE-binding segment also binds purified GS27 and STX6; purified binary SNAP29 binding is not claimed.

Proposed replacements: SNARE binding

Summary: Refine the positive SNAI1 association by partner class.

Reason: The exact COG6–O95863 pair is positive in HuRI binary Y2H (3 matched records). SNAI1 is a sequence-specific DNA-binding transcriptional repressor. Its exact mapped protein supports the partner-class term; COG6 transcriptional regulation is not inferred.

Proposed replacements: DNA-binding transcription factor binding

Summary: Refine the positive DRC4 association by partner class.

Reason: The exact COG6–O95995 pair is positive in HuRI binary Y2H (3 matched records). DRC4/GAS8 is a structural component of the axonemal nexin-dynein regulatory complex. Its component identity, beyond ciliary proximity, supports cytoskeletal protein binding.

Proposed replacements: cytoskeletal protein binding

Summary: Retain the positive BCL10 interaction.

Reason: The exact COG6–O95999 pair is positive in HuRI binary Y2H (3 matched records). BCL10 is an immune-signaling adaptor; the positive binary interaction remains useful without assigning its signaling function to COG6.

Summary: Refine the positive OAS1 association by partner class.

Reason: The exact COG6–P00973-2 pair is positive in HuRI binary Y2H (3 matched records). The mapped OAS1 isoform 2 is the 364-residue oligoadenylate synthetase product, retaining the enzyme region. This is enzyme binding, without an inferred COG6 antiviral catalytic function.

Proposed replacements: enzyme binding

Summary: Retain the positive SNRPB interaction.

Reason: The exact COG6–P14678-2 pair is positive in HuRI binary Y2H (3 matched records). The exact SNRPB isoform 2 is a spliceosomal protein. This pair does not demonstrate binding to a complete ribonucleoprotein complex.

Summary: Retain the positive PSMB1 interaction.

Reason: The exact COG6–P20618 pair is positive in HuRI binary Y2H (3 matched records). PSMB1 is explicitly a noncatalytic proteasome subunit. Binding one component does not establish binding to an assembled proteasome or a protease.

Summary: Retain the positive PSMA1 interaction.

Reason: The exact COG6–P25786 pair is positive in HuRI binary Y2H (3 matched records). PSMA1 is a noncatalytic proteasome structural subunit. The pair is retained without an enzyme-binding refinement based only on complex membership.

Summary: Refine the positive QARS1 association by partner class.

Reason: The exact COG6–P47897 pair is positive in HuRI binary Y2H (3 matched records). QARS1 is a catalytic glutamine–tRNA ligase. Binding to that exact protein supports enzyme binding without assigning aminoacylation to COG6.

Proposed replacements: enzyme binding

Summary: Refine the positive GEM association by partner class.

Reason: The exact COG6–P55040 pair is positive in HuRI binary Y2H (3 matched records). GEM belongs to the small monomeric GTPase family, and current GOA includes an experimental GTPase-activity record. UniProt also notes undetectable intrinsic activity in an assay context; no nucleotide-state preference or stimulation of hydrolysis is inferred from this pair.

Proposed replacements: small GTPase binding

Summary: Retain the positive MFAP1 interaction.

Reason: The exact COG6–P55081 pair is positive in HuRI binary Y2H (3 matched records). MFAP1 is a spliceosomal component; no direct splicing or assembled-spliceosome binding activity is established for COG6.

Summary: Refine the positive LMO4 association by partner class.

Reason: The exact COG6–P61968 pair is positive in HuRI binary Y2H (3 matched records). LMO4 is a transcriptional cofactor that acts through protein complexes. Its partner class supports coregulator binding without a COG6 transcriptional role.

Proposed replacements: transcription coregulator binding

Summary: Retain the positive TMSB4X interaction.

Reason: The exact COG6–P62328 pair is positive in HuRI binary Y2H (3 matched records). TMSB4X sequesters actin monomers and regulates actin assembly. That regulatory role is not by itself proof of the protein-component membership required by GO:0008092.

Summary: Refine the positive ZBTB16 association by partner class.

Reason: The exact COG6–Q05516 pair is positive in HuRI binary Y2H (3 matched records). ZBTB16 is a DNA-binding transcriptional repressor, supporting the specific class of this independently retained HuRI pair.

Proposed replacements: DNA-binding transcription factor binding

Summary: Refine the positive RUNX1T1 association by partner class.

Reason: The exact COG6–Q06455-2 pair is positive in HuRI binary Y2H (3 matched records). RUNX1T1 is a transcriptional corepressor. The exact isoform-2 mapping and recorded 11–577 interaction region retain its corepressor architecture; no isoform-selective repression is inferred.

Proposed replacements: transcription corepressor binding

Summary: Retain the positive ARHGEF5 interaction.

Reason: The exact COG6–Q12774 pair is positive in HuRI binary Y2H (3 matched records). The recorded ARHGEF5 1079–1597 region supports a positive regulator interaction; neither GEF binding as a distinct available term nor enzyme binding is established here.

Summary: Retain the positive RTP5 interaction.

Reason: The exact COG6–Q14D33 pair is positive in HuRI binary Y2H (3 matched records). The RTP5 label alone does not demonstrate that the bound protein transports a substance across a membrane.

Summary: Retain the positive CSRP2 interaction.

Reason: The exact COG6–Q16527 pair is positive in HuRI binary Y2H (3 matched records). CSRP2 has LIM domains and regulatory associations, but this source does not identify a mapped domain-specific interface or a securely resolved binding subclass.

Summary: Refine the positive TSHZ3 association by partner class.

Reason: The exact COG6–Q63HK5 pair is positive in HuRI binary Y2H (3 matched records). TSHZ3 is a homeobox-containing transcriptional repressor. The recorded 184–1081 region retains its annotated DNA-binding homeobox, supporting the partner class without a COG6 transcription mechanism.

Proposed replacements: DNA-binding transcription factor binding

Summary: Retain the positive KANK2 interaction.

Reason: The exact COG6–Q63ZY3 pair is positive in HuRI binary Y2H (3 matched records). KANK2 regulates protein localization and cytoskeletal behavior. The interaction alone does not settle its membership as a protein component of the cytoskeleton.

Summary: Retain the positive UFSP1 interaction.

Reason: The exact COG6–Q6NVU6 pair is positive in HuRI binary Y2H (3 matched records). The current UFSP1 record and ROGID map to a 218-residue catalytic product, but a later initiation-site correction leaves the historical HuRI clone start unresolved. Preserve the positive pair without asserting either a tested 142-residue inactive fragment or a verified protease-binding refinement.

Summary: Retain the positive PPP1R18 interaction.

Reason: The exact COG6–Q6NYC8 pair is positive in HuRI binary Y2H (3 matched records). PPP1R18 targets phosphatase activity to the actin cytoskeleton. Its regulatory association does not establish cytoskeletal protein-component membership or a catalytic enzyme partner.

Summary: Refine the positive SAXO4 association by partner class.

Reason: The exact COG6–Q7Z5V6-2 pair is positive in HuRI binary Y2H (3 matched records). SAXO4 is a microtubule inner-protein component of flagellar doublets. The mapped 405-residue isoform 2 differs by a 20-residue internal deletion and preserves that protein architecture. This is a partner-class inference from the official ortholog-grounded component annotation, not a COG6 ciliary function. The official human component assignment traces to mouse Q148A4, whose experimentally supported subunit annotation is explicit; this is not a direct assay of human Q7Z5V6-2 component membership. The exact tested product and positive interaction remain independently mapped.

Proposed replacements: cytoskeletal protein binding

Summary: Retain the positive TTC7B interaction.

Reason: The exact COG6–Q86TV6 pair is positive in HuRI binary Y2H (3 matched records). TTC7B recruits a lipid-kinase complex; this individual interaction does not establish that COG6 binds the assembled enzyme complex or performs its recruitment.

Summary: Retain the positive IGFN1 interaction.

Reason: The exact COG6–Q86VF2-5 pair is positive in HuRI binary Y2H (3 matched records). The recorded IGFN1 isoform-5 region is 2841–3708. Gene-level Z-disc association is insufficient to settle cytoskeletal component membership of that exact assayed region.

Summary: Retain the positive FAM90A1 interaction.

Reason: The exact COG6–Q86YD7 pair is positive in HuRI binary Y2H (3 matched records). The exact FAM90A1 pair is positive, but its functional partner class remains unresolved.

Summary: Retain the positive MGC50722 interaction.

Reason: The exact COG6–Q8IVT4 pair is positive in HuRI binary Y2H (3 matched records). The mapped MGC50722 protein has no sufficiently characterized partner class for a specific refinement.

Summary: Retain the positive CCDC116 interaction.

Reason: The exact COG6–Q8IYX3 pair is positive in HuRI binary Y2H (3 matched records). CCDC116’s coiled-coil designation does not identify a functional binding class or a mapped interface.

Summary: Retain the positive CNNM3 interaction.

Reason: The exact COG6–Q8NE01 pair is positive in HuRI binary Y2H (3 matched records). CNNM3 has transporter-family annotations, but its own membrane-crossing activity versus regulatory role remains insufficiently resolved here for transmembrane-transporter binding.

Summary: Retain the positive DOCK8 interaction.

Reason: The exact COG6–Q8NF50-4 pair is positive in HuRI binary Y2H (3 matched records). DOCK8 isoform 4 is a guanine-nucleotide exchange regulator. The positive pair does not make COG6 a GEF or identify an enzyme-binding subclass.

Summary: Retain the positive ZNF417 interaction.

Reason: The exact COG6–Q8TAU3 pair is positive in HuRI binary Y2H (3 matched records). ZNF417 has predicted transcription-regulatory features; a specific sequence-recognition role is not established by this interaction source.

Summary: Retain the positive RAB3IL1 interaction.

Reason: The exact COG6–Q8TBN0 pair is positive in HuRI binary Y2H (3 matched records). RAB3IL1 is a guanine-nucleotide exchange regulator. Retain its positive association without assigning its regulatory activity to COG6.

Summary: Retain the positive CCHCR1 interaction.

Reason: The exact COG6–Q8TD31-3 pair is positive in HuRI binary Y2H (3 matched records). The mapped CCHCR1 isoform-3 region is 90–835; coiled-coil structure does not establish domain-specific binding or cytoskeletal protein-component membership.

Summary: Retain the positive CUL5 interaction.

Reason: The exact COG6–Q93034 pair is positive in HuRI binary Y2H (3 matched records). CUL5 is a noncatalytic cullin scaffold. Its individual interaction is not equivalent to binding a complete ubiquitin-ligase enzyme.

Summary: Refine the positive SMARCD1 association by partner class.

Reason: The exact COG6–Q96GM5 pair is positive in HuRI binary Y2H (3 matched records). SMARCD1 is a noncatalytic transcriptional coregulator of SWI/SNF complexes that connects transcription factors to chromatin-remodeling machinery. The partner-class refinement does not assign ATPase activity or transcriptional regulation to COG6.

Proposed replacements: transcription coregulator binding

Summary: Retain the positive POM121 interaction.

Reason: The exact COG6–Q96HA1-2 pair is positive in HuRI binary Y2H (3 matched records). The exact POM121 isoform 2 is a nuclear-pore component. This does not itself establish binding to an assembled transmembrane transporter.

Summary: Refine the positive NTAQ1 association by partner class.

Reason: The exact COG6–Q96HA8 pair is positive in HuRI binary Y2H (3 matched records). NTAQ1 is a catalytic N-terminal glutamine amidase; the exact HuRI pair supports this enzyme-binding refinement independently of the earlier BFG record.

Proposed replacements: enzyme binding

Summary: Refine the positive CCDC120 association by partner class.

Reason: The exact COG6–Q96HB5 pair is positive in HuRI binary Y2H (3 matched records). CCDC120 is a centriolar subdistal-appendage component, with direct component-class evidence in current GOA. This supports the bound cytoskeletal-protein class, not a COG6 role in centriole assembly.

Proposed replacements: cytoskeletal protein binding

Summary: Refine the positive TSNAX association by partner class.

Reason: The exact COG6–Q99598 pair is positive in HuRI binary Y2H (3 matched records). TSNAX is the catalytic endonuclease partner of translin. Binding the mapped TSNAX protein supports enzyme binding without assigning RNA cleavage to COG6.

Proposed replacements: enzyme binding

Summary: Retain the positive FAM110A interaction.

Reason: The exact COG6–Q9BQ89 pair is positive in HuRI binary Y2H (3 matched records). FAM110A localization and coiled-coil features do not establish a specific functional partner class.

Summary: Retain the positive SYT17 interaction.

Reason: The exact COG6–Q9BSW7 pair is positive in HuRI binary Y2H (3 matched records). SYT17 is a synaptotagmin-family membrane-associated protein; this pair does not establish receptor or SNARE binding.

Summary: Refine the positive MRI1 association by partner class.

Reason: The exact COG6–Q9BV20 pair is positive in HuRI binary Y2H (3 matched records). MRI1 is a catalytic methylthioribose-1-phosphate isomerase. The positive pair supports enzyme binding, without a metabolic function being transferred to COG6.

Proposed replacements: enzyme binding

Summary: Retain the positive NUP58 interaction.

Reason: The exact COG6–Q9BVL2 pair is positive in HuRI binary Y2H (3 matched records). NUP58 is a nuclear-pore component. The individual pair is distinct from binding an assembled transport complex.

Summary: Retain the positive FNDC11 interaction.

Reason: The exact COG6–Q9BVV2 pair is positive in HuRI binary Y2H (3 matched records). FNDC11’s fibronectin-domain annotation alone does not demonstrate extracellular-matrix structural activity or a mapped domain interface.

Summary: Retain the positive SCNM1 interaction.

Reason: The exact COG6–Q9BWG6 pair is positive in HuRI binary Y2H (3 matched records). SCNM1 is a minor-spliceosome component; retain the positive pair without a splicing mechanism for COG6.

Summary: Retain the positive TSGA10 interaction.

Reason: The exact COG6–Q9BZW7 pair is positive in HuRI binary Y2H (3 matched records). TSGA10 has processed sperm-tail products with distinct localization. The mapped full-length protein cannot automatically inherit the cytoskeletal-component assignment of a processed fragment.

Summary: Retain the positive PLEKHA2 interaction.

Reason: The exact COG6–Q9HB19 pair is positive in HuRI binary Y2H (3 matched records). PLEKHA2 binds phosphoinositides; the interaction does not transfer lipid binding to COG6 or establish a receptor class.

Summary: Refine the positive NEK6 association by partner class.

Reason: The exact COG6–Q9HC98-4 pair is positive in HuRI binary Y2H (3 matched records). The mapped NEK6 isoform 4 is 338 residues; the recorded 26–338 binding region equals the canonical 313-residue sequence and retains the complete kinase domain. Kinase binding is therefore supported without an activation or substrate claim.

Proposed replacements: kinase binding

Summary: Refine the positive RAB6B association by partner class.

Reason: The exact COG6–Q9NRW1 pair is positive in HuRI binary Y2H (3 matched records). The exact RAB6B pair supports small GTPase binding. Nucleotide-state selectivity is not determined by this screen. The added ortholog corroboration is assessed separately in PMID:18256213; it does not change the original human screen scope.

Proposed replacements: small GTPase binding

Summary: Retain the positive TBC1D22B interaction.

Reason: The exact COG6–Q9NU19 pair is positive in HuRI binary Y2H (3 matched records). TBC1D22B has a putative Rab-regulatory role. Preserve the interaction without inferring an exact catalytic or GTPase-binding class.

Summary: Retain the positive SPG21 interaction.

Reason: The exact COG6–Q9NZD8 pair is positive in HuRI binary Y2H (3 matched records). SPG21/maspardin is an alpha/beta-hydrolase-fold protein without an established catalytic activity in this source; fold membership alone does not justify enzyme binding.

Summary: Refine the positive RSPH14 association by partner class.

Reason: The exact COG6–Q9UHP6 pair is positive in HuRI binary Y2H (3 matched records). RSPH14 is an axonemal radial-spoke stalk component in the official ortholog-grounded annotation. Its mapped full-length product supports the cytoskeletal partner class without a COG6 axonemal role. The official human assignment traces to mouse Q9D3W1 with experimentally supported radial-spoke complex membership. The class transfer is qualified ortholog evidence, not direct human radial-spoke assembly or a reason to count inference steps as disproof.

Proposed replacements: cytoskeletal protein binding

Summary: Retain the positive BAIAP2 interaction.

Reason: The exact COG6–Q9UQB8-6 pair is positive in HuRI binary Y2H (3 matched records). BAIAP2 isoform 6 is a membrane/cytoskeleton regulatory adaptor. The pair does not establish that this exact product is a protein component of the cytoskeleton.

Summary: Retain the positive WASHC3 interaction.

Reason: The exact COG6–Q9Y3C0 pair is positive in HuRI binary Y2H (3 matched records). WASHC3 belongs to an actin-regulatory complex. Retain the binary pair without equating actin regulation with demonstrated cytoskeletal protein-component membership.

Summary: Refine the positive SNAP29 association by partner class.

Reason: The exact COG6–O95721 pair is positive in BioPlex AP-MS (2 matched records). The exact BioPlex COG6–SNAP29 association is supported by two AP-MS records. Independent directed Y2H and HeLa co-immunoprecipitation establish COG6–SNAP29 association, supporting SNARE binding while preserving the co-complex scope of BioPlex.

Proposed replacements: SNARE binding

Summary: Refine the positive rep association by partner class.

Reason: The exact COG6–P0DTD1-PRO_0000449631 pair is positive in SARS-CoV-2–human binary Y2H (3 matched records). The source maps the processed viral nsp14 product (polyprotein residues 5926–6452), a catalytic RNA-processing enzyme. Retain enzyme binding at the virus–host assay scope; COG6 is not thereby a viral methyltransferase or exonuclease.

Proposed replacements: enzyme binding

Summary: Refine the positive rep association by partner class.

Reason: The exact COG6–P0DTD1-PRO_0000449633 pair is positive in SARS-CoV-2–human binary Y2H (3 matched records). The source maps the processed viral nsp16 product (polyprotein residues 6799–7096), an RNA 2′-O-methyltransferase. Retain enzyme binding at the virus–host assay scope, without transferring its chemistry or infection requirement to COG6.

Proposed replacements: enzyme binding

Summary: COG6 participates in intra-Golgi vesicular transport.

Reason: The InterPro COG6-family assignment is consistent with its role in the COG complex, Golgi SNARE recognition and curated intra-Golgi tethering/SNARE events. PMID:23057818 supports the machinery interaction. PMID:21807881 separately demonstrates endosome-to-TGN retrieval and does not directly assay this intra-Golgi route. Preserve the source assertion without assigning glycosyltransferase chemistry or autonomous fusion to COG6.

Summary: Conserved participation in Golgi organization.

Reason: Current Arabidopsis donor data include an experimental Golgi-organization annotation, and human COG6 loss alters Golgi cisternal structure. Direct COG6–SNARE recognition supplies a performed molecular role. This supports the target assertion without reconstructing the historical PAINT node.

Summary: COG6 helps maintain Golgi architecture.

Reason: The human HEK293T COG6 knockout has strongly dilated and fragmented Golgi cisternae by electron microscopy; independent COG6 depletion in HeLa cells also fragments the Golgi ribbon. Its demonstrated SNARE-recognition role supports participation beyond a nonspecific loss phenotype.

Summary: COG6 is a conserved component of the COG complex.

Reason: Current yeast and fly donor records include experimental COG-complex membership, and the human subunit is independently assembled with other COG proteins. These positive premises agree with the inherited complex assertion; historical IBD placement remains uninspected.

Summary: Human COG6 joins the reconstructed COG assembly.

Reason: The original reticulocyte-lysate reconstruction uses human COG constructs. COG6 incorporates cooperatively with COG5 plus COG7, and its omission reduces COG8 incorporation. This establishes complex participation without purified binary COG6–COG5/7 interfaces.

Summary: The COG6-family mapping identifies a COG subunit.

Reason: The family-based complex assignment is supported by human COG6 assembly and endogenous complex evidence. It does not imply that every short alternative product has been tested.

Summary: COG6 is part of the Golgi COG complex.

Reason: The complete-knockout study describes the octameric COG complex and directly examines COG6-dependent lobe-B stability. Original human assembly experiments independently establish the complex assignment.

Summary: COG6 acts at the trans-Golgi network membrane.

Reason: The Reactome event contains the TGN-localized COG complex and its exact COG6/Q9Y2V7 entity. COG6-specific STX6 association and endosome-to-TGN transport experiments support this contextual membrane assignment. COG6 is not itself a membrane-spanning SNARE.

Summary: COG6 acts at the trans-Golgi network membrane.

Reason: The Reactome event contains the TGN-localized COG complex and its exact COG6/Q9Y2V7 entity. COG6-specific STX6 association and endosome-to-TGN transport experiments support this contextual membrane assignment. COG6 is not itself a membrane-spanning SNARE.

Summary: COG6 performs SNARE-recognition work supporting endosome-to-TGN retrieval.

Reason: Human HeLa COG6 depletion delays STx-B and TGN38 delivery to the TGN. Combined with endogenous association, purified N-terminal COG6–STX6 binding and disruption of STX6-containing SNARE assemblies, this identifies COG6 as part of the transport machinery rather than the transported substrate. The mechanism is complex-mediated; depletion also affects other COG subunits, and no isolated fusion activity or equal activity of all isoforms is inferred. Exact-term comparator records for VPS53/VPS54 show that analogous SNARE-recognizing tether components can carry this process; COG3/COG4 lack it in the checked snapshot. Typed is_a/part_of closure finds no ancestor/descendant overlap with the original or replacement annotations or core terms. The current GO-CAM index has no target entry.