COL2A1

Summary: Retain the skeletal-development context of human collagen-II disease.

Reason: The complete original abstract describes a human COL2A1 Gly853Glu substitution with hypochondrogenesis, altered collagen assembly/secretion and abnormal cartilage matrix. This supports the preserved developmental assertion through matrix structure, without assigning a separate developmental signaling activity. Full experimental text was not recovered.

Summary: Retain ossification as a matrix-related developmental context.

Reason: The current exact mouse donor annotation is IEP-based and measures embryonic expression. Independent mouse Col2a1-null experiments directly show altered vertebral matrix and endochondral development, supporting a contextual structural contribution. Neither expression alone nor this judgment makes collagen a mineralizing enzyme.

Summary: The precise chondrocyte-differentiation assertion remains unresolved.

Reason: The complete human achondrogenesis abstract reports replacement of type-II collagen by types I and III, but the original full experimental evidence for the differentiation step was not recovered. A production GO-CAM models COL2A1 structural activity in this process using the same reference; this considered curator assertion is retained, not treated as independent experimental confirmation. Mouse Col2a1-null cartilage can differentiate and hypertrophy despite disorganized matrix, which does not rule out a context-dependent matrix contribution. Preserve the source pending a direct assessment of the human differentiation evidence.

Summary: Specify the established collagen tensile-strength structural class.

Reason: The original human matrix-disruption evidence is consistent with a structural collagen constituent. Official GO:0030020 explicitly includes extracellular-matrix collagens. Human fibril incorporation and independent matrix ultrastructure support this more informative term; no isolated-chain tensile assay is attributed to the original source.

Proposed replacements: extracellular matrix structural constituent conferring tensile strength

Summary: Specify the established collagen tensile-strength structural class.

Reason: The InterPro mapping correctly identifies the collagen structural class. Official GO:0030020 explicitly includes extracellular-matrix collagens. Human fibril incorporation and independent matrix ultrastructure support this more informative term; no isolated-chain tensile assay is attributed to the original source.

Proposed replacements: extracellular matrix structural constituent conferring tensile strength

Summary: Retain the positive KEAP1 binary-screen interaction.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. HuRI establishes association of the tested products. KEAP1 is an adaptor of a ubiquitin-ligase complex, so this pair alone does not establish contact with an assembled ligase or a collagen ubiquitination mechanism. No native colocalization or pair-specific mammalian validation is inferred.

Summary: Retain the positive HIP1 interaction at the tested-product scope.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. The exact canonical HIP1 product is identified, with curated endocytic and actin-associated functions. The saved evidence establishes F-actin association, but does not resolve a cytoskeletal-reorganization regulatory class for this partner; actin filament binding is not itself an actin-regulation assertion. Retain the supported pair without assigning a native cytoskeletal or endocytic role to COL2A1.

Summary: Retain the interaction with the recorded PRPF40A fragment.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. The partner is PRPF40A-2 residues 716–930, not the full protein. N-terminal WW-domain functions and whole-protein spliceosomal or cytoskeletal roles cannot identify the contacted class/domain within this C-terminal fragment. Preserve the positive interaction without a domain-specific refinement.

Summary: Specify binding to an actin-regulatory protein.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. The exact 782-residue gelsolin product is the displayed secreted/plasma precursor, distinct from cytoplasmic isoform 2. Its official function includes actin capping, nucleation and severing, satisfying the cytoskeletal-regulatory partner class of GO:0005519. This refinement describes the bound protein, not COL2A1-driven actin remodeling, mature plasma processing or a mapped contact with the active actin-binding site.

Proposed replacements: cytoskeletal regulatory protein binding

Summary: Retain the exact LAMP2B interaction.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. The partner sequence is LAMP2-2, the B product. LAMP2A-specific chaperone-mediated autophagy receptor activity is not transferred to this isoform or to collagen. The assay does not establish a native lysosomal role for the collagen fragment.

Summary: Specify binding to the receptor tyrosine kinase FGFR3.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. The exact canonical FGFR3 product has the independently curated receptor protein-tyrosine-kinase class. GO:0030971 describes that partner class; it does not assert receptor activation, mature glycosylation, a kinase-domain contact or physiological extracellular affinity of full-length collagen.

Proposed replacements: receptor tyrosine kinase binding

Summary: Specify the positive interaction with a caspase-family protein.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. The exact partner is the 293-residue CASP6 precursor. GO:0089720 explicitly denotes binding to a caspase-family protein and does not require an activated protease. The Y2H result does not establish proteolytic cleavage of collagen, an active-site interface or a COL2A1 apoptosis function.

Proposed replacements: caspase binding

Summary: Specify the ionotropic glutamate-receptor partner class.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. The exact canonical GRIN2C product is the GluN2C ionotropic NMDA-receptor subunit. This supports GO:0035255 at the positive binding-assay scope, without claiming an assembled native receptor, altered ion current, surface processing or a neuronal signaling role for COL2A1.

Proposed replacements: ionotropic glutamate receptor binding

Summary: Retain the positive COQ8A interaction without a protein-kinase claim.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. Official COQ8A identity and the primary biochemical literature distinguish an atypical kinase-like protein from established trans-protein-kinase activity. Substrate specificity remains unsettled, and gain-of-function autophosphorylation or yeast Coq8 ATPase assays are not wild-type human protein-kinase proof. Keep the positive association without forcing a disputed catalytic partner subclass or a collagen role in ubiquinone synthesis.

Summary: Retain the positive UBQLN1 interaction.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. The exact ubiquilin-1 product is a ubiquitin-associated proteostasis shuttle. The saved partner identity establishes a degradation shuttle but does not resolve a protein-folding chaperone class; the positive pair also does not establish a collagen proteasomal-delivery mechanism. Generic binding remains a supported secondary assay result.

Summary: Retain the positive interaction with the recorded short protein.

Reason: Three literal positive array, prey-pooling and validated Y2H records have human participants and yeast host. The tested COL2A1 source is P02458-1 residues 1151–1418, not a full mature collagen fibril; the 268-residue sequence also matches current product 3, but that does not re-identify the tested isoform. Q9Y649 is the exact 63-residue reference product in the source record; a more specific functional class is not established by its current identity record. Preserve the interaction without inventing a gene name, protein function or native mechanism.

Summary: Retain extracellular localization from the human venous study.

Reason: Actual Methods and Results describe normal and varicose human saphenous-vein ECM extraction and proteomics. The exact COL2A1 supplemental target row was not independently recovered, so source-level detection is not overstated. Independent direct human ECM target tables and cartilage incorporation support the location.

Summary: The orthology-derived extracellular location agrees with direct human evidence.

Reason: Current mouse donor identity and same-term records are consistent with secreted collagen-II in human cartilage and cell-derived ECM. This is a location assertion, not an inference of every activity performed by enzymes or receptors that act on collagen.

Summary: True: entailed by collagen type II trimer, collagen type XI trimer and extracellular matrix, kept in this review.

Reason: The claim is true: collagen type II trimer, collagen type XI trimer and extracellular matrix, kept in this review, entail this term. Other rows for this term are ACCEPTed. This row's reference does not establish it; that is recorded in source_defects, not grounds for REMOVE, which judges the claim.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-1454757 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Gelatin degradation by MMP1, 2, 3, 7, 8, 9, 12, 13, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-1474197 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Collagen type II degradation by MMP1,3,8,13,PRSS2, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-1474266 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Formation of collagen fibrils, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2002428 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Removal of fibrillar collagen N-propeptides, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2002440 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Removal of fibrillar collagen C-propeptides, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The secretion event resolves exact P02458 procollagen in the ER and extracellular collagen output. This original row records the extracellular output location; the separate ER-lumen source is preserved. Cargo membership does not establish an autonomous secretion activity.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-216043 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Collagen type II binds integrin alpha10beta1, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2213201 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Formation of collagen fibres, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2213205 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Type XII and XIV collagens associate with type I and type II fibrils, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2213208 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Type X collagen associates with collagen type II fibrils, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2213210 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Collagen IX is cross-linked to the surface of collagen type II fibrils, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2299620 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Collagen XI cross-links with collagen II, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2327733 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, FN1 binds Collagen types I-V, VII, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2327738 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, DDR1 binds collagen type I, II, III, IV, V, XI fibrils, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2327909 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, DCN binds collagen I, II, III, VI fibrils, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2396370 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, VTN binds collagens II, III and V, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2424252 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, COMP binds collagen, fibronectin, aggrecan and matrilins, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2465890 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, DDR2 binds collagen type I, II, III, V, X fibrils, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-2473594 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Collagen type II degradation by MMP15, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-375151 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, Interaction of NCAM1 with collagens, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-382054 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, PDGF binds to extracellular matrix proteins, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Retain the extracellular compartment of the curated collagen participant.

Reason: The current Reactome:R-HSA-5696356 hierarchy reaches exact P02458 collagen-II chains or processed fragments in the extracellular region through authenticated physical-entity paths. The event, OSCAR binds collagen and SP-D, supplies curated location context. Receptor activation, proteolysis and processing performed by other participants are not assigned to COL2A1 merely because it is bound or consumed.

Summary: Specify type-II collagen trimer membership.

Reason: The broad orthology-derived collagen-trimer assertion is correct. The alpha-1(II) chain is directly established as the repeated chain in type-II homotrimers; it also contributes the alpha-3 chain of type XI, retained separately. Refinement of this existing broad source does not add a redundant annotation.

Proposed replacements: collagen type II trimer

Summary: Specify type-II collagen trimer membership.

Reason: The broad ARBA fibrillar-trimer assertion is correct. The alpha-1(II) chain is directly established as the repeated chain in type-II homotrimers; it also contributes the alpha-3 chain of type XI, retained separately. Refinement of this existing broad source does not add a redundant annotation.

Proposed replacements: collagen type II trimer

Summary: Retain the ancestral type-II trimer membership.

Reason: PAINT represents a curator-placed ancestral assertion, not a vote among its WITH/FROM entries. Human COL2A1 appearing in that list is legitimate target experimental grounding. Direct human fibril-incorporation and mouse isoform-assembly evidence support the biological placement; the historical node was not independently reconstructed.

Summary: The source directly identifies mouse IIA chains in type-II collagen.

Reason: Actual Methods and Results analyze mice engineered to express the exon-2-containing alpha-1(IIA) form, using biochemical chain analysis, immunoprecipitation, mass spectrometry and matrix ultrastructure. This is positive ortholog evidence for type-II assembly, not a direct assay of every current human alternative product.

Summary: The human cartilage study includes genuine collagen-II incorporation.

Reason: Original indexed Methods and Results analyze human fetal cartilage and directly sequence alpha-1(II) telopeptides in type-II/type-IX cross-links. The title foregrounds type IX, but the experiments include COL2A1-derived material. This supports type-II matrix membership without asserting that the same study established an isolated COL2A1 homodimer.

Summary: Retain orthology-derived type-II collagen membership.

Reason: The exact mouse Col2a1 identity and current source records agree with experimentally observed type-II chain assembly. Human cartilage evidence independently supports the target membership, without generalizing the tested mouse splice construct to all human products.

Summary: Retain type-II chain association with the engineered construct scope explicit.

Reason: The actual source uses recombinant collagen-II peptide constructs with stabilizing GPP segments and a T4 foldon, with partial SDS-stable trimers and circular-dichroism evidence. These model constructs support collagen-chain association and are not a structure of full-length native human type-II collagen. Native type-II membership is independently established.

Summary: Retain the alpha-3(XI) contribution of the COL2A1 chain.

Reason: The official original abstract explicitly analyzes fetal bovine cartilage type-XI cross-linking; the full original paper remained inaccessible after bounded attempts. Independently read mouse IIA experiments directly establish COL2A1-derived chains in type XI. Retain the established membership with species/access limits, rather than reject a curator assertion from the bovine abstract alone.

Summary: Retain the established type-XI collagen membership.

Reason: The complete original review abstract supplies collagen-family context, rather than a new target assay. Direct ortholog biochemical evidence independently supports the COL2A1-derived alpha-3(XI) chain. No full review-paper read or all-isoform assembly experiment is claimed.

Summary: The exact basement-membrane localization remains unresolved.

Reason: The current mouse donor points to an ocular immunolocalization paper whose complete abstract reports broad collagen distribution. The specific basement-membrane target readout was not recovered. Established cartilage/vitreous ECM localization does not by itself prove this specialized subcompartment, so preserve the original source as undecided rather than deny possible ocular basement-membrane association.

Summary: Retain source-scoped intracellular detection without assigning an unresolved subcompartment.

Reason: The current mouse evidence includes intracellular retention of Dmm mutant procollagen in the complete PMID:14614991 abstract and intracellular IIA-reactive staining in the actual PMID:22248926 Results. The latter Discussion explicitly leaves the identity of the intracellular material unresolved between procollagen-related chains and cleaved N-propeptide; reuptake is proposed, not demonstrated. GO:0005737 includes subcellular structures, whereas GO:0005788 specifically requires the ER-enclosed volume. These source observations do not resolve that narrower location. Retain the broad orthology assertion as non-core; the separate ER-lumen annotations describe established biosynthetic intermediates and are not evidence that every intracellular signal in these experiments is ER-luminal.

Summary: Retain ER-luminal procollagen localization.

Reason: The electronic location matches the biosynthesis and assembly of a secreted collagen precursor. It is a genuine intermediate location rather than a second principal mature-protein compartment or an autonomous ER folding-catalyst function.

Summary: Retain the ER-luminal precursor and assembly-intermediate location.

Reason: The current Reactome:R-HSA-2022073 hierarchy resolves exact P02458 precursor chains in the ER lumen during Procollagen triple helix formation. This location is appropriate for procollagen assembly, distinct from the mature extracellular structural role. Processing, transport and folding-enzyme activities remain with the proteins performing them.

Summary: Retain the ER-luminal precursor and assembly-intermediate location.

Reason: The current Reactome:R-HSA-2089971 hierarchy resolves exact P02458 precursor chains in the ER lumen during Secretion of collagens. This location is appropriate for procollagen assembly, distinct from the mature extracellular structural role. Processing, transport and folding-enzyme activities remain with the proteins performing them.

Summary: Retain the ER-luminal precursor and assembly-intermediate location.

Reason: The current Reactome:R-HSA-8944263 hierarchy resolves exact P02458 precursor chains in the ER lumen during Association of procollagen type II. This location is appropriate for procollagen assembly, distinct from the mature extracellular structural role. Processing, transport and folding-enzyme activities remain with the proteins performing them.

Summary: Retain the ER-luminal precursor and assembly-intermediate location.

Reason: The current Reactome:R-HSA-8944266 hierarchy resolves exact P02458 precursor chains in the ER lumen during Association of procollagen type XI. This location is appropriate for procollagen assembly, distinct from the mature extracellular structural role. Processing, transport and folding-enzyme activities remain with the proteins performing them.

Summary: The precise central-nervous-system developmental role is unresolved.

Reason: The current mouse exact-term source is an IEP annotation grounded in Sox9/Col2a1 embryonic expression. The complete original abstract shows expression patterns, but full target-specific experiments establishing work performed in CNS development were not recovered. Preserve the positive expression and electronic source without inventing a signaling role or declaring the function impossible.

Summary: Retain the structural ocular contribution to visual function.

Reason: The actual human study includes COL2A1 mutations and Stickler ocular phenotypes as well as COL11A1/Marshall cases. It is not a wrong-paralog paper. Vitreoretinal matrix integrity can contribute to vision without collagen performing phototransduction; preserve this contextual visual phenotype.

Summary: Retain the COL2A1-associated visual phenotype.

Reason: The complete original abstract reports COL2A1 screening in families selected by vitreoretinal features. This is target-specific clinical evidence consistent with a structural ocular role. Full experimental text was not recovered, so no separate retinal biochemical mechanism is attributed to the source.

Summary: Retain the auditory phenotype as a secondary structural context.

Reason: The original human COL2A1/Stickler source has limited full-text access. The established connective-tissue disease context supports retaining the curator-recorded hearing phenotype without converting clinical hearing loss into a directly assayed sensory transduction activity of collagen.

Summary: Retain the collagen tensile-strength structural class.

Reason: Exact original human bone-marrow ECM workbook rows identify COL2A1. The HDA experiment establishes detection, not tensile modulus; the mechanical class is supported by the verified collagen term convention, human fibril incorporation and independent structural matrix evidence.

Summary: Retain the ancestral collagen structural activity.

Reason: The PAINT assertion reflects ancestral placement across matrix collagen homologs. It is supported by target structural evidence and the official collagen-inclusive definition, not by donor count or a reconstructed historical node. No target-specific loss of the collagen structural role is demonstrated.

Summary: Type-II matrix incorporation supports the structural inference.

Reason: The IC source uses type-II trimer membership as the inference premise. Actual human cartilage cross-linking verifies COL2A1 incorporation in the fibrillar matrix; this is compatible with the collagen structural class. The original source does not directly measure tensile force of purified alpha-1 chains.

Summary: Retain collagen structural classification with proteomics limits explicit.

Reason: Available original cardiac glycoproteomics text does not expose the exact COL2A1 target table. The RCA classification is independently supported by collagen-II matrix architecture and the current GO collagen convention. No tensile assay is inferred from detection or matrisome membership.

Summary: Retain collagen structural classification with proteomics limits explicit.

Reason: Available original Discussion explicitly includes type-II collagen among abundant human marrow-MSC matrix collagens, while the target workbook was not read. The RCA classification is independently supported by collagen-II matrix architecture and the current GO collagen convention. No tensile assay is inferred from detection or matrisome membership.

Summary: Retain collagen structural classification with proteomics limits explicit.

Reason: Authentic tissue-ECM Methods and discussion were read, but the target workbook remained inaccessible; attempted workbook files are HTML, not experimental tables. The RCA classification is independently supported by collagen-II matrix architecture and the current GO collagen convention. No tensile assay is inferred from detection or matrisome membership.

Summary: Retain the ancestral fibril-organization contribution.

Reason: The original PAINT source includes multiple collagen-family experimental descendants and the target itself; target self-inclusion is legitimate grounding. Human cartilage cross-links and mouse matrix ultrastructure support a performed structural contribution to fibril organization. No historical ancestral-node reconstruction or donor-count threshold is claimed.

Summary: Retain fibril organization while distinguishing the actual binding assays.

Reason: The original recombinant peptide study maps COMP/TSP-4 interactions with a collagen motif and tests model-trimer formation. It does not directly reconstruct complete native fibrillogenesis or solve a collagen-COMP complex. COL2A1 nevertheless performs the structural assembly underlying the preserved fibril-organization assertion, corroborated by human incorporation and mouse ultrastructure.

Summary: Retain combined electronic support for structural fibril organization.

Reason: ARBA and current mouse orthology are consistent with measured abnormal fibril architecture in Col2a1-null cartilage and directly observed collagen-II matrix incorporation. The collagen chain contributes structural material and association, rather than merely being processed by another protein.

Summary: Human collagen-II mutation supports altered fibril organization.

Reason: The complete original human Gly853Glu abstract links a COL2A1 mutation to abnormal cartilage matrix and collagen assembly/secretion. Independent directly read fibril ultrastructure supports the structural interpretation. No uninspected full-paper figure or separate fibril-remodeling enzyme is claimed.

Summary: Retain a performed matrix contribution to notochord remodeling.

Reason: The current ortholog source is expression-based, but independent Col2a1-null mouse experiments directly show persistent notochord, failed intervertebral-disc formation and altered vertebral matrix after initially normal early development. Collagen-II matrix structure contributes to that remodeling context. The proposed swelling-pressure mechanism is a Discussion explanation, not a measured pressure assay or a new human signaling activity.

Summary: Retain extracellular-matrix localization.

Reason: Original Table I identifies COL2A1 in the human HUES1-derived ECM column (normalized spectral count 1.2); feeder and mouse columns are not conflated. Independent human collagen-II matrix incorporation and direct target detection support this spatial assertion; detection does not establish every molecular activity assigned by matrisome classification.

Summary: Retain extracellular-matrix localization.

Reason: The available cardiac source provides ECM-enrichment/proteomics context, but the exact COL2A1 target table was not recovered. Independent human collagen-II matrix incorporation and direct target detection support this spatial assertion; detection does not establish every molecular activity assigned by matrisome classification.

Summary: Retain extracellular-matrix localization.

Reason: The available Discussion explicitly identifies type-II collagen in the human marrow-MSC matrix; exact supplemental target rows and complete Methods remain uninspected. Independent human collagen-II matrix incorporation and direct target detection support this spatial assertion; detection does not establish every molecular activity assigned by matrisome classification.

Summary: Retain extracellular-matrix localization.

Reason: Original S2/S3 workbooks contain exact COL2A1 rows in human bone-marrow ECM, with sample-category and aggregate identifications distinguished. Independent human collagen-II matrix incorporation and direct target detection support this spatial assertion; detection does not establish every molecular activity assigned by matrisome classification.

Summary: Retain extracellular-matrix localization.

Reason: Authentic human tissue-ECM extraction text was read; the target workbook was not recovered, and HTML download failures are not treated as tables. Independent human collagen-II matrix incorporation and direct target detection support this spatial assertion; detection does not establish every molecular activity assigned by matrisome classification.

Summary: Retain the ancestral extracellular-matrix location.

Reason: The broad collagen-family PAINT assertion is consistent with direct human COL2A1 matrix detection and cartilage incorporation. Target self-inclusion among the experimental descendants is valid, not circular. Current identity/annotation checks do not substitute for the uninspected historical phylogenetic node, and donor number is not used to judge support.

Summary: Human cartilage evidence supports matrix localization.

Reason: The complete original human achondrogenesis abstract reports replacement of type-II by types I and III in cartilage. Its full text was not recovered, but independent directly inspected human ECM tables and cartilage incorporation substantiate the correct location.

Summary: Retain the orthology-derived ECM location.

Reason: The exact mouse Col2a1 identity and current source annotations agree with independently observed human COL2A1 in cartilage and ECM-enriched preparations. No product-wide claim is made for the short alternative product merely from the gene-level location.

Summary: Retain the collagen-family review context for ECM localization.

Reason: The complete original review abstract discusses collagen supramolecular assemblies and matrix functions. Direct target evidence independently establishes COL2A1 matrix membership; no unavailable full review text or new primary assay is implied.

Summary: Retain MHC-II binding at the processed-peptide scope.

Reason: The authentic abstract reports the crystal structure of human HLA-DR4 with a peptide from human collagen II. This directly supports peptide-MHC binding, not docking of an intact native collagen fibril, collagen antigen-presentation machinery or MHC enzymatic activity. The original source participants remain intact.

Summary: Self-association enables the integrated collagen structural role.

Reason: The current mouse source and independent type-II homotrimer/chain-assembly evidence support identical-chain association. This is an enabling part of the single structural core, not a separate generic binding core or proof of a homodimer stoichiometry. Mouse IIA and engineered foldon constructs are not generalized to all human alternative products.

Summary: The exact homodimerization assertion is not established by the inspected source.

Reason: Original human cartilage Methods and Results in PMID:8660302 establish type-II/type-IX and type-IX/type-IX cross-links. They do not directly establish a COL2A1 homodimer. Type-II homotrimer assembly and broader identical-chain association are positive but do not settle this specific dimer activity. Preserve the original IPI/self-partner object as undecided, not as a denial of COL2A1 self-association. The verified GO:0070207 protein-homotrimerization term describes a distinct biological process; substituting it would not resolve the original molecular-function/source assertion. Established trimer assembly remains represented by the structural core and complex annotations.

Summary: Retain the experimentally supported proteoglycan interaction.

Reason: Actual human osteoarthritic-cartilage proximity assays support collagen-II/lubricin association. Direct solid-phase experiments use purified bovine collagen II with human recombinant lubricin and BSA controls; these are not a purified human COL2A1 affinity experiment. The source supports the binding capability without a universal cartilage-lubrication mechanism or all-product claim.

Summary: Retain platelet-derived growth factor binding as a secondary capability.

Reason: The complete original abstract reports radiolabeled PDGF isoforms binding collagens I–VI and constituent chains, with bioavailability consequences. Full reagent species and processing details were not recovered. Preserve the direct binding assertion without making collagen a PDGF receptor or assigning an unreported human construct.

Summary: Retain broad organ-development context as secondary to matrix structure.

Reason: The electronic developmental classification is consistent with human skeletal/ocular disease and directly observed developmental matrix defects in mice. It adds no independent molecular core and does not establish a generic organ-patterning signal.

Summary: Retain inherited skeletal morphogenesis as a structural context.

Reason: The PAINT placement is concordant with human COL2A1 skeletal disease and measured mouse vertebral architecture. The target appearing in WITH/FROM is expected experimental grounding. Retain the developmental context without claiming that the historical node was reconstructed or that collagen is a transcriptional patterning regulator.

Summary: Retain the cartilage-development phenotype of human COL2A1 disease.

Reason: The complete original abstract identifies COL2A1 mutations in three people with spondyloepimetaphyseal dysplasia, Strudwick type. This target-specific clinical context is consistent with collagen-II matrix structure, but is not an independent chondrocyte-differentiation or cartilage signaling assay.

Summary: The precise limb-bud-formation role remains unresolved.

Reason: The current mouse exact-term donor is IEP-based and points to a Znf219/Sox9 expression and transcriptional-regulation study. The complete abstract does not establish work performed by collagen II in limb-bud formation, and full target-specific experimental text was not recovered. Preserve the positive expression/source record without converting it into either a proven mechanism or a negative biological claim.

Summary: Clinical joint disease does not yet resolve the exact embryonic morphogenesis assertion.

Reason: The complete original abstract concerns COL2A1 mutation detection and Stickler clinical features. Full experimental text establishing the specified embryonic skeletal-joint morphogenesis was not recovered. Structural matrix participation is plausible and the curator assertion is preserved, but the precise developmental process remains undecided at this source scope.

Summary: The specific otic-vesicle developmental contribution remains unresolved.

Reason: The current mouse source is an embryonic-expression annotation. The complete original Sox9/Col2a1 abstract does not establish a performed collagen-II step in otic-vesicle development, and the relevant full experiments were not recovered. Hearing phenotypes do not automatically resolve this distinct embryonic process; preserve the original inference without a mechanistic invention or removal.