EMILIN1 combines a matrix-organizing role with integrin-mediated cell anchorage. The structural evidence includes interactions with elastin and fibulins, deposition rescue and organized extracellular assemblies. The adhesion evidence includes purified human alpha4beta1 binding, ligand-domain mutagenesis and receptor-blocking experiments. These are complementary activities; EMILIN1 is not the alpha4beta1 receptor heterodimer or the LOX enzyme that performs oxidative crosslinking. PMID:12456677(https://pubmed.ncbi.nlm.nih.gov/12456677/), PMID:14701737(https://pubmed.ncbi.nlm.nih.gov/14701737/), PMID:18463100(https://pubmed.ncbi.nlm.nih.gov/18463100/), PMID:28717224(https://pubmed.ncbi.nlm.nih.gov/28717224/).
The 99 imported contexts are preserved: 29 ACCEPT, 17 MODIFY, 48 KEEP_AS_NON_CORE, four UNDECIDED and one REMOVE. The 115 raw QuickGO records project to 103 normal TSV rows and then 99 source contexts; original identifiers, extension variation and multiplicity remain available in the frozen data. No NEW annotation or publication quotation is added. Supported generic interaction-screen pairs remain non-core with their assay limits, rather than being discarded or assigned a function from partner class alone.
The seven elasticity annotations are generalized to extracellular matrix structural constituent. This retains the positive mouse donor phenotype and matrix-mechanical evidence while distinguishing structural integrity from a resolved contribution to reversible recoil. The 2022 paper does include femur three-point bending and altered knockout stiffness. Proteomic detections corroborate matrix membership but are not mechanical assays. PMID:36351433(https://pubmed.ncbi.nlm.nih.gov/36351433/).
Human proteomic targets were checked against genuine tables or author-deposited outputs. The PXD000359 endothelial-matrix output contains an EMILIN1 group with positive intensity in all three experiments. The recovered OUP supplement for the human-vein study independently resolves its target. The tissue and cell-derived matrix studies retain protein-group, sample/pool, serum and fractionation limitations. The trabecular-meshwork vesicle location has specific immunoblot and density-gradient support, but remains a contextual non-core location. PMID:23979707(https://pubmed.ncbi.nlm.nih.gov/23979707/), PMID:27068509(https://pubmed.ncbi.nlm.nih.gov/27068509/), PMID:21362503(https://pubmed.ncbi.nlm.nih.gov/21362503/).
Two natural products remain distinct: Q9Y6C2-1 (1,016 residues) and Q9Y6C2-2 (347 residues). The long-product signal peptide, mature chain 22–1016, recombinant gC1q fragments, tags and disease constructs are not interchangeable product identities. In the 2008 domain study, paper residues 846–995/Glu933 correspond to current long-product 867–1016/Glu954; twelve tag/cloning residues are separate. Shared protein groups do not establish equivalent activity of both natural products. PMID:18463100(https://pubmed.ncbi.nlm.nih.gov/18463100/).
The four unresolved annotations are explicit. Coagulation has unresolved homolog-complex source provenance; the antitoxin paper's operative experiment is unavailable; the two macrophage-migration rows retain genuine infiltration observations but an unresolved process-specific interpretation. Access is not used as a substitute explanation for that scientific uncertainty. The two collagen-organization rows are generalized to unsigned regulation using direct matrix-interaction work and human/mouse collagen evidence. The mouse knockout study actually increases angiogenesis, supporting correction of the two positive-sign transfers to contextual negative regulation. PMID:17988845(https://pubmed.ncbi.nlm.nih.gov/17988845/), PMID:25056700(https://pubmed.ncbi.nlm.nih.gov/25056700/).
Three structured source queries are included without reporting them upstream: the CAFA receptor-complex membership mismatch and the GO Central/Ensembl signed angiogenesis transfers. All are NOT_REPORTED; the original source fields remain exact. The coagulation provenance and inherited/current PAINT ligand-node discrepancy remain questions rather than asserted database errors.
Normal publication-cache availability and actual reading are separate. A true full_text_unavailable flag records an abstract-only normal cache even where an original was recovered separately; each reference assessment and the source guide specifies the actual access. Genuine provider attempts failed and no synthetic provider report was created. The visible analysis replays frozen database and tabular evidence without redistributing original article bodies or figure images.
The collagen-biosynthesis assessments now include direct human fibroblast evidence from Figure 6A of PMID:36351433. The same paper distinguishes increased production from impaired collagen network assembly/crosslinking, supporting the unsigned organization replacement without assigning LOX chemistry to EMILIN1. Macrophage accumulation remains distinct from measured migration. The source guide now supplies direct original-paper links and exact experiment/figure locators; cache flags and the no-new-quotation policy remain unchanged. The YAML records concrete questions about collagen assembly, macrophage movement, antitoxin preparations and the two unresolved provenance issues.