This guide records what was inspected for the FN1 annotation augmentation. The source evidence explains the biological decisions; the context archive preserves the earlier review, current original source objects, exact projection and all inherited quotation associations. The older fetched UniProt record, GOA table and provider report are preserved unchanged.
The target is human FN1, HGNC:3778, UniProt P02751. The current QuickGO annotation service supplied 222 original contexts. The archive records every original object and assigned-by, partner and extension field. Forty original contexts have extensions. Alternative OR branches are represented as separate annotations, while each conjunction remains together, following GAF 2.2 column 16. This produces 223 current typed contexts, including 48 extension groups and 55 atoms.
The final 235 rows also retain eleven historical machine-source contexts not represented by the current response, including the older heart-development donor group, and one individually re-evaluated authored matrix-organization assertion. Absence from a newer response is not a retirement criterion. Only the two unsupported or redundant authored proposals are retired. The archive supplies the proposed-slot-to-final-index map; the earlier indices used below are zero-based and remain unambiguous there.
Predicate identifiers are CURIE strings, with typed target terms. Official OLS/QuickGO metadata resolved BFO:0000050, BFO:0000066, RO:0002233, RO:0002565 and RO:0002411 and all target labels. The exact BFO part-of IRI was used rather than an anomalous returned obo_id. Original source relation names and OR/conjunction objects are preserved alongside their schema representation.
The exact hashes and paths of all 148 normal dependencies are in normal_source_identity_manifest in the context archive. Existing normal sources were preserved; six previously absent publications were acquired through the repository fetcher. A further needed donor paper, PMID:36812915, was hydrated from its existing Git object. A normal cache's full_text_available flag is an acquisition fact, not a claim that every section, figure or supplement was read.
Selected Methods, Results and text captions were read where specified below. Paragraph or line locators identify the retained text extraction, not printed publication page numbers. No figure-pixel or supplement inspection is implied by reading a caption.
| Source | Inspected evidence and scope |
|---|---|
| PMID:12388756(https://pmc.ncbi.nlm.nih.gov/articles/PMC129965/) | Original Methods/Results/Discussion, especially Figure 6 text for human fibroblast and aortic smooth-muscle FN-matrix maintenance; Figures 7/9 distinguish scaffold organization from integrin ligation. Human plasma FN, rat-fragment interventions and separate thrombospondin/collagen reconstitution are distinguished. |
| PMID:33962943(../../../publications/PMID_33962943.md) | Selected original human α5β1/FN7–10 structural Results, Methods and captions. The RGD/synergy mechanism is not transferred to every integrin or natural FN1 product. |
| PMID:24658351(../../../publications/PMID_24658351.md) | Selected original Methods/Results and captions distinguish wild-type human FN10 from an engineered high-affinity antagonist. Natural products are not all assigned antagonist activity. |
| PMID:10075919(https://pubmed.ncbi.nlm.nih.gov/10075919/) | Authentic abstract and retained duplicate text only. Human FNIII12–14 heparin/integrin-binding structure informs distinct sites; no complete Methods or figure audit is claimed. |
| PMID:3024962(https://pubmed.ncbi.nlm.nih.gov/3024962/) | Authentic abstract: human recombinant fragments bind gelatin-Sepharose. This is not an unrestricted assay of all native collagen types. |
| PMID:20541508(https://pubmed.ncbi.nlm.nih.gov/20541508/) | Authentic abstract: six-module gelatin-binding domain, zinc-dependent assembly and collagen-peptide binding. The separate soluble fibrinogen-complex assertion remains unresolved. |
| PMID:29030641(https://pmc.ncbi.nlm.nih.gov/articles/PMC5640667/) | Selected original Methods/Results: human plasma FN binds recombinant human PRG4/lubricin fragments in solid-phase assays; cartilage proximity is separate supporting context. |
| PMID:19126672(https://pmc.ncbi.nlm.nih.gov/articles/PMC2724728/) | Selected original paragraphs 23–40 and 77–80. Human plasma FN on mouse NIH3T3 cells supports β1-dependent FAK/paxillin/spreading. PI3K/Akt signals also occur on poly-D-lysine and are not assigned exclusively to integrin engagement. |
| PMID:20123964(https://pmc.ncbi.nlm.nih.gov/articles/PMC2838065/) | Selected Results/Methods paragraphs 18,20,22,36,38,40,71,74,76,77,79,82,84. Added FN rescues migration/wound closure of LPP3-depleted human dermal microvascular endothelial cells. Separate branching rescue is not an FN-alone result. |
| PMID:26571399(https://www.jci.org/articles/view/83178) | Selected Results/Methods paragraphs 120,126,136: embryonic mouse enteric neural-crest explants migrate on FN with GDNF; excess collagen VI inhibits that response. Commercial FN coating species is unresolved in the inspected wording. Human disease observations are separate. |
| PMID:25834989(../../../publications/PMID_25834989.md) | Actual selected Methods/Results and captions distinguish human plasma FN controls from FBLN1C1 treatment. Cell counts, attachment, EDA-FN/perlecan deposition and secreted total TGF-β1 were measured. Deposition/ELISA do not isolate transcription or synthesis from turnover. Five inherited excerpts describing FBLN1C1 are archived rather than presented as FN-specific results. |
| PMID:36812915(https://pmc.ncbi.nlm.nih.gov/articles/PMC10079584/) | Selected paragraphs 34,38,39,41,45,46,48,52–54,68,81,83,95,99 and exact mouse donor GOA. Muscle secretion and liver-response evidence inform the two human IEAs. A human cell host does not establish the species of supplied protein. |
| PMID:10742111(https://pubmed.ncbi.nlm.nih.gov/10742111/) | Original abstract explicitly reports FN on the apical surface of mouse hair cells. Full Methods and figure pixels were not read. The human annotation retains its transferred evidence code and donor alternatives. |
| PMID:16061471(https://www.researchgate.net/publication/7686990_Fibulin-1_Acts_as_a_Cofactor_for_the_Matrix_Metalloprotease_ADAMTS-1) | The normal cache is abstract-only, recorded by the optional full_text_unavailable: true reference flag. Separately inspected author-uploaded original article Methods (indexed line 855), Results (1310–1342), Figure 6A caption and Discussion (1590–1596) explicitly include fibronectin in proteoglycan-digestion experiments and report approximately twofold enhancement of ADAMTS1 activity relative to BSA. Rat aggrecan and the original donor transfer are distinguished from a direct human FN1 experiment. PDF download and figure-pixel inspection were unavailable; the actual selected original text, rather than the abstract or title, supports this bounded capacity. |
| PMID:22952693(../../../publications/PMID_22952693.md) | Actual normal Methods/Results and supplemental-text description support human plasma FN degradation by cathepsins. Inspected binding experiments chiefly concern nidogen. The three exact protease-binding assertions remain unresolved; substrate status is not evidence against binding. |
| PMID:23154389(../../../publications/PMID_23154389.md) | Actual normal body: human FN/vitronectin coatings, Activin A/Wnt3a and integrin interventions support the hESC endoderm context, without FN-alone fate instruction. |
| PMID:34089617(https://pubmed.ncbi.nlm.nih.gov/34089617/) | Authentic abstract explicitly reports human/mouse receptor binding. The exact monocyte readout was not accessible; publisher challenge pages were not treated as primary text. Binding and process assertions have separate dispositions. |
| PMID:12069579(https://pubmed.ncbi.nlm.nih.gov/12069579/), PMID:567240(https://pubmed.ncbi.nlm.nih.gov/567240/), PMID:7989369(https://pubmed.ncbi.nlm.nih.gov/7989369/) | Authentic abstracts only. The first distinguishes binding to fibrin from immobilized fibrinogen under its tested conditions; the older paper reports positive fibrinogen/fibrin interactions. The inherited 7989369 source assays fibrin. Together these limit the authored soluble-fibrinogen claim without establishing universal nonbinding. |
| PMID:3997886(https://pubmed.ncbi.nlm.nih.gov/3997886/) | Authentic abstract describes FN incorporation into polymerized fibrin clots through cross-linking. That result does not independently resolve membership in the soluble fibrinogen precursor complex. |
Authentic abstracts also informed the remaining source-specific annotations, including PMID:11792823, PMID:16236823, PMID:12167537, PMID:12421310 and PMID:19429745. RCA/HDA matrix papers establish a human proteomics context, but individual target peptide/tissue supplementary tables were not replayed. Their accepted structural function rests on independent FN evidence while every original tissue extension remains intact.
All 52 binding-source abstracts were inspected. The complete paginated official IntAct target query yielded 632 records. Exact PMID/partner joins, including construct, host and assay fields, were used where available. A missing match was never treated as a negative biological result. Official UniProt records established partner classes; a class-specific binding term does not give FN the partner's catalytic activity.
| Source | Additional inspected material and boundary |
|---|---|
| PMID:18042364 | Normal lines 82–90, authentic abstract and exact solid-phase/cross-link records. Human plasma/cellular FN binds full TSG-6/CUB_C; the cache omits full original Methods/Results. The exact FN self-pair specifies fragments 1420–1813, not physiological stoichiometry. |
| PMID:18411296 | Normal selected lines 71–85 identify EF-Tu C-terminal binding regions. FN recognition is a moonlighting interaction of the bacterial GTPase. |
| PMID:19220903 | Results 98–126/Table 1 and 134–154. EnoA1 is individually supported; five proteins co-migrate in a second FN-reactive band, and EnoA2 identity alone is not a separate binding assay. Six assertions remain undecided. |
| PMID:19542224 | Normal line 67 and exact FN/THBS1 SPR-array record; no pair-specific kinetic constant is invented. |
| PMID:19738201 | Normal Results 77–85 and exact target record: chemically cross-linked FN adhesion complexes in K562-based assays. TGM2 association is not a purified binary interface. |
| PMID:22442151 | Methods 145/149 and Results 232–236 distinguish immobilized-FN/anti-His binding from catalytic assays in 111/115. Similar WT/mutant association and low activity when FN-bound do not imply catalytic activation. |
| PMID:24136289 | Exact yeast two-hybrid, pull-down and split-luciferase records identify a 599-nt FN prey and viral fragments. Official processed chain PRO_0000045603 of Q99IB8, residues 2443–3033, is the RNA polymerase. The original cache omits full Methods/Results. |
| PMID:24388360 | Selected Discussion 98,106–110 and abstract distinguish SERPINA5 association from downstream migration and the partner's inhibitor activity. |
| PMID:25034023(https://doi.org/10.1016/j.febslet.2014.06.061) | Authentic indexed publisher Methods 2.4–2.6, Results 3.2–3.3/Table 2/Figure 3 caption explicitly assay FBN1/FBN2 fragments against FN by solid-phase and SPR methods. Direct-page access failures are distinct from recovered indexed original text. |
| PMID:25290767(https://febs.onlinelibrary.wiley.com/doi/full/10.1111/febs.13092) | Selected original Results 307–323 and 352–367/Figure 3 describe human gelatin-binding domain, collagen-I Gly778–Gly798 peptide, anisotropy and displacement by FNE. |
| PMID:26848503(https://pmc.ncbi.nlm.nih.gov/articles/PMC4910512/) | Original BioC/XML Results paragraphs 9–12,20–24, Methods 43/55 and Table 1 paragraph 141 identify FN1 with nine peptides and enrichment 5.2/4.5. Human HT-1080 DSP cross-linking/tagged-collagen IP-MS supports association without FN-specific binary validation. |
| PMID:26627825 | Table 1/Results 182,188,194 and Figure 5 caption 197 support fibrinogen-mediated Efb–FN association, distinct from direct Efb/P-selectin assays. |
| PMID:30082873 | Results 172/Table 3/Discussion 188 support LOX-propeptide binding to anastellin. The tested propeptide is not the mature catalytic enzyme. |
| PMID:33961781 | Exact tagged-bait co-IP record identifies human HEK293T and Supplementary Table 1; a co-complex association is not recast as a purified interface. |
PMID:20080707 lacks the exact FN-specific assay in the accessible text; the attempted BioC response was an error body. PMID:10788510 does not resolve the specific FN–ClfA pair from its abstract. PMID:17849409 directly supports GlnA1–FN but leaves its separate DnaK/Ag85B pairs unresolved. These are source-specific uncertainties, not claims that the interactions are biologically impossible.
The specific replacements use official collagen binding, integrin binding, proteoglycan binding and lipoprotein-particle binding classes. The intact lipoprotein(a) assay is distinguished from an isolated apo(a)-kringle screen. Twelve microbial adhesion interactions remain non-core even though their partners are enzymes: catalytic classification does not characterize the adhesion operation measured in those sources. The original interaction evidence and construct limits remain. Supported interactions without an established informative subtype remain non-core under the campaign instructions.
Forty-seven selected Reactome content-service event paths were followed to their physical-entity leaves. A cytosolic parent set does not overwrite a leaf compartment: the actual FN1–ALK monomer/dimer leaves are plasma-membrane entities. All twelve fusion-derived contexts are marked over-annotated at the unqualified native-FN1 scope. Their source-defect records request a curator check of disease-fusion context in the GO export; they do not dispute the Reactome fusion model, guess a replacement identifier or assign ALK catalysis to native FN1. Soluble plasma/precursor/release contexts are distinguished from assembled matrix; only the thirteen source-supported matrix refinements are retained. Reactions consuming FN or acting on another participant do not give FN the enzyme's function.
The actual PTHR46708 family has 218 tree nodes and 218 aligned sequences. Human PTN002498994 contains the exact 2477-aa P02751 sequence. All six incoming IBA origins are PTN002709787, on the actual target path, with positive original IBD evidence. All 16 preserved normal PAINT rows match the original GAF, including five outside the current API node projection. Neither a short donor list nor the target appearing among sources is treated as invalidating the ancestry. Family records are retained unchanged; no additional GO assertion is manufactured from other family branches.
All 17 official product accessions and reconstructed FASTA identities are preserved. The displayed P02751-15 sequence contains EDA and EDB and is not labelled a plasma-specific product. Three incomplete VSP lists and a truncated synonym are corrected from the official table. Domain boundaries and splice boundaries are distinguished: VSP_060349 replaces residues 1722–1813 with TI. No binding or localization experiment is assigned to every product.
All 149 earlier quotation associations remain in the archive. Of these, 134 remain active byte-for-byte. Fifteen are individually archived with reasons: two absent UniProt strings, five excerpts about a different intervention, one wrong-partner excerpt, three associations attached to retired authored assertions/core, two titles that do not support the updated muscle-to-liver receptor-ligand claim, and two associations on the angiogenesis and endothelial-migration rows whose identical excerpt describes LPP3-induced FN expression rather than added-FN rescue. The valid enolase paper title is retained for the unresolved OppD row solely as source identification. No new quotation is introduced.
The snapshot manifest requires the unchanged original GOA table and the exact current projection. Their 226 and 219 data rows are parsed separately, checked against their declared hashes and target identity, then jointly passed to the existing strict matcher. The current heart-development context retains three WITH/FROM entities; the separate historical context at final index 234 preserves the older four-entity list, including ZFIN:ZDB-GENE-030131-6545. It is an existing source-backed non-core assertion, not a new functional proposal. The archive records its original data index 3, physical line 5 and every source field. See the snapshot validation documentation for identity limits and scoped source-only CI validation.
The original mouse Fn1 GO-CAM model supplies a receptor-ligand role during muscle activity, indexed under MGI:MGI:95566. Its source is PMID:36812915. This is separate from the scaffold contribution assessed for the retained extracellular-matrix-organization proposal.
The context archive is a gene-specific provenance companion for this augmentation, not a new general interchange format or an input to the GOA matcher. It keeps the original review and quotation associations together with the exact current raw response, conjunction/OR mapping, product corrections and two TSV snapshots. Git preserves prior committed files; the companion additionally joins the newly retrieved contexts to the individual review decisions and quotation retirements. Its presence does not imply that other gene reviews should acquire a similarly sized archive. Existing source-only consumers continue to read the original single GOA file; they are not advertised as union-aware.
The six non-microbial enzyme-partner contexts were also reassessed individually. TGM2 appears in cross-linked adhesion complexes, a direct immobilized-FN assay and a separate AP-MS record. The direct assay also measures low transamidating activity in the adhesive complex; that observation is retained rather than converted into a general requirement for catalytic regulation. The viral-polymerase fragment assay, LOXL2 cellular association and TSG-6 CUB_C/bridging experiments support their stated interactions without establishing a corresponding catalytic FN1 operation. All eighteen former enzyme-binding replacements are now non-core retentions; no original source assertion already used GO:0019899.
The apical-membrane transfer remains non-core with reference-only PMID:10742111 support. The official external side of apical membrane is part of apical plasma membrane and includes peripherally associated proteins; the term does not require an intrinsic membrane segment. This preserves the mouse-donor and abstract-only access limits above.
The retained matrix-organization and peptidase-activator judgments use the selected original primary evidence and public locators already listed above. The campaign instructions prohibit new quotations for this review, including newly fetched papers; reference-only evidence is therefore intentional. The angiogenesis and endothelial-migration rescue judgments remain supported by the inspected PMID:20123964 Results, while both mismatched inherited excerpt associations are preserved only in the archive.
The offline projection verifier passes the 222 archived original QuickGO objects through the repository's actual fetch_goa_data formatter and compares the resulting 219-row TSV byte-for-byte with the committed additive snapshot and its archive pin. Run python genes/human/FN1/FN1-replay-goa-snapshot.py from the repository root in the existing project environment. It performs no network request or file write. This proves reproducible projection of the retained response, not independent upstream authenticity or full-extension serialization.