Fibronectin supplies extracellular structural material and distinct ligand-binding surfaces. The review represents five core activities: extracellular matrix structural contribution, integrin binding, collagen binding, heparin binding and proteoglycan binding. The source guide gives the actual access and assay boundaries; the context archive preserves the original review and complete current source projection.
The structural core rests on fibronectin polymerization and scaffold maintenance, including actual human-cell matrix-blockade experiments. Separate reconstitution experiments establish retention of thrombospondin and collagen fibrils. Integrin binding is supported by defined human receptor/domain assays. The FNIII10 RGD mechanism is not generalized to α4β1 or every other receptor. Collagen, heparin and proteoglycan recognition have independent molecular evidence and are not each assigned speculative developmental processes.
The previous authored extracellular-matrix-organization proposal is retained with individually rebuilt primary IMP support from PMID:12388756. Fibronectin itself provides the scaffold, beyond merely being required as a substrate. Same-role comparator annotations and official process ancestry were checked. The GO-CAM index includes mouse Fn1 (MGI:MGI:95566) in model 645d887900000439, with receptor ligand activity during the response to muscle activity. That muscle-to-liver signaling role is already considered in the donor-based GO:0048018 annotation; it does not model or contradict the distinct extracellular-matrix scaffold operation supporting GO:0030198. The earlier no-entry statement was too broad. The two other authored proposals are retired: fibrinogen binding was supported by a fibrin paper, and broad coagulation duplicated the existing source-specific GO:0072378 process. Older positive fibrinogen-binding literature prevents a universal negative conclusion. Their complete original objects remain in the archive.
The 235 active rows comprise 223 typed current source contexts, eleven retained historical machine-source contexts and the rebuilt authored proposal. Counts are 53 ACCEPT, 120 KEEP_AS_NON_CORE, 24 MODIFY, 22 UNDECIDED, twelve MARK_AS_OVER_ANNOTATED, three TOO_GENERAL and one NEW. There are no remaining REMOVE decisions. This is the result of individual source review; current-response absence, a different gene in a title, an unavailable experiment or a missing database pair is not used as a blanket removal rule.
Seventy generic-binding contexts were individually assessed, including two historical source rows. The resulting 52 non-core retentions, eight specific replacements and ten uncertainties preserve each original partner and evidence context. Twelve experimentally supported microbial adhesion interactions are retained as non-core: the partner’s catalytic class does not describe the adhesion function measured in those assays. The other six enzyme-partner contexts are also retained as non-core after individual review of their adhesive, structural or co-complex evidence. The enzyme-binding definition does not require catalytic regulation, but classification by the partner’s enzyme identity does not add a demonstrated FN1 functional distinction in these contexts. Processed viral chains, inactive propeptides, co-complex assays and co-migrating overlay bands are distinguished. The separate nonbinding and location reviews retain source-specific soluble, matrix, fragment, donor and fusion boundaries.
Source reconciliation preserves the old fetched GOA, UniProt and provider snapshots. The archive records all 196 original annotations, every one of the 222 current original QuickGO objects, each OR branch, all conjunctions and the final index map. Only authored slots 194 and 195 are removed from the active array; all 193 historical machine-source rows remain represented. Current source extensions and partner fields enrich their matched historical rows, while genuinely distinct contexts are added individually.
The 17 alternative-product identities are retained with verified sequences. Corrections address inverted EDA/EDB statements, unsupported plasma-product assignments, incomplete sequence-feature lists and a truncated synonym. No product is given a new experimental function. Existing full-length and fragment results retain their construct limits.
The archive preserves all 149 inherited reference/text associations. Exact deterministic matching supports 147; two strings attributed to UniProt are absent from that record. Additional source alignment identifies five FBLN1C1 excerpts attached to FN-specific endpoints and a GlnA1 excerpt attached to DnaK. These are archived individually, together with three associations on the retired authored proposals/core. Two further inherited paper titles are retained only in provenance because they do not support the updated muscle-to-liver receptor-ligand claim. Both inherited associations of the LPP3-induced FN-expression excerpt, on angiogenesis and endothelial cell migration, are also archived because they do not support the separate added-FN rescue claims. The 134 remaining active excerpts are byte-identical to their originals; no quotation is generated, extended, normalized in storage or transferred to an appended annotation.
Unresolved questions remain explicit. The exact soluble-fibrinogen-complex, selected protease/self-binding, monocyte readout and several bacterial-pair assays are not reconstructed from titles or general dataset validation. These annotations remain undecided with authentic source identity and the limits of the accessible evidence documented.
The historical heart-development IBA is also retained as a separate source-backed row with its exact older four-donor group. The current three-donor row remains unchanged. Both represent the same verified ancestral developmental placement with contextual scope; changing a donor list does not establish target loss or a new human experiment. Joint validation uses the explicit immutable-snapshot manifest, preserving both source files and all earlier review evidence.
The retained historical rows audit source assertions that were actually observed; they are not a claim that every assertion remains in the latest live GOA response. The archive identifies current and historical-only contexts separately. The two heart-development donor groups represent the same ancestral biological role, and obsolete-term/current-term pairs can likewise converge on one reviewed function. Counts of source contexts must not be reported as counts of distinct functions or independent experiments.
The twelve disease FN1–ALK events are over-annotated when exported as unqualified native P02751 localization. The underlying fusion graphs remain valid evidence for the fusion entities themselves. Source-defect records request a check of that export context without inventing a corrected identifier. Independent native-FN membrane association, including the donor-derived apical surface context, is assessed separately.
A suggested question now distinguishes fibrin recognition from soluble or immobilized fibrinogen assays. It introduces no new annotation, core or quotation. The description states biological product variation; assay-to-product transfer limits remain in individual reasons and this source guide.