Thirty direct contexts across 16 whole PMID groups were reviewed. The lane retains all 60 imported source objects and their order, both seeded splice objects, all 32 reference identities/titles, and zero quotations. It proposes no NEW annotation and leaves database decisions, description, core synthesis and computed status to their separate stages.
FBN2 forms microfibrillar matrix with structural and growth-factor-organizing roles. Human coassembly/localization and biochemical interaction experiments are kept distinct from mouse developmental signaling and bovine tissue ultrastructure. Broad structural function remains useful beyond elasticity. Eye-development claims retain their IEP and structural-inference scope, without a new causal morphogenesis claim. The FBN1-focused renal paper actually assays human FBN2.
The sole undecided direct context is the PMID28327460 HDA target identification. Its ambiguous table is not proof of a curator error. Independent same-gene scaffold evidence supports the generic structural classification but cannot reconstruct that assay. In contrast, the PMID23658023 main table and PMID28675934 original workbook supply explicit FBN2 detections. Fourteen positive generic-binding rows remain non-core under the standing project instruction.
Placensin is a distinct released C-terminal product with human secretion evidence and recombinant human ligand activity in mouse metabolic assays. Exact endogenous processing remains unresolved: Methods 2780–2912 and Figure 2 caption 2779–2912 differ, and the furin site was predicted rather than mapped. The displayed precursor and the truncated 1,473-aa splice product must not share an assumed C-terminal hormone output. Cleavage-product proposals remain separate from the unchanged splice records.
Detailed biological limits and stable article/figure locators are in the direct source guide, with source-keyed decisions, binding assays and product evidence. The genuine failed deep-research provider attempt remains preserved in the intake; no provider report was fabricated.
All 60 contexts are now reviewed: 15 ACCEPT, 34 KEEP_AS_NON_CORE, nine TOO_GENERAL, one MODIFY and one UNDECIDED. The 32 complete reference assessments retain all six independently read shared-source additions. The earlier direct-lane entry above is preserved as a journal record of that stage.
Three distinct molecular functions organize the biology: extracellular matrix structure, calcium binding in the EGF-like domain architecture, and hormone activity of placensin. Matrix sequestration provides the mechanism for contextual TGF-β limitation; it is not receptor enzymology or an obligatory mineralization substrate. The hormonal product regulates recipient-cell glucose metabolism rather than carrying out glucose-conversion chemistry. Elasticity, bone and ocular developmental contexts remain positive non-core roles. No generic protein-binding core or new annotation was added.
The two cleavage-product classes are now applied in functional_isoforms, while both original splice records remain exact. Their chain coordinates are curated metadata, not newly mapped endogenous termini. In particular, the truncated 1,473-aa splice product lacks the placensin interval and is assigned no assumed hormone production. The biological gap and suggested experiment concern endogenous terminal mapping and furin dependence with secretion controls, limited to what the examined source establishes.
The genuine research-provider attempt failed, including its configured fallback. That failure is retained as provenance; no provider report was fabricated. Literature reading, primary-source recovery and database-source evaluation supplied the evidence used here. The one unresolved HDA target join is retained explicitly, so the final status follows the repository's computed convention rather than completion of every reading task.
See the integrated source guide, database evidence, database journal, and applied product evidence. The complete source tables preserve all original machine assertions and their reviewed scope.
Six generic binding annotations are refined from actual partner-resolved assays: five BMP-family prodomain interactions use GO:0036122 (PMID18339631, Figure 6 and Table 3), and the FN1 interaction uses GO:0001968 (PMID25034023, Results 3.3 and Figure 3B). The BMP term has no mature-domain restriction; GDF5 is also BMP14 in the verified human partner record. The experiments do not establish mature growth-factor receptor activity or pathway direction. Other supported generic binding remains non-core, and no extra binding core or natural splice-product function is introduced.
The description and structural core now explicitly include the microfibrillar scaffold for early elastic-fiber assembly as well as nonelastic matrices. UniProt FUNCTION uses ortholog evidence for early assembly; PMID8120105 directly establishes target sequence and preferential elastic-matrix localization. Its target EGF-like architecture and exact UniProt FT DOMAIN features now support the calcium core in addition to the InterPro mapping-method reference. This does not turn a domain inference into a new target calcium-affinity experiment. CCA and early-onset macular degeneration are included as variant-associated disease contexts, without asserting one loss-of-function mechanism for all dominant variants (UniProt DISEASE; PMID24899048 family Results/Discussion).
The processing gap explicitly engages UniProt P35556: SUBCELLULAR LOCATION annotates linked secretion followed by furin separation with ECO:0000269/PMID32329225. The curated model and positive placensin release remain accepted. The actual article identifies a predicted PiTou site (Figure 1), furin colocalization (Figure EV1), recombinant fragment expression (Figure 2 and Methods) and secreted-product IP-MS after digestion (Figure EV3 and Methods). These data leave intact endogenous termini and furin-specific dependence to be measured, despite supporting the model. The two applied functional-product objects and both splice records remain unchanged. The shorter splice form is not assigned placensin production.
Stable original locators and their scope are in the updated direct source guide. All 60 imported source objects remain exact; no supporting quotation, new annotation, cache alteration or new core function was introduced. Three existing reference assessments receive append-only clarifications. The separate database feedback owns same-term action coherence and the placensin metabolic-process refinement.
The previous integrated counts are superseded: all 60 source contexts now have 22 ACCEPT, 29 KEEP_AS_NON_CORE, seven MODIFY, one TOO_GENERAL and one UNDECIDED decisions. Six supported generic binding rows use specific BMP or fibronectin binding replacements. Eleven extracellular-region records are consistently accepted; positive matrix records are non-core, while the unresolved PMID28327460 target identification remains undecided. Placensin uses positive regulation of gluconeogenesis instead of its broader parent term.
The revision adds bounded elastic-fiber and disease context, actual target-domain support for calcium binding, and explicit reconciliation of the curated furin-processing model with the original experiment limitations. Both independently read feedback scopes are retained in the PMID32329225 reference assessment. All 60 imported source objects, 32 reference identities and inherited fields, two splice records, two functional products, original caches and prior journal entries are preserved. No new annotation or supporting quotation is added.