The review links each judgment to its supporting references. Six excerpts are centralized in reference findings. This guide identifies primary assay locations and limits. The current YAML evidence archive preserves complete machine contexts and partial-projection semantics. The original JSON archive remains unchanged as historical provenance and to preserve the original journal link.
The official Q9UK05 record and direct FASTA identify a 429-residue precursor, signal sequence 1–22, prodomain 23–319 and mature ligand 320–429 (PRO_0000033903). No alternative splice products are declared. The functional class describes this cleavage product; it does not convert processing, prodomain association or redox states into splice isoforms. The two IntAct contexts explicitly target Q9UK05-PRO_0000033903 and remain distinguishable from whole-gene records.
17068149. Human microvascular endothelial experiments used BMP9 with ALK1 perturbation/trapping. Scratch closure measures migration; WST-1 indirectly estimates viable-cell number. It does not trace division or separate survival from proliferation. Figure 1 reports SMAD phosphorylation/ALK1 dependence; Figure 5A reports scratch closure and 5B the WST-1 response. Selected text/captions were read, not images.
20406889. Mouse allantoic, progenitor-derived and adult endothelial assays differ from human BMP9-transgene xenografts. Increased vascular area can be disorganized. Human HMVEC-d inhibition is reported without displayed data. Differentiation is confounded by cell yield. Figure 2 shows the allantoic vascular response; the Discussion identifies ALK1/endoglin induction and proposed feedback. Selected Methods/Results/Discussion were read.
Original main text, Experimental Procedures, “Thymidine Incorporation”: 18 hours of BMP9 exposure preceded a six-hour labeling period, followed by bulk scintillation counting. Cell-number normalization was not stated. Figures 2–3 and selected Results identify phosphorylation and transcript responses. These original-text locators are outside the partial cache. IL-8 protein methodology alone does not establish a positive protein result. No figure pixels were inspected.
32561494. Primary human colon monolayers received BMP9 with or without ALK1-Fc. Notch-target, colonocyte and junctional transcripts plus barrier measurements support contextual responses. TEER is not junction-assembly imaging; SMAD discussion is background rather than a new phospho-SMAD experiment. Selected Results/Methods and captions were read, not images or supplements.
39453386. Selected JATS Results/Methods concern combined stellate-cell Gdf2/Bmp10 deletion. Sinusoidal identity loss and focal hepatic iron loading do not isolate GDF2-specific differentiation or intracellular iron handling.
25751889. Selected main text and deposition sequences identify mouse prodomain with human mature ligand in 4YCG. The 4YCG entity records label entity 1 as growth-factor domain and entity 2 as prodomain, opposite their sequence assignments. The reference assessment records this metadata issue for checking; full supplementary methods were not inspected.
23972370. Selected original Methods/Results and captions distinguish human variant processing/signaling from zebrafish vascular remodeling. No human-ligand rescue of fish was established in the inspected text. Direct payload failures were retained; indexed original text was read. No images or supplementary experiments were independently inspected.
19903896. Selected original Methods/Results separate human BMP9 signaling from mouse receptor-trap phenotypes. A receptor trap can sequester more than one ligand. Correlated phospho-SMAD/ID1 observations are described as data not shown; human lymphatic-cell proliferation was not increased in the reported two-dimensional assay.
32082362. The authentic original is PMC7002562. Selected Methods/Results assay human BMP9 stimulation of HUVECs. The microRNA intervention targets ALK1 rather than GDF2; the separate xenograft experiments do not establish a ligand-specific intervention.
28564608. Selected original Methods/Results identify human BMP9 and ENG constructs, direct binding/interface evidence and a separate ALK1–BMP9–ENG pull-down. Ternary geometry is modeled using prior structural information; it is not a new experimentally resolved full ternary structure.
19956691. Selected Methods/Results explicitly include mouse Bmp9 constructs and recombinant human BMP9 in distinct chicken micromass/limb experiments. The GDF5-centered title does not exclude these BMP9 assays. Cartilage induction and engineered antagonist resistance are kept separate from natural human product claims.
11748585. The authentic abstract reports cartilage-marker induction in human mesenchymal cells. Full original methods were not recovered; cartilage markers alone are not evidence of ossification.
12925636. The authentic abstract reports human BMP constructs and mineralized nodules. A linked 2004 erratum remains inaccessible; no quantitative hierarchy or full-methods verification is claimed.
16801541. The authentic abstract includes human BMP9 treatment of mouse hepatocytes and human HepG2 cells. Hepcidin transcription, rather than intracellular iron concentration, is the recovered endpoint; original full-text attempts failed.
21710321. Only the authentic abstract was read. The linked erratum DOI10.1007/s00018-011-0776-5 remains unresolved. Qualitative extracellular precursor/procomplex distinctions are used without adopting exact circulating percentages.
19056867. Selected original main text describes urinary proteomics. The target-level supplemental accession/peptide entry was not recovered. Neither absence of GDF2 nor vesicle-specific residence is inferred from that access limitation.
22799562. Selected full Methods/Results distinguish direct human ligand–receptor SPR from antagonist phospho-SMAD assays in mouse NIH-3T3 cells. NMR resolves unbound ALK1; ligand orientation is a docking model. No images or supplements were independently inspected.
25237187. Selected full Methods/Results distinguish two solution dimers and redox-dependent stability. Human-serum experiments spike recombinant ligand; they do not directly measure endogenous clearance. C73S concentration was inferred from prodomain immunoblotting. No images or supplementary experiments were independently inspected.
The other reference assessments state when only an authentic abstract was inspected, even if a full cache is available. Failure to retrieve full text does not establish a wrong gene, false result or absent experiment. Reported recipient species, expressed constructs and ligand species remain separate.
The actual PTHR11848 tree, MIA, alignment and attributes place the sequence-verified human target at AN350/PTN002483793. Cytokine activity and extracellular location descend from AN0/PTN000218063; BMP signaling and positive endothelial proliferation descend from AN340/PTN008330932. The four literal donor sets are archived. Human self-evidence and the single human BMP-pathway donor are legitimate inputs to the ancestral assertion. No negative assertion occurs on the target lineage. This is a node-placement review, not a donor-count score.
The official R-HSA-8858369 event has input set R-HSA-8858373, including GDF2 dimer R-HSA-3222350. Its repeated component R-HSA-3222335 maps to Q9UK05 residues 320–429 in the extracellular region. The receptor's membrane compartment is not assigned to its soluble ligand. R-HSA-201451 supplies pathway context rather than a separate target experiment.
All 61 original objects and the 16-column export are retained. All 61 actual native indices are reconciled against the genuine seed and export. Existing term labels, qualifiers, evidence codes, original references and support memberships are not silently modernized. In particular, the Notch assertion retains acts_upstream_of and two extracellular assertions retain is_active_in.
Thirteen source records contain 30 literal extension-atom occurrences. Ten single-group records project completely, including the two-atom conjunction on the 19903896 SMAD row. The 32561494 BMP row has three alternative groups with a common occurs_in colon epithelial cell atom; the 19366699 SMAD row has three alternatives with a common part_of BMP signaling atom. Only those entailed common atoms are projected. The seven alternative transcriptional inputs have no shared atom and receive no flat projection. These 13 projected atoms are not equivalent to the three original OR expressions: all alternatives remain intact in the archive. No extra original is manufactured.
Predicate identifiers were verified against official OLS property results: BFO:0000066 occurs in, BFO:0000050 part of, RO:0002233 has input and RO:0002629 directly positively regulates. The containing ontology name does not replace the CURIE namespace. Native predicate values use CURIE strings, as required by the current schema; their labels remain in the archive. Processed-chain target restoration uses the literal target identifier in the existing isoform slot, with the cleavage-product interpretation stated explicitly.
Current GO ancestry does not place positive regulation of gene expression (GO:0010628) above GO:0045944 through is_a/part_of; the latter does descend from GO:0045893. The retained DNA-replication regulation claim is limited to the population response. It does not assert per-cell replication control. WST-1 instead fails to measure the cell-size increase meant by GO cell growth; these are different readout/definition comparisons. Iron-homeostasis and tight-junction-assembly uncertainties concern specific claims, not whether a source could be opened. GO:0060586 concerns extracellular body fluids and is not substituted for intracellular iron solely on the basis of hepcidin induction.
The single core uses cytokine activity, a receptor-ligand activity, alongside BMP signaling and positive SMAD signaling regulation. Receptor recognition is its mechanism, not a separate duplicated core. This follows Decision 4 in the current ligand guidance. The three generic-binding replacements retain the experimentally supported BMP-receptor partner class. The two core processes are not strict ancestors of one another; growth-factor activity remains contextual.
The literal has_input GDF2 extensions describe input to the signaling process. They do not assert self-binding, and no source defect is inferred merely from the matching subject accession.
Source acquisition and validation outcomes belong to the evidence archive and append-only native history. The historical JSON is retained once alongside its YAML successor to preserve the original journal link and complete provenance.
GO:0032924 is defined by an extracellular ligand engaging an activin receptor and downstream signaling. It is not restricted to activin-family ligands. The current official definition and the ActRIIA perturbation result support the retained pathway assignment.