The review is the decision authority for all 179 imported source contexts. The source tuples remain unchanged. Human UniProt P31371 is a 208-residue protein with a curated 4–208 chain; no alternative splice products were seeded. Experimental truncations and engineered monomeric mutants below are not natural splice products. Citations are reference-only, without new publication quotations.
PMID:8321227(https://pubmed.ncbi.nlm.nih.gov/8321227/) explicitly reports secretion from human-FGF9-cDNA-transfected COS cells in its abstract. The recovered author-upload original provides Results around Figure 5, printed page 4255: three CHO-derived product peaks have mitogenic activity and react with the C-terminal antibody. The 30/27-kDa products start at Ala2; a minor 21-kDa product starts at residue 33. These preparation-dependent products do not establish uniform processing of every FGF9 molecule. The Figure 4 operative passage was not recovered; the COS secretion statement is explicitly abstract-level. The immutable normal cache remains abstract-only. PMID:8428960(https://pubmed.ncbi.nlm.nih.gov/8428960/) supplies corroborating human glia-activating-factor purification/mitogenesis at abstract level only.
PMID:10506195(https://www.sciencedirect.com/science/article/pii/S0021925819519810), Results and Figures 1, 4 and 5, was read through indexed publisher text. Figure 4 identifies rat FGF9. Microsomal protection/glycosylation and COS-cell secretion support ER translocation despite lack of a conventional cleaved signal peptide. Central hydrophobic and amino-terminal contributions must not be reduced to a claim of cytosolic confinement. PMID:10713129(https://pubmed.ncbi.nlm.nih.gov/10713129/) reports ER/Golgi transit and a non-cleaved amino-terminal signal in its abstract; its operative body and construct species were not independently established here. Broad GO:0005737 cytoplasm includes organelles, so ER/Golgi transit is compatible with that compartment. Compatibility is not evidence for molecular activity there: these trafficking experiments do not show active human FGF9 in cytoplasm. The retained non-core compartment inference is supported by the literal positive PAINT IBD and verified target ancestry described below, not by transit. An endogenous human cytoplasmic mechanism remains experimentally unestablished.
The actual PMID:8663044 author PDF, Methods and Tables I–III (pages 15293–15296), was read; Table III pixels were inspected. Table II specifies mouse FGF9. The endpoint is receptor-dependent BaF3 mitogenesis, not direct affinity. Receptor species differ, and FGFR3/4 extracellular domains were joined to FGFR1 signaling domains. Responses depend on splice form and include FGFR3b; the experiment does not establish absolute c-isoform exclusivity.
PMID:28757146(https://pmc.ncbi.nlm.nih.gov/articles/PMC5587394/), Results, Figures 1 and 4–6 captions, and STAR Methods, supplies direct binding and signaling evidence. Human FGF9 constructs include residues 35–208, an amino-terminal truncation, and engineered monomeric mutants. SPR and receptor-dependent ERK responses distinguish receptor recognition from downstream activation. The crystal complex uses engineered FGFR1c P252R. These constructs are not equivalent to an unmodified 208-residue ligand or natural isoforms.
PMID:19219044(https://pmc.ncbi.nlm.nih.gov/articles/PMC2676118/), Results, Figures 3–8 captions and relevant Methods, separates two effects of mouse N143T: weaker dimerization/heparin binding with broader tissue distribution, but reduced mitogenic potency for several receptor forms. The 2017 study also finds an intrinsic receptor-potency effect. Thus increased signaling range need not mean greater receptor potency. Human multiple-synostoses association in PMID:19589401(https://pubmed.ncbi.nlm.nih.gov/19589401/) was read at abstract level; inconsistent S99D/S99N labels in that record and later narrative are not converted into a confident variant assertion. The separate structural/dimer abstracts PMID:11060292, PMID:11223514 and PMID:19564416 are corroboration, not claimed full experimental reads.
PMID:17848411(https://doi.org/10.1210/jc.2007-1690), author-upload Methods, FGF9 Results and Figure 4 caption, measures human second-trimester fetal-testis transcripts. Expression increases per testis but not per Sertoli cell. The retained IEP context is observational, not an FGF9 perturbation or proof of a human sex-determination mechanism.
PMID:22926577(https://doi.org/10.1007/s11064-012-0874-2), author-upload Methods and Tables 1–2, explicitly identifies P31371/glia-activating factor in adult postmortem human substantia-nigra disease/control proteomics. Table 2 is a positive target join, but its functional category does not demonstrate formation or maturation of the substantia nigra. This is the basis for marking the developmental HEP interpretation over-annotated, without alleging a wrong-gene result.
PMID:29118903 publisher PDF, Methods and Figure 3E (page 4407, pixels inspected), shows FGF9 expression rescuing a miR-665-associated CCK-8 decrease in a human aortic smooth-muscle cell line. The purchased construct is not sequence-resolved. This does not assign every miRNA-dependent migration/Wnt phenotype to FGF9. Figure 2 caption directions disagree with Results, and Figure 3A labels 5′ UTR while Methods/caption state 3′ UTR; neither changes the visible rescue direction in Figure 3E.
PMID:28259995(https://pmc.ncbi.nlm.nih.gov/articles/PMC5360430/), rat aortic SMC Methods and Results/Figures 5–6, supports the P36364 donor annotations: FGF9 knockdown reduces proliferation/migration and increases SMA/SM22α contractile markers. Interpretation follows these readouts despite inconsistent differentiation wording. PDGFRβ expression changes do not demonstrate FGF9–PDGFRβ binding. PMID:21858205(https://pmc.ncbi.nlm.nih.gov/articles/PMC3157460/), Results/Figures 8–9 and ocular Methods, supports mouse donor eye-development scope; a neomorphic Y162C phenotype is not assumed to be simple loss of function or a direct human experiment.
FGF9-phylogeny.yaml records the complete PTHR11486 target path, exact 208-residue MSA match, nine inherited positive IBD assertions and zero target-path IRDs after checking all 35 saved rows. FGF12 loss nodes are on another branch. The positive cytoplasm IBD supplies the support for the inherited compartment assertion. Neither a selected subset of donor functions nor different review actions for other FGF genes demonstrates target-specific loss or, by itself, overturns that node placement. Absence of a recorded loss is not treated as an independent positive experiment. The official IPR002209 growth-factor mapping matches P31371 residues 48–193 and agrees with mitogenic evidence.
FGF9-reactome-contexts.yaml retains all 152 event-source joins. Every event participant expansion was recovered and checked for P31371. The explicit ligand route R-HSA-190256 → R-HSA-189953 → R-HSA-189870 identifies extracellular human P31371 residues 4–208. A ligand-containing receptor set can participate in intracellular enzyme reactions without making FGF9 a kinase, GEF or lipid enzyme. These records support the same ligand-location claim, not 152 distinct biological activities. Earlier 429/521 responses remain preserved alongside the eventual successful captures.
Official typed ancestry places GO:0008543 below GO:0007165; no strict relationship with GO:0007267 was established. Nevertheless the two PMID:8575785 TAS contexts remain UNDECIDED, because only bibliographic metadata was recovered. Independent signaling evidence is retained elsewhere rather than substituted for that unavailable source. The single core combines receptor recognition, growth-factor activity, FGFR signaling and proliferation; downstream and tissue-specific effects remain contextual. Sequence-collection references were assessed as provenance with uninspected target clones, not functional experiments.