The review covers 58 original QuickGO contexts, which the genuine repository formatter and seed project to 56 annotations. It preserves all three natural-product records and restores the three original extension conjunctions (seven atoms). Source identities, evidence codes, qualifiers, partners and order are retained. The dependency set contains 19 publications and three PANTHER files; 16 dependencies reuse exact main bytes. No publication quotations or NEW annotations are introduced.
The evidence record gives source hashes, complete extension atoms, sequence reconstruction and target-path checks. The notes record the scientific decisions and subsequent validation history.
The two activity summaries distinguish direct Hsp90 ATPase inhibition/cochaperone activity from FNIP1 participation in the FLCN–FNIP1 RagC/D GAP complex. PMID:24095279 directly assays purified FLCN–FNIP1 complex activity, reporting an apparent RagD preference. Core 2 retains the subunit-specific FLCN binding term and explicitly records contributes_to GO:0005096. The original enzyme-activator annotation is modified to that complex contribution; its original qualifier and extension conjunction remain source provenance. Recruitment, Rag association and nucleotide-state epistasis alone are not substituted for the direct complex assay. The isolated inactive controls tested FLCN and FNIP2, not FNIP1.
FNIP1-dependent effects on B-cell development and survival are retained with species, developmental-stage and nutrient limits. The positive-apoptosis ISS term is modified to negative regulation of B-cell apoptotic process: actual mouse loss experiments show increased death, not a pro-apoptotic role for wild-type Fnip1. The human source assertion remains intact. The two 2012 mouse studies have different mTOR/rapamycin outcomes and are not merged into a universal TOR response.
| Official product | Length | Sequence interpretation |
|---|---|---|
| Q8TF40-1 | 1166 aa | Displayed sequence; exact ungapped PANTHER target match. |
| Q8TF40-2 | 508 aa | VSP_028982 and VSP_028983 reconstructed and matched to official FASTA. |
| Q8TF40-3 | 1138 aa | VSP_028984 reconstructed and matched to official FASTA. |
These are sequence identities, not three experimentally equivalent activities. The C-terminal 929–1166 cochaperone fragment is an experimental construct, not a fourth natural product. The BC001956.1 accession cited in the 2013 Methods is a human 508-aa clone that differs from current Q8TF40-2 at residue 76. Neither its description as full-length nor its length resolves the construct as canonical FNIP1 or current product 2. No sequence in an immutable source is repaired.
The preserved PTHR21634 tree contains 116 nodes and its MSA 116 sequences. Human target PTN002502545 has the exact 1166-aa Q8TF40-1 sequence. All three original IBD assertions map to ancestor PTN001092003, whose recorded species and taxonomic range are Unikonts. ATPase inhibition and protein-folding chaperone binding remain supported; the broad cytoplasm assertion is treated as TOO_GENERAL alongside more specific accepted locations, without disputing the positive descent. Target self-evidence is legitimate and donor count is not a quality score. API release 19 and the original PAINT GAF snapshot retain separate version provenance. IBA citations to the evidence record specifically refer to its phylogeny section; machine_contexts preserves the original source assertions and is not new functional support.
A full-text flag describes cache acquisition, not how much was inspected. The following assessments name selected original sections and limits. No complete supplement, whole-paper or figure-pixel audit is implied. Separate authentic originals were inspected for several abstract-only caches; those immutable caches remain unchanged.
| Source | Access actually used | Scope |
|---|---|---|
| Original study 17028174 | Selected original cached Results and Methods | The human 1166-aa construct was assembled from overlapping cDNAs. Tagged and endogenous FLCN association, reticulocyte-translation/GST capture and AMPK-complex association are distinct assays. FNIP1 is an AMPK substrate; regulatory gamma/beta-subunit capture does not make those partners protein kinases or establish FNIP1 catalysis. Nutrient/AMPK experiments here primarily examine phosphorylation, not FNIP1-loss TOR inhibition. No SI or figure pixels independently inspected. |
| Original study 18403135 | Selected original cached Results and Methods | Human FNIP1/FNIP2 reciprocal endogenous and tagged capture, reticulocyte-labeled FNIP2 capture by GST-FNIP1, and sequential affinity capture support association/multimers. These assays do not establish a fixed dimer stoichiometry or equal behavior of all current FNIP1 splice products. |
| Original study 18663353 | Authentic normal abstract only; selected source-specific assay not recovered | The abstract foregrounds FNIP2/FNIPL. The particular FNIP1 localization experiment remains unverified; no wrong-gene or paralog-error claim is made. The cytoplasmic location is independently supported by other FNIP1 studies. |
| Original study 19914239 | Selected official publisher-indexed Results and supplementary caption; normal cache remains abstract-only | Indexed primary text describes COS7 cells expressing Flag-tagged rat FLCN variants with Myc-FNIP1/FNIP2 and anti-Myc capture. S62 phosphorylation-state variants did not materially alter association. This supports the original rat-FLCN partner context. Complete Methods, supplementary figure pixels and exact FNIP1 construct sequence were not inspected. Direct publisher access returned a block; the successful indexed text is a separate access route. |
| Original study 20562859 | Normal full article plus exact curated target-specific interaction records; no complete supplementary target validation audit | Exact positive target records support human FNIP1 associations with the two ATG8-family proteins, including affinity-capture contexts. Retain their distinct partner identities. This is not a demonstration that FNIP1 is a canonical autophagy cargo receptor, nor proof that every listed high-throughput edge was independently validated in native cells. |
| Original study 21209915 | Selected original cached Results and relevant assay text | FNIP1 reduces TFE3-driven GPNMB reporter activity in HT1080 cells; FLCN/FNIP combinations have stronger selected effects. FNIP1 alone does not suppress GPNMB in FLCN-null UOK257 cells. These source-specific results support contextual regulation rather than DNA binding or universal repression of RNA polymerase II. Preserve the original TFE3 participant extension. |
| Original study 22190034 | Normal abstract and three exact positive IntAct target records; original full-body access unsuccessful | Curated records preserve the HIV protease partner, D25-mutant bait and affinity/coimmunoprecipitation/spoke context. They support a measured association without demonstrating FNIP1 cleavage, substrate status or binding to a physiologically active protease conformation. No target-specific full-paper or SI validation is claimed. |
| Original study 22608497 | Selected original normal Results, Discussion and Methods | Mouse Fnip1 loss causes a cell-autonomous early B-cell defect and increased susceptibility to apoptosis under tested stresses; wild-type Fnip1 re-expression restores development. AMPK activation persists but fails to restrain S6 in the knockout. Nutrient and rapamycin rescue are conditional. The donor annotation has positive-apoptosis direction, whereas these experiments support a survival-promoting/negative-apoptosis contribution. Preserve human ISS provenance and document the source-side discrepancy; do not silently change its original term. |
| Original study 22709692 | Selected original normal Results, Discussion and Methods | Independent mouse Fnip1-loss models arrest early B-cell development and increase death. Bcl2 and prearranged-receptor experiments support survival/developmental contributions with distinct endpoints. This study did not observe the same mTOR increase or rapamycin rescue as PMID:22608497; the models must remain distinct. Conditional Flcn comparisons are separate from Fnip1-specific rescue experiments. |
| Original study 24081491 | Selected original normal FNIP1 Results, Discussion, figure captions and Methods | FNIP1 depletion and overexpression support FLCN lysosomal recruitment and Rag association in human cells. They do not directly measure FNIP1 increasing FLCN GAP catalytic activity. The FNIP1 construct accession resolves to a human 508-aa sequence, differing by one residue from current Q8TF40-2; the paper calls it full-length. Do not assign canonical or current isoform identity from this unresolved discrepancy. Mouse FLCN constructs and endogenous human FLCN experiments are separate. |
| Original study 24095279 | Selected original cached Results and Methods | FLAG-FNIP1/HA-FLCN complexes from HEK293T cells were affinity- and size-exclusion-purified for radiolabeled Rag nucleotide-hydrolysis/TLC assays. FLCN–FNIP1 activity was reported as lower than the FNIP2 complex with apparent RagD preference. This is target-specific complex activity, not independent FNIP1 catalysis. The inactive isolated controls are FLCN and FNIP2; no FNIP1-alone or current-natural-product equivalence is established. Figure pixels and SI were not inspected. |
| Original study 25126726 | Selected actual original Results/captions from retained PMC body beyond incomplete normal cache | FNIP1 increases FLCN–GABARAP capture in human cells, with FNIP2 compared separately. GST-GABARAP capture uses cell lysate; it is not an isolated three-purified-protein affinity measurement. Tagged MDCK localization emphasized FNIP2 and cannot be reassigned to FNIP1. Retain both original participant inputs without adding autophagy-adaptor or cleavage functions. |
| Original study 27353360 | Selected original full XML Results and ATPase/binding Methods | Purified bacterial FNIP1 binds Hsp90 and FLCN and permits their association. Human-cell complex experiments identify Hsp90, HOP, p23, CDC37, GR, BRAF and CDK4 partners. Full-length FNIP1 and residues 929–1166 inhibit Hsp90 ATPase in the reported assay; that ATPase experiment used cell-purified, salt-washed components, distinct from bacterial direct-binding experiments. Client maturation/loading is supported, without assigning every co-complex edge direct binary binding or every natural product equivalent activity. |
| Original study 29848618 | Selected original full XML via independent CC consultation, beyond abstract-only normal cache | Endogenous-locus 2×HA-FNIP1 in human HeLa cells shows nutrient-regulated LAMP1 colocalization and RagC-dependent lysosomal recruitment, while remaining associated with FLCN. Fed cytosolic and starved membrane-associated pools are compatible. No complete splice-product assignment, quantitative soluble-fraction census, figure-pixel or SI inspection is claimed. |
| Original study 30699359 | Selected original normal Results and Methods | OGT, PP5 and CK2 associate with FNIP1 and modify it; the matched glycosylation/phosphorylation/dephosphorylation assays establish regulated-substrate roles rather than FNIP1 catalysis. FNIP1 domain/mutant effects on Hsp90 ATPase support a cochaperone rheostat within the tested constructs. |
| Original study 32905580 | Selected original PMC Results and captions; main cache remains abstract-only | Human biallelic FNIP1 deficiency is associated with severe B-cell deficiency. The sampled marrow has relative early-stage enrichment and fewer immature B cells without an overt complete block. Mitochondrial number/activity increased in the measured cells; selected signaling readouts differ rather than demonstrating uniform TOR activation. Supplementary Methods were not inspected. Clinical evidence supports a human developmental context without a purified molecular mechanism. |
| Original study 34597140 | Selected original normal Results, captions and construct/SPR Methods | Human FNIP1 wild-type and Y583A/V586A LIR mutants support GABARAP-dependent sequestration and TFEB-selective signaling control. FNIP2 peptide structures and tight-binding measurements are distinct from FNIP1 full-complex experiments. This supports an FNIP1 role in basal FLCN-complex function and condition-dependent inhibition by sequestration, not a uniform sign for every mTOR output. |
| Original study 37079666 | Selected original normal Results, Discussion and Methods | Human-cell FNIP1 phosphorylation mutants, AMPK stimulation and locked-state RagC epistasis place FNIP1 in the FLCN–Rag–TFEB/TFE3 pathway. This supports positive TORC1 control toward these substrates in the basal setting and release after AMPK activation. It does not directly establish isolated FNIP1 GAP catalysis or the exact direct enzyme-activation conjunction. Lysosomal gene/volume responses are downstream contextual functions. |
| Original study 39423796 | Selected original normal Results and captions | The inspected STING–GABARAP–RagC experiments are compatible with the FLCN–FNIP signaling route and cite the earlier direct FNIP1 work. An FNIP1-specific perturbation establishing the exact curated IDA assertion was not isolated in the selected text. Retain the distinction between that source-specific access limit and independent positive FNIP1 evidence in PMID:34597140 and PMID:37079666. |
The published project instruction requires supported generic binding to remain KEEP_AS_NON_CORE when no evidence-backed informative replacement is established. This user instruction overrides the skill default for this campaign; validators remain unchanged. Ten stored binding contexts receive specific partner-class replacements. Ten retain generic binding with their assay limits. The NR3C1, CDK4 and BRAF client associations from PMID:27353360 are retained as non-core because the source-specific Hsp90-network assays do not isolate a FNIP1-proximal client contact; this is not a universal requirement for purified binary assays. An AMPK regulatory gamma/beta chain is not relabeled as a catalytic kinase, and co-chaperone or viral-protease-mutant capture is not promoted to an unsupported enzymatic mechanism.
All indices are zero-based. Predicates use schema-supported CURIE strings, while original relation names and atom order are preserved in the evidence record. No OR alternatives were flattened.
A genuine configured deep-research attempt with falcon and perplexity-lite fallback failed authentication/access. No provider research file was invented. Manual selected-primary reading, official target/partner metadata and original annotation/PAINT records supplied the review evidence.
The human apoptosis-direction assertion is recorded against its assigned resource, UNIPROT. The two separate upstream MGI donor records retain their exact QuickGO observation identifiers, explicitly distinguished from native MGI annotation IDs. All three proposed corrections are NOT_REPORTED with unknown cause. The donor IMP extension remains unchanged and is flagged for curator consistency review.
The retained BC001956.1 accession sequence is established. Its mismatch to current natural-product sequences is a construct-provenance issue for source clarification; the experimental knowledge question concerns the activities of the three natural products.