Human ALX1, HGNC:1494, UniProt Q15699, is the approved symbol; CART1 is its previous symbol. The seed6 record contains 44 annotations, 13 original reference identities and no seeded alternative-product block. The initial current-main and separate ALX1/CART1 open-PR checks were empty. Original annotation terms, evidence, supporting entities and ordinary qualifiers are preserved. UniProt, GOA, normal publication caches and family outputs are not edited. The source baseline and checks are retained in the campaign's temporary audit receipts.
The normal Falcon research command with perplexity-lite fallback and the installed matching client was actually attempted and failed; it produced no provider report. The normal gene-PMID command found all six seeded publication records. One ordinary fetch of seven additional, primary-verified citations returned DNS failures and cached zero of seven. Manual research below is explicitly curator-authored notes, not a provider artifact. Missing local records remain publication gates even where primary text was read externally.
The initial decision set is 27 ACCEPT, 2 MODIFY, 1 KEEP_AS_NON_CORE and 14 UNDECIDED, with zero NEW rows and one integrated transcription-factor core. The unresolved set comprises 13 source-limited protein-binding assertions and the PAINT neuron-development assignment. Source limitation is not evidence that an interaction is false. The general DNA-binding and general transcription-factor rows are refined to the established Pol II-specific activity, without changing their original machine assertions.
PMID:8756334 (DOI 10.1089/dna.1996.15.531) clones human Cart1 from a HeLa library and tests repression of a rat prolactin promoter through its AT-rich footprint II element. Human protein identity and promoter species are distinct. The accessible abstract does not identify the cotransfection host: GH3 and 10T1/2 appear in a separate expression survey and are not used to fill that gap. The original body was not recovered through publisher/search attempts. Primary identity/access route: https://pubmed.ncbi.nlm.nih.gov/8756334/.
PMID:9753625 (DOI 10.1006/bbrc.1998.9257) provides positive human sequence-selection, recombinant EMSA, nuclear localization and HeLa reporter evidence. Selected palindromes contain two TAAT sequences separated by three or four bases. Activation is site- and dose-dependent. Its abstract says the DNA-bound multimer is possibly a homodimer, which is weaker than an obligatory native dimer claim. The protected UniProt summary's more definite wording is not used to manufacture a NEW homodimer activity. The primary body was not recovered; the positive abstract assays suffice for the established binding/transcriptional function. Primary route: https://pubmed.ncbi.nlm.nih.gov/9753625/. A second reviewer independently read both original abstracts and confirmed these human-protein, reporter-host and oligomeric limits.
Activation in this setting and repression of a different promoter are compatible. The activator TAS row citing the earlier repression paper is retained because direct human activation is established independently; unrecovered full text is not grounds to label the original citation wrong. Likewise, independently verified nuclear localization supports retained source-specific nuclear assertions without inventing unseen imaging details.
PMID:38262408 (DOI 10.1016/j.cell.2023.12.032) adds endogenous human chromatin evidence. Indexed original Results/Methods at https://pmc.ncbi.nlm.nih.gov/articles/PMC10872279/ describe H9-derived cranial neural crest cells, tagged endogenous ALX1 occupancy and perturbation of developmental regulatory programs. ALX1 and ALX4 are distinguished in the experimental design. The purified structural complex contains ALX4, and the ALX4 biochemical/structural results are not presented as purified ALX1 measurements. This published Cell version supports direct human regulatory work; its earlier preprint is not separately counted. Chromatin co-occupancy and functional cooperation do not by themselves specify a fixed soluble ALX1 complex stoichiometry.
PMID:12929931 (DOI 10.1359/jbmr.2003.18.8.1419) has a substantive structured abstract on the original publisher page: https://academic.oup.com/jbmr/article/18/8/1419/7592439. Reporter, GST pull-down, yeast-two-hybrid, immunoprecipitation and acetylation experiments connect Cart1 to p300/CBP, with a conserved lysine affecting physical association and transactivation. Full construct/species/host Methods were not recovered. The exact mouse-donor experiment is therefore not recoded as direct purified-human evidence. Together with conserved human transcriptional assembly context, this supports the existing broad transcription-regulator-complex transfer while leaving its exact human composition unresolved. It does not assign p300's acetyltransferase chemistry to ALX1.
PMID:23288509 (DOI 10.1158/0008-5472.CAN-12-2377) is locally abstract-only, but the indexed original AACR Methods/Results/Figure 5 were inspected at https://aacrjournals.org/cancerres/article/73/5/1581/586425/ALX1-Induces-Snail-Expression-to-Promote. Human ALX1 expression/attenuation, SNAI1 reporter regulation and Snail epistasis provide positive regulatory participation in EMT, beyond a knockout phenotype alone. SKOV3/HEY respond differently from ES2. The study's reporter experiments are not relabeled as ALX1 ChIP, and nuclear Snail staining is not reused as an ALX1 localization assay. The transcriptional step belongs to the central mechanism; EMT is retained as a contextual non-core output. No external body passage was inserted into a field reserved for noncommittable full text.
PMID:20451171 (DOI 10.1016/j.ajhg.2010.04.002) links biallelic human ALX1 disruption to severe frontonasal dysplasia. Indexed primary Results at https://pmc.ncbi.nlm.nih.gov/articles/PMC2869009/ distinguish a multigene deletion family from an independent splice-site family. Patient RNA was unavailable: a predicted splice consequence is not a measured transcript or binding defect. This clinical evidence supports human developmental importance but does not substitute for an activity assay.
PMID:32914578 (DOI 10.15252/emmm.202012013) was inspected through indexed primary Results/Methods at https://pmc.ncbi.nlm.nih.gov/articles/PMC7539331/. Human p.L165F patient-derived neural crest cells have altered developmental state, survival and migration, with BMP manipulations; zebrafish loss and dominant-negative experiments are distinct. Neural crest cells are not automatically mature neurons, and a dominant-negative construct may interfere with related proteins. No direct human neuron-maturation term is inferred.
PMID:8673125 (DOI 10.1038/ng0796-275) is titled Prenatal folic acid treatment suppresses acrania and meroanencephaly in mice mutant for the Cart1 homeobox gene. The original Nature abstract at https://www.nature.com/articles/ng0796-275 describes mouse forebrain mesenchymal survival and cranial neural-tube closure defects. Folate rescue does not make ALX1 a folate-binding protein. PMID:35127681 (DOI 10.3389/fcell.2022.777887) provides mouse craniofacial skeletal and mesenchymal marker/survival evidence; indexed original Methods/Results are at https://pmc.ncbi.nlm.nih.gov/articles/PMC8815032/. Exon-2 deletion and the residual truncated N-terminal product are distinguished from intact Alx1. These model-organism findings support the existing skeletal-morphogenesis transfer alongside human transcriptional and disease evidence.
PMID:41670220 (DOI 10.1242/dmm.052241) establishes a temporally restricted mouse cranial mesoderm requirement for extraocular muscle development. Primary identity/abstract verified at https://pubmed.ncbi.nlm.nih.gov/41670220/ and https://pmc.ncbi.nlm.nih.gov/articles/PMC12937918/; subsequent body retrieval was challenged. It motivates a human developmental question, not an automatic NEW human myogenesis assertion. No direct promoter-binding experiment or identical human lineage requirement is claimed from this access.
PMID:16189514 (DOI 10.1038/nature04209) is an authentic human interaction-map source. Its local abstract does not expose the ALX1–IPO13 assay record. PMID:32296183 (DOI 10.1038/s41586-020-2188-x; primary https://pmc.ncbi.nlm.nih.gov/articles/PMC7169983/) is cached with a substantive full body: human ORFeome screening in several yeast-two-hybrid configurations, retesting and dataset-level orthogonal validation were read. The ALX1 pair-specific Supplementary Table 9 records and validation assignments were not recovered through the primary interactome/IntAct routes. Protected UniProt lists the same curated partners, including specific donor isoforms; this is positive curated evidence, not grounds for claiming misattribution. All 13 IPI rows remain source-specific UNDECIDED under the access rule. Generic binding is not converted into an invented adapter or enzyme activity.
PMID:28473536 (DOI 10.1126/science.aaj2239; primary https://pmc.ncbi.nlm.nih.gov/articles/PMC8009048/) is an authentic human TF methylation/SELEX source. Its canonical full_text_available: true accompanies only a partial HTML extraction; the ALX1-specific target supplement was not inspected. The derived full-text-unavailable flag remains false, while the extraction limitation is explicit. Recombinant human sequence recognition in PMID:9753625 independently supports the existing DNA-binding annotation. No ALX1 methyl-CpG preference is asserted.
Official HPA data at https://v22.proteinatlas.org/ENSG00000180318-ALX1/subcellular support nuclear-body and nucleoplasmic localization with HPA001598 in the tested human cell-line panel. The archived Golgi call is uncertain and is not added as a new location. The source method, antibody context and limited cell panel are distinct from universal tissue distribution.
The normal PANTHER/InterPro family record identifies human Q15699, mouse Q8C8B0 and rat Q63087 in PTHR24329:SF359. Its broad family prose is explicitly marked llm: true and unchecked. That prose is not a primary experiment and supplies no neuronal or developmental evidence. No family label is guessed or edited. PAINT rows are assessed as ancestral-node assertions; self-evidence in WITH/FROM is not circular, and small donor lists are not treated as weak. Exact original IBD placements were not reconstructed.
One integrated core captures sequence-specific Pol II transcriptional regulation in nuclear chromatin. Its process term GO:0006357 is the common regulatory parent of already present positive/negative Pol II annotations, not a new uncovered process claim. The core's actor is ALX1 itself binding regulatory DNA and regulating transcription; it is not merely a substrate of a developmental pathway. Existing broad chromatin and nuclear locations are retained. An accepted broad complex assertion need not manufacture a separate core or fixed complex composition.
The live GO:0005667 definition, inspected through the official MGI ontology page, permits a complex that associates with DNA directly or through another DNA-binding component and regulates transcription. The separate GO:0048666 definition concerns neuronal development toward a mature functional cell. Neural crest migration, mesenchymal patterning and closure of a neural tube do not settle that latter assertion. Its ancestral experimental basis remains an explicit question. Local GO-CAM index search returned no ALX1/Q15699 match; this is a local search result, not a claim that no curator model exists anywhere. Zero NEW rows avoids adding redundant DNA-binding parents, uncertain homodimer stoichiometry or untested human lineage processes.
The finite authored census currently contains 13 PMIDs: six original records and seven new primary sources. All DOI/PMC links above map to those same identities. No genuine provider report was generated, so there is no hidden provider citation corpus. Unused sequence/project and phosphoproteomic bibliography in immutable UniProt is not recursively promoted to a functional dependency. The seven absent normal caches are 8673125, 12929931, 20451171, 32914578, 35127681, 38262408 and 41670220. They remain explicit DRAFT gates; identity verification and external primary access do not erase cache availability requirements.
Full ordinary validation passed with two warning categories: the seven missing caches and the core's regulatory parent GO:0006357 not appearing as an exact source row. The latter is an intentional synthesis of the existing signed transcription annotations, so an additional redundant NEW row is not warranted. Source-preservation checks passed for all 44 annotation objects, 13 original reference identities, raw gene records and six existing publication bytes. All 39 ordinary supporting quotes match their own cached source after whitespace normalization with case preserved; none require case folding. Rendering and the append-only history record are checked separately. The bounded independent source consultation confirmed the two early human assays and the qualified complex-transfer interpretation; it did not certify every original donor experiment.
The seven normal publication records have now been recovered, inspected and imported without changing the six pre-existing caches. The earlier missing-cache statements document the draft stage. Three recovered records remain abstract-only: PMID:8673125, PMID:12929931 and PMID:20451171. Their complete cached abstracts confirm the identity and bounded interpretations above. The structured Cart1/p300 abstract directly reports physical association, reporter regulation and p300-dependent acetylation; its full donor construct and host details remain unavailable. The human clinical paper remains distinct from a direct DNA-binding experiment.
Recovered XML bodies for PMID:32914578, PMID:35127681, PMID:38262408 and PMID:41670220 were inspected for the relevant Results, corresponding Methods and interpretive limits. This does not claim that all sections or supplemental datasets were reviewed. The patient neural crest study measured increased susceptibility to heat-shock-induced apoptosis, without a difference in basal apoptosis. It explicitly acknowledges uncertainty about the developmental stage represented by the cultured cells. In contrast to the migration effects in the patient-cell and zebrafish experiments, the mouse exon-2 deletion study detected no overt early cranial neural crest migration defect, while leaving subtle effects open. Its demonstrated regional mesenchymal-patterning and skeletal defects support the existing developmental assertion without making a universal migration claim.
The human H9-derived cranial neural crest study confirms endogenous tagged ALX1 occupancy, loss-dependent changes in chromatin accessibility, and transcriptional changes associated with anterior versus posterior facial identity. The DNA-depleted IP-MS experiments did not detect stable ALX–TWIST interactions. Thus, DNA-associated cooperation is not recast as a stable soluble ALX1 complex. The purified structural and EMSA experiments used ALX4 domains, and remain distinct from the human cellular ALX1 evidence.
The extraocular-muscle study now has a readable body. Conditional lineage and timed mouse deletion distinguish an early cranial mesoderm requirement for muscle formation from a neural crest requirement for organization and attachment. The authors leave direct versus indirect regulation of Pitx2 for future work, and report that extraocular muscles were not specifically examined in the cited human patients. This strengthens the existing developmental question without creating a human myogenesis annotation.
Seven source findings and exact supporting quotations now accompany the recovered records, with access flags reflecting the actual abstracts or XML bodies. All 44 original assertion objects and their actions, the core terms, original reference identities, raw gene records and prior history remain preserved. The neuron-development assertion and 13 source-limited interaction rows remain UNDECIDED. No GO assertion or new functional citation was added.
Final full validation passes with the one intentional common-ancestor core advisory and no missing-cache warning. The review remains DRAFT for that advisory. Rendering and source-preservation checks pass; all 55 supporting quotations are literal cache substrings, with only whitespace reconciled in the earlier quotations. The complete authored dependency census remains 13 PMIDs.
The starting point is the exact published head f3c21940e42ad9feb81a5841616ce81609fe69f0. The three curated gene files, both existing histories, immutable seed records and all 13 PMID dependencies were copied and checked against that head before editing. The earlier sections above remain a chronological account; the adjudications in this section supersede their draft-stage action/status statements.
The 13 experimental protein-binding rows remain UNDECIDED. The explicit source-access rule applies when the relevant experimental evidence cannot be adjudicated, including the standing exception in the annotation-reviewer skill. Generic binding is uninformative, but that alone does not resolve whether a source supports a more specific ALX1 function. The HuRI body PMID:32296183 describes systematic ORFeome/Y2H screening, pairwise retesting and dataset-level orthogonal validation. Its locally cached text contains no ALX1, CART1 or Q15699 record; it points to Supplementary Table 9 for the interaction set, while Table 8 addresses validation. A body labeled full text therefore does not expose the ALX1 pair entries or their individual validation assignments. The 2005 IPO13 source PMID:16189514 remains abstract-only. No further retrieval success or pair-level validation is claimed here.
The revised reasons identify the actual curated partner and retain explicit partner-isoform scopes. Biological plausibility differs, but does not replace the missing pair evidence. The immutable RARA UniProt record describes its nuclear receptor/transcriptional context. The official KAT5 UniProt entry, inspected for identity/function/location, describes a nuclear chromatin acetyltransferase. These make nuclear overlap with ALX1 plausible without transferring either partner's activity to ALX1. In contrast, the official OR52L1 UniProt entry identifies a multipass cell-membrane receptor; it also preserves the curated ALX1 interaction. The localization mismatch motivates physiological and construct-level checks, not a wrong-interaction verdict. No additional biological claims about uninspected partner experiments were introduced. All partner names and isoform identifiers already occur in the ALX1 source record.
The direct human localization evidence was re-read on the archived HPA v22 ALX1 subcellular page. With antibody HPA001598, the page lists HEK293, U-251MG and U2OS and reports both nuclear-body and nucleoplasmic staining. Its short exact classification excerpts are:
Nuclear bodies (supported)
Nucleoplasm (supported), Golgi apparatus (uncertain)
These excerpts are cited through this local notes file solely to preserve attribution and make the evidence checkable; the notes are not an independent experiment. Golgi localization remains outside the accepted claims. The nucleus-only hCART1 quote was removed from the two more specific location rows. For nuclear body, the exact abstract statement about Cart1 recruiting p300 to a subnuclear compartment PMID:12929931 is added as bounded corroboration. The accessible abstract does not identify the species of its Cart1 construct or the host for that imaging; neither mouse nor human construct identity is assumed. The original publisher full-text route remained inaccessible during this follow-up. No complete-paper or supplementary-material read is claimed.
The existing GO:0048704 embryonic skeletal system morphogenesis assertion is now included in the single transcription-factor core's directly_involved_in, with no new annotation or duplicated molecular-function core. The official GO definition concerns generation and organization of embryonic skeletal structures; its MGI parent view places it under embryonic organ and skeletal-system morphogenesis and within embryonic skeletal development. This is synthesis of an existing accepted assertion, not an inference from other genes having a missing term. The actor is ALX1 performing regulatory work: endogenous ALX1 ChIP/CUT&RUN occupancy and ALX1-dependent chromatin/transcriptional changes in human H9-derived cranial neural crest cells PMID:38262408 connect its TF activity to regional facial programs. The relevant Results paragraphs and associated perturbation design were re-read. Long-term ALX1 depletion altered accessibility at 6,195 peaks; ALX1/ALX4 perturbations shifted anterior/posterior regulatory expression. The mouse skeletal defects PMID:35127681 and human disease PMID:20451171 anchor developmental consequences. These phenotypes alone are not the participation argument. Long-term experiments can contain indirect effects, acute ALX depletion had much weaker effects, combined ALX1/ALX4 results are not all ALX1-specific, and ALX4 purified structures remain excluded from ALX1 biochemistry. No additional process is proposed.
The broad InterPro homeodomain GO:0003677 DNA binding row changes from MODIFY to ACCEPT. The mapping is true at its intended level and agrees with direct human binding assays; the separate GO:0000981 row and core already provide the more specific transcription-factor activity. This is the only action change. The sole remaining MODIFY is the general TF activity refined to Pol II-specific TF activity. Workflow route strings already documented above were removed from the touched reasons. Evidence-access and PAINT-node limits remain explicit where they affect interpretation.
A suggested question now asks which Cart1 constructs support the p300/CBP association and whether matched human assays establish a more specific cofactor-binding function. The reviewer's example GO:0003713 is transcription coactivator activity, not coactivator binding, as confirmed in the repository ontology cache and official ontology record. It was not added: p300 coactivation/acetylation does not make ALX1 the coactivator or acetyltransferase. The PAINT neuron-development row remains UNDECIDED; a different donor list or an unverified node-depth argument cannot establish target loss or incorrect IBD placement.
The review is now COMPLETE because all 44 source assertions have been adjudicated and all required caches are present. The known GO:0006357 common-parent advisory is intentional and does not call for a redundant NEW row. COMPLETE does not claim that the 14 source-specific uncertainties have disappeared. The final action census is 28 ACCEPT, 1 MODIFY, 1 KEEP_AS_NON_CORE, 14 UNDECIDED, 0 NEW, with one integrated core. The finite citation census remains 13 PMIDs and seven GO_REF methods, plus this attributed local provenance reference. Raw GOA/UniProt, all cached publications and both earlier histories remain unchanged.
Verification for this follow-up: full ordinary validation passes with only the intentional GO:0006357 common-parent advisory; rendering and new-history validation pass. All 61 supporting-text occurrences are literal substrings of their cited publication or attributed local notes. Integrity checks preserve all 44 original source objects, all 20 original reference objects, the 13 publication caches, both raw seed files and both prior histories.