Siderius Type X-Linked Intellectual Disability

1. Disease Information

1.1 Overview

Siderius type syndromic X‑linked intellectual developmental disorder (MRXSSD), increasingly called PHF8‑XLID in the recent literature, is a rare X‑linked recessive chromatinopathy caused by loss‑of‑function variants in PHF8 at Xp11.22. PHF8 encodes a JmjC‑domain, Fe(II)/2‑oxoglutarate‑dependent histone lysine demethylase (KDM7B). The classical clinical description — derived from the founding families — is mild-to-borderline intellectual disability with cleft lip and/or cleft palate and mild facial dysmorphism in affected males. The 2022 multicenter expansion (n = 16 new affected males) substantially reframed the phenotype: developmental delay is universal, ID is near-universal, autism spectrum disorder and ADHD are frequent and were previously under-recognized, and orofacial clefting is much less common than the founding-cohort literature implied (3/16, 19%).

Key framing quote (Sobering et al. 2022, HGG Adv, PMID:35469323) ✅:

"Loss-of-function variants in PHD Finger Protein 8 (PHF8) cause Siderius X-linked intellectual disability (ID) syndrome, hereafter called PHF8-XLID. PHF8 is a histone demethylase that is important for epigenetic regulation of gene expression. PHF8-XLID is an under-characterized disorder with only five previous reports describing different PHF8 predicted loss-of-function variants in eight individuals."

1.2 Identifiers

1.3 Synonyms

• MRXSSD (OMIM preferred abbreviation)
• Siderius X-linked mental retardation syndrome
• Siderius–Hamel syndrome / Siderius‑Hamel cleft lip‑palate syndrome
• Mental retardation, X-linked, syndromic, Siderius type (legacy OMIM title)
• PHF8-XLID (preferred in Sobering et al. 2022 and increasingly the working name)
• X-linked mental retardation with cleft lip/palate (descriptive, historical)
• Gene aliases used as disease-adjacent labels: JHDM1F, KDM7B, ZNF422

1.4 Data provenance character

Information is aggregated disease-level and case-series derived, not EHR-derived. The entire literature base is ~29 affected individuals across ~15 families in 6 primary reports plus contiguous-deletion cases. There is no registry, no natural-history study, no EHR cohort, and no clinical trial for this disorder. Any prevalence, frequency, or prognostic statement in this report is therefore case-series-derived and subject to strong ascertainment bias (see §3.4).

2. Etiology

2.1 Primary causal factor

Monogenic, germline, X-linked recessive loss of PHF8 function. No environmental, infectious, or multifactorial etiology is established. The mechanism of pathogenicity is loss of function; both truncating variants and catalytically inactivating missense variants have been shown to abolish demethylase activity.

Laumonnier et al. 2005 (J Med Genet, PMID:16199551) ✅ — the founding molecular report:

"Truncating mutations were found in the PHF8 gene (encoding the PHD finger protein 8) in two unrelated families with X linked mental retardation (XLMR) associated with cleft lip/palate (MIM 300263)." "The association of XLMR and cleft lip/palate in these patients with mutations in PHF8 suggests an important function of PHF8 in midline formation and in the development of cognitive abilities, and links this gene to XLMR associated with cleft lip/palate."

Original linkage (Siderius et al. 1999, Am J Med Genet, PMID:10398231) ✅:

"A family is described in which X-linked mild to borderline mental retardation (MR) is associated with cleft lip/palate. Linkage analysis showed a maximum LOD score of Z=2.78 at straight theta=0.0 for the DXS441 locus with flanking markers DXS337 and DXS990, defining the region Xp11.3-q21.3 with a linkage interval of 25 cM."

2.2 Genetic risk factors

• Causal: hemizygous PHF8 LoF variant in a male (see §4 for the variant catalogue).
• Carrier status: heterozygous mother. In Sobering 2022, 10/12 tested mothers were unaffected carriers; 2 probands' mothers carried no variant (de novo) ⚠.
• De novo occurrence: 3 of 11 LoF families in Sobering 2022 were de novo (c.596+1G>A, c.1627-1G>A, c.1965_1966dup) ⚠ — clinically important, since absence of family history does not exclude the diagnosis.
• Contiguous gene deletions at Xp11.22 encompassing PHF8 ± FAM120C ± WNK3 produce an overlapping/expanded phenotype (Qiao et al. 2008, PMID:18498374, two brothers with ASD; De Wolf et al. 2014, PMID:25258334, "The deletion of PHF8 most likely explains the cleft palate and mild intellectual disability" ⚠; Huang et al. 2020, PMID:32219840, prenatal CMA detection of Xp11.22 deletion in a fetus with cleft lip and palate ⚠).
• Modifier genes: none identified. A pathway-level candidate framework exists: PHF8 physically and functionally interacts with two other XLID proteins, ZNF711 and KDM5C/JARID1C (Kleine-Kohlbrecher et al. 2010, PMID:20346720) ✅ — "our results functionally link the XLMR gene PHF8 to two other XLMR genes, ZNF711 and JARID1C, indicating that MR genes may be functionally linked in pathways, causing the complex phenotypes observed in patients developing MR." Poeta et al. 2019 (PMID:31691806) place KDM5C at the crossroads of ARX, ZNF711 and PHF8 transcriptional axes ⚠. Whether variation in these partners modifies PHF8-XLID severity is untested.
• X-inactivation as a female modifier: skewed XCI tested in 3 carrier mothers — 2 completely skewed, 1 uninformative ⚠ (Sobering 2022). This is the presumed explanation for carrier females being unaffected.

2.3 Environmental risk factors

None established for this disorder. There is one mechanistically motivated hypothesis worth curating as a knowledge gap, not as fact: because PHF8 is a 2‑oxoglutarate/Fe(II)/O₂-dependent oxygenase, its catalytic output is oxygen-dependent, and maternal hypoxia is an established modifier of cleft lip/palate risk in mice.

Loenarz et al. 2010 (Hum Mol Genet, PMID:19843542) ✅ (abstract):

"The dependence of PHF8 activity on oxygen availability is interesting because the occurrence of fetal cleft lip has been demonstrated to increase with maternal hypoxia in mouse studies. Cleft lip and other congenital anomalies are also linked indirectly to maternal hypoxia in humans, including from maternal smoking and maternal anti-hypertensive treatment."

and from the cached full text ✅:

"Episodes of maternal respiratory hypoxia in mice also correlate with increased incidence of cleft lip/palate (gestational day 10-11, 10% O2), while hyperoxia rescued mouse strains that are genetically susceptible for cleft lip/palate (gestational day 10-11, 50% O2) (5), suggesting that gestational oxygen levels mediate genetic and environmental factors."

Curation guidance: this is a plausible gene–environment interaction (hypomorphic PHF8 × gestational hypoxia → clefting penetrance) but has never been tested in PHF8-mutant humans or mice. Model it as a discussions entry with kind: KNOWLEDGE_GAP (or as a mechanistic_hypotheses group with status: EMERGING), not as an etiologic claim.

2.4 Protective factors

• Genetic: none identified. No protective alleles, no gnomAD LoF-tolerant signal reported.
• Environmental: none disease-specific. Periconceptional folic acid supplementation is protective for orofacial clefting in the general population but has not been studied in PHF8-related clefting — do not import that claim into this entry.
• A counterintuitive "protective" finding in mice: Phf8-null mice show resilience to stress-induced anxiety/depression-like behavior (Walsh et al. 2017, PMID:28485378) ⚠ — "we report a striking resiliency to stress-induced anxiety- and depression-like behaviour on loss of Phf8." This has no demonstrated human counterpart and should be curated as evidence_source: MODEL_ORGANISM with an explicit HUMAN_MODEL_MISMATCH discussion (see §15.4).

3. Phenotypes

3.1 Cohort-derived frequency table (Sobering et al. 2022, n = 16 affected males with LoF variants) ⚠

All frequencies below are from the PMC full text of PMID:35469323 and must be re-verified against the paper before entering as frequency: with evidence (see docs/frequency-evidence-guidelines.md — a frequency band needs its own quantitative snippet).

3.2 HPO disease annotations currently attached to OMIM:300263

Retrieved from the HPO annotation API (ontology.jax.org) ⚠ — note these derive from the older, clefting-ascertained families, which is why cleft frequencies are far higher than in Sobering:

Additional features named in OMIM/GTR/MalaCards summaries but not in the HPO table above ⚠: cryptorchidism, preaxial polydactyly, broad nasal bridge. Treat these as low-confidence single-case observations.

3.3 Neuroimaging phenotypes (6 individuals imaged, Sobering 2022) ⚠

• 1 normal
• 1 mildly increased subarachnoid space + white matter changes
• 2 (identical twins) polymicrogyria and cortical dysplasia → HP:0002126 Polymicrogyria ⚠
• 1 abnormal striatal signal (caudate/globus pallidus)
• 1 cranio-occipital malformation + thin corpus callosum → HP:0033725 / HP:0002079 ⚠

The twin polymicrogyria observation is mechanistically interesting given the 2026 finding that PHF8 loss arrests neurogenesis in mouse embryos (§6.6), but n = 2 in one family — curate as OCCASIONAL at most.

3.4 The clefting discordance — flag this explicitly in the KB entry

The single most important curation nuance for this disease: earlier cohorts were ascertained because of clefting (Koivisto et al. 2007 screened 7,712 cleft-surgery patients), so cleft frequency in the pre-2022 literature (~70–75%) is inflated by design. Sobering et al. found 3/16 (19%) and concluded ⚠:

"Orofacial clefting was seen in three individuals from our cohort, suggesting that this feature is less common than previously reported."

Recommendation for the dismech entry: annotate cleft lip/palate as a characteristic but not obligate feature (frequency: OCCASIONAL per Sobering) with a notes: field recording the ascertainment-bias discordance, and cite both frequency sources.

3.5 Phenotype characteristics

• Onset: clefting is congenital (HP:0003577 Congenital onset ⚠); developmental delay is apparent in infancy/early childhood (HP:0011463 Childhood onset ⚠); facial gestalt evolves with age (elongated face "tends to worsen with age" ⚠).
• Severity: classically mild-to-borderline ID; Sobering broadened this to borderline through severe, with two individuals having no ID (one with dyscalculia, one with mild learning difficulties) ✅ — "All affected individuals exhibited developmental delay and all but two had borderline to severe ID."
• Progression: static / non-progressive neurodevelopmental disorder. No neurodegeneration, no regression, and no progressive organ involvement has been reported in any of the ~29 published individuals. Use clinical_course: STABLE for the ID node.
• Frequency among affected individuals: see tables above.

3.6 Quality-of-life impact

❌ No disease-specific QoL data exist (no EQ-5D, SF-36, PROMIS, or disease-specific instrument has been applied). Inferable, non-citable impacts: educational support needs (universal DD), speech intelligibility (hypernasal speech + clefting), feeding difficulty in infancy (63%), behavioral/social burden of ASD+ADHD (~44% each), and seizure burden (31%). Do not assert QoL numbers.

4. Genetic / Molecular Information

4.1 Gene

PHF8 (PHD finger protein 8), Xp11.22, NCBI Gene 23133, HGNC:20672, OMIM 300560. Aliases: JHDM1F, KDM7B, ZNF422, KIAA1111, MRXSSD.

Protein: UniProt Q9UPP1, canonical isoform 1,060 aa (Sobering et al. describe the disease-relevant isoform as 1,024 aa; multiple transcript variants exist — note the discrepancy when curating). Domains: N-terminal PHD finger (H3K4me3 reader), central JmjC catalytic domain (aa 231–387 per UniProt), multiple nuclear localization signals, and a serine-rich region. Subcellular localization: nucleus and nucleolus; "recruited to H3K4me3 sites on chromatin during interphase" and dissociates during mitosis ⚠ (UniProt).

4.2 Reported pathogenic variants

Previously reported (the "five previous reports," 8 individuals) — as tabulated in Sobering et al. Table 1 ⚠:

New in Sobering et al. 2022 (11 LoF variants / 11 families, 16 males) ⚠:

VUS reported (5 individuals / 4 families, all maternally inherited missense) ⚠: c.143A>G p.(Tyr48Cys); c.257C>T p.(Thr86Met); c.808C>T p.(Arg270Cys); c.1150G>A p.(Glu384Lys). Sobering explicitly separates these from the LoF cohort — do not curate them as pathogenic.

A benign in-frame variant to know about: p.Ser969del — "it is abundantly found in gnomAD in both heterozygous males and homozygous females" ⚠ (Sobering 2022). Useful as a negative control / interpretation caveat.

4.3 Variant classification, spectrum, and origin

• Origin: germline, exclusively. No somatic PHF8 disease variants relevant to this disorder (somatic/overexpression PHF8 biology is a cancer story — see §6.9 — and must be kept out of the disease mechanism graph).
• Types: nonsense, frameshift, canonical splice-site, intragenic multi-exon deletions, contiguous Xp11.22 deletions, and one recurrent catalytically inactivating missense (F279S). The older literature emphasized that mutations cluster in the JmjC-encoding exons; the 2022 cohort shows LoF variants distributed across the gene, including C-terminal frameshifts.
• Functional consequence: loss of function — via (a) truncation removing the JmjC domain and NLSs, (b) catalytic inactivation, and (c) mislocalization.

Loenarz et al. ✅ on both loss-of-catalysis and mislocalization:

"Clinically observed mutations to the PHF8 gene cluster in exons encoding for the double stranded beta-helix fold and will therefore disrupt catalytic activity." "This mutant encodes a F279S variant of PHF8 that modifies a conserved hydrophobic region; assays with both peptides and intact histones reveal this variant to be catalytically inactive." (full text) "In contrast to the wild-type HA-PHF8, analogous studies revealed that the clinically observed F279S variant did not show clear nuclear localization, with apparent cytoplasmic localization."

Abidi et al. 2007 ✅ on the truncation mechanism:

"The mutation results in a truncated PHF8 protein lacking the Jumonji-like C terminus domain and five nuclear localization signals."

Koivisto et al. 2007 ✅ on F279S:

"A novel missense mutation c.836C>T of the PHF8 gene was identified in a Finnish family with multiple-affected male patients. The mutation resides in exon 8 and changes phenylalanine to serine (F279S) in the functionally important Jmonji C domain of the protein." "The mutation was not present in 200 anonymous blood donors (approximately 300 X-chromosomes)."

Independent confirmation of F279S catalytic death: Qi et al. 2010 ✅ ("a point mutation in the catalytic domain (phenylalanine to serine, F279S) abolished PHF8 demethylase activities"); Qiu et al. 2010 ✅ ("a mutant PHF8 (phenylalanine at position 279 to serine) identified in the XLMR patients is defective in enzymatic activity, indicating that the loss of histone demethylase activity is causally linked with the onset of disease"); Fortschegger et al. 2010 ✅ ("a PHF8 disease mutant was defective in demethylation and in coactivation").

4.4 Population / database data

• ClinVar: 537 total records for PHF8[gene]; 251 returned by a pathogenic[clinical significance] filter ⚠ — this filter over-counts (it captures likely-pathogenic and conflicting interpretations); do not quote 251 as "pathogenic variants." Get the exact breakdown from ClinVar Miner or a fresh ClinVar query before curating.
• gnomAD constraint (pLI / LOEUF / o-e LoF): ❌ not retrievable this session (gnomAD's GraphQL endpoint requires POST; DECIPHER blocked). This must be filled in manually before curating any constraint claim.
• Allele frequency of pathogenic variants: all reported pathogenic variants are private/family-specific. No founder allele is known. F279S absent from ~300 control X chromosomes (Koivisto) ✅.
• ClinGen Gene-Disease Validity: reported as Definitive for PHF8 – syndromic X-linked intellectual disability Siderius type, X-linked, supported by ≥7 variants from 5 publications plus mouse, zebrafish and cell-culture models with rescue; mechanism = loss of function ⚠. Recommended action: rather than citing a web page, run just clingen-refresh and cite the CGGV: assertion as a structured-source evidence row.

4.5 Epigenetic information

Two distinct senses matter here and should not be conflated in the KB:

1. PHF8 is itself an epigenetic effector — its loss changes the histone-methylation landscape (see §6). This is the disease mechanism.
2. Episignature (methylation signature) for diagnostic classification: ❌ No published DNA-methylation episignature for PHF8-XLID was found. Many chromatinopathies (Kabuki, Sotos, CHARGE, etc.) have EpiSign classifiers; PHF8 does not appear among them in the retrieved literature. This is a genuine, curatable knowledge gap and a natural proposed_experiments item.

4.6 Chromosomal abnormalities

Xp11.22 microdeletions removing PHF8 (± FAM120C, WNK3) — Qiao 2008 (PMID:18498374, familial deletion in two brothers with ASD; "complete deletion of the plant homeodomain finger protein 8 (PHF8) gene" ⚠), De Wolf 2014 (PMID:25258334, syndromic autism ⚠), Huang 2020 (PMID:32219840, prenatal detection in a fetus with cleft lip and palate ⚠). These support CMA as a diagnostic modality and support haploinsufficiency/nullisomy as the mechanism.

5. Environmental Information

• Environmental factors: none established. The only mechanistically grounded candidate is gestational hypoxia acting on an O₂-dependent 2‑OG oxygenase (§2.3) — hypothesis only.
• Lifestyle factors: none disease-specific. Maternal smoking is cited by Loenarz et al. only as a general hypoxia-mediated cleft risk factor in the population, not as a PHF8 interaction.
• Infectious agents: not applicable.
• Toxicological note (model-organism only, do not curate as human etiology): chronic exposure to the PFAS replacement GenX induced transgenerational motor deficits via the C. elegans PHF8 ortholog jmjd-1.2 (PMID:40803444, Environ Pollut 2025) ⚠. Interesting as a conservation-of-mechanism datapoint (§14), not as a human risk factor.

6. Mechanism / Pathophysiology

6.1 The core enzymatic lesion (MOLECULAR scale)

PHF8 is a PHD-finger reader + JmjC-domain eraser: it binds H3K4me3 at active promoters through its PHD finger and removes repressive mono-/di-methyl marks through its Fe(II)/2‑OG-dependent JmjC domain, thereby acting as a transcriptional coactivator.

Loenarz et al. 2010 ✅:

"We report that recombinant PHF8 is an Fe(II) and 2-oxoglutarate-dependent N(epsilon)-methyl lysine demethylase, which acts on histone substrates. PHF8 is selective in vitro for N(epsilon)-di- and mono-methylated lysine residues and does not accept trimethyl substrates." (full text) "our results reveal that PHF8 is a 2OG oxygenase with selectivity for H3K9me2/me1, H3K27me2 and H3K36me2 residues."

Kleine-Kohlbrecher et al. 2010 ✅:

"the XLMR protein PHF8 and a C. elegans homolog F29B9.2 catalyze demethylation of di- and monomethylated lysine 9 of histone H3 (H3K9me2/me1). The PHD domain of PHF8 binds to H3K4me3 and colocalizes with H3K4me3 at transcription initiation sites."

Qi et al. 2010 (Nature) ✅ — the H4K20me1 activity and the genome-wide picture:

"Here we provide multiple lines of evidence establishing PHF8 as the first mono-methyl histone H4 lysine 20 (H4K20me1) demethylase, with additional activities towards histone H3K9me1 and me2." "PHF8 is located around the transcription start sites (TSS) of approximately 7,000 RefSeq genes and in gene bodies and intergenic regions (non-TSS). PHF8 depletion resulted in upregulation of H4K20me1 and H3K9me1 at the TSS and H3K9me2 in the non-TSS sites, respectively, demonstrating differential substrate specificities at different target locations." "Importantly, patient mutations significantly compromised PHF8 catalytic function."

Fortschegger et al. 2010 ✅ — the RNAPII coupling:

"Chromatin immunoprecipitation followed by high-throughput sequencing indicated that PHF8 is enriched at the transcription start sites of many active or poised genes, mirroring the presence of RNA polymerase II (RNAPII) and of H3K4me3-bearing nucleosomes... we present evidence for direct interaction of PHF8 with the C-terminal domain of RNAPII."

Feng et al. 2010 (PMID:20208542) adds the rDNA/nucleolar arm: PHF8 "activates transcription of rRNA genes through H3K4me3 binding and H3K9me1/2 demethylation" ⚠.

Suggested GO annotations (verify all with OAK before use):

CHEBI cofactors/substrates: 2-oxoglutarate (CHEBI:16810 ⚠), iron(2+) (CHEBI:29033 ⚠), dioxygen (CHEBI:15379 ⚠), L-serine (CHEBI:17115 ⚠), sirolimus/rapamycin (CHEBI:9168 ✅ verified).

6.2 Downstream arm 1 — neuronal/cognitive: the mTOR–RSK1 axis

Chen et al. 2018 (Nat Commun, PMID:29317619) ✅ is the strongest mechanistic chain from gene to cognition, and it is druggable:

"Here we report that Phf8 knockout mice displayed impaired learning and memory, and impaired hippocampal long-term potentiation (LTP) without gross morphological defects. We also show that mTOR signaling pathway is hyperactive in hippocampus in Phf8 knockout mouse. Mechanistically, we show that demethylation of H4K20me1 by Phf8 results in transcriptional suppression of RSK1 and homeostasis of mTOR signaling. Pharmacological suppression of mTOR signaling with rapamycin in Phf8 knockout mice recovers the weakened LTP and cognitive deficits."

Causal chain: PHF8 loss → failure to demethylate H4K20me1 at RSK1 → RSK1 de-repression → mTOR hyperactivation in hippocampus → impaired LTP → learning/memory deficit (rescued by rapamycin). Note the paper places PHF8 in the same "mTORopathy-adjacent" company as tuberous sclerosis, Fragile X and Down syndrome ✅ (intro).

6.3 Downstream arm 2 — neuronal differentiation and cytoskeleton

• RAR coactivation / neuronal differentiation — Qiu et al. 2010 (Cell Res, PMID:20548336) ✅: "knockdown of PHF8 in mouse embryonic carcinoma P19 cells impairs RA-induced neuronal differentiation, whereas overexpression of the wild-type but not the F279S mutant PHF8 drives P19 cells toward neuronal differentiation. Furthermore, we show that PHF8 interacts with RARalpha and functions as a coactivator for RARalpha."
• Cytoskeleton / neurite outgrowth — Asensio-Juan et al. 2012 (NAR, PMID:22850744) ✅: "PHF8 controls the expression of genes involved in cell adhesion and cytoskeleton organization such as RhoA, Rac1 and GSK3β... Further analysis in neurons shows that depletion of PHF8 results in down-regulation of cytoskeleton genes and leads to a deficient neurite outgrowth. Overall, our results suggest that the mental retardation phenotype associated with loss of function of PHF8 could be due to abnormal neuronal connections as a result of alterations in cytoskeleton function."
• REST/NRSF co-occupancy at neuronal gene promoters — Wang et al. 2014 (Sci Rep, PMID:24852203) ✅ (title/abstract: "PHF8 and REST/NRSF co-occupy gene promoters to regulate proximal gene expression").

6.4 Downstream arm 3 — glial biology (newer, and under-appreciated)

• Astrocytes — Iacobucci et al. 2021 (Development, PMID:34081130) ✅: "we investigate the contribution of the XLID-associated histone demethylase PHF8 to astrocyte differentiation and function. Using genome-wide analyses and biochemical assays in mouse astrocytic cultures, we reveal a regulatory crosstalk between PHF8 and the Notch signaling [pathway]…"
• Oligodendrocytes — Kremp et al. 2024 (Glia, PMID:38613395) ✅: "Phf8 promotes the proliferation of rodent oligodendrocyte progenitor cells and impairs their differentiation to oligodendrocytes... Phf8 has a strong positive impact on Olig2 expression by acting on several regulatory regions of the gene and changing their histone modification profile... We conclude that Phf8 may impact nervous system development at least in part through its action in oligodendroglial cells." Companion review: Kremp & Wegner 2026, Neural Regen Res (PMID:40145966) ⚠.

This gives a non-neuronal (glial) contribution to the neurodevelopmental phenotype — worth its own pathophysiology node with CL:0002453 (oligodendrocyte precursor cell ✅ verified) and CL:0000127 (astrocyte ⚠).

6.5 Downstream arm 4 — craniofacial / midline development

Qi et al. 2010 ✅ provides the only direct in-vivo craniofacial mechanism:

"PHF8 regulates cell survival in the zebrafish brain and jaw development, thus providing a potentially relevant biological context for understanding the clinical symptoms associated with PHF8 patients. Lastly, genetic and molecular evidence supports a model whereby PHF8 regulates zebrafish neuronal cell survival and jaw development in part by directly regulating the expression of the homeodomain transcription factor MSX1/MSXB, which functions downstream of multiple signalling and developmental pathways. Our findings indicate that an imbalance of histone methylation dynamics has a critical role in XLMR."

MSX1 is a well-established human orofacial-clefting gene, which makes the PHF8→MSX1 link the most credible molecular explanation for the clefting phenotype. Kremp et al. 2024 ✅ independently characterize Phf8 as "implicated by mutation in mice and humans in neural crest defects and neurodevelopmental disturbances." Bone-forming relevance: Han et al. 2015 (PMID:25923143) ⚠ — "PHF8, a major H4K20/H3K9 demethylase, plays a critical role in craniofacial and bone development."

Causal chain: PHF8 loss → H4K20me1/H3K9me1-2 accumulation at MSX1/msxb → reduced MSX1 expression in cranial neural crest-derived facial primordia → impaired fusion of facial prominences / palatal shelves → cleft lip ± cleft palate, in parallel with increased apoptosis of brain cells.

6.6 Downstream arm 5 — metabolic control of neurogenesis (2026, newest)

Artés/Iacobucci et al. 2026 (EMBO Rep, PMID:41714361) ⚠ identifies "PHF8 as a key driver of the serine biosynthesis pathway, safeguarding the intracellular serine pool essential for neural progenitor proliferation," with PHF8 depletion causing disrupted metabolism, blocked autophagy, replication defects and proliferation arrest, and PHF8 deficiency halting neurogenesis and brain development in mouse embryos. This is the newest and arguably most important mechanistic advance for this disease — it supplies a proliferation/metabolism node upstream of the cognitive phenotype and is consistent with the polymicrogyria seen in the twins. Verify the abstract before quoting.

6.7 Downstream arm 6 — cell cycle, DNA damage, transcription recovery

• Cell cycle: PHF8 controls the G1–S transition with E2F1/HCF-1/SET1A via H4K20me1 demethylation (Liu et al. 2010, Nature, PMID:20622854) ⚠.
• DNA damage: Kim et al. 2024 (NAR, PMID:39087553) ⚠ — PHF8 is "the major demethylase that reverses transcriptionally repressive epigenetic modification laid down by the DYRK1B-EHMT2 pathway," concentrating at damage tracks and promoting "timely resolution of local H3K9me2 to facilitate the resumption of transcription," including at rDNA. Relevance to the neurodevelopmental phenotype is unestablished — curate as mechanistic context, not as disease mechanism.

6.8 Immune / metabolic / tissue-damage dimensions

• Immune: no immunodeficiency or autoimmunity in patients. A 2026 paper (PMID:41709745) ⚠ reports "the Znf711-Phf8 complex functions as a transcriptional rheostat essential for neutrophil development" in a model system — no human hematologic phenotype has been reported in PHF8-XLID, so this is a prediction to watch, not a curatable human phenotype.
• Metabolic: the serine-biosynthesis finding (§6.6) is the only metabolic axis. No metabolic decompensation phenotype; do not conform this entry to metabolic_intoxication_decompensation.
• Tissue damage: none — this is a developmental, not a degenerative or inflammatory, disorder. No fibrosis, ischemia, oxidative-injury or necrosis mechanism applies.

6.9 ⚠ Scope guardrail: the PHF8 cancer literature

A large and growing body of work (≥20 papers in 2025–2026 alone: TNBC, gastric, prostate, hepatocellular, colorectal) treats PHF8 as an overexpressed oncogenic driver and drug target. This is the opposite direction of effect from the disease (LoF), involves somatic/expression biology rather than germline lesions, and must not be imported into this disease entry's pathophysiology graph. If any of it is curated at all, it belongs in a separate context with an explicit note that the disease is loss-of-function.

6.10 Suggested pathophysiology node chain for the dismech entry

1. PHF8 Loss-of-Function Variant (MOLECULAR)
     ↓ CAUSES
2. Loss of JmjC Histone Demethylase Activity (MOLECULAR)
     [H3K9me1/2, H4K20me1, H3K27me2 fail to be erased; PHD-H3K4me3 reading uncoupled from erasure]
     ↓ CAUSES
3. Failure of PHF8-Dependent Transcriptional Coactivation at Target Promoters (MOLECULAR)
     ↓ branches into:
   3a. RSK1 de-repression → mTOR hyperactivation (CELLULAR)  → impaired hippocampal LTP (CELLULAR/TISSUE) → ID, learning/memory deficit (ORGANISM)
   3b. Impaired serine biosynthesis / autophagy block in neural progenitors (CELLULAR) → arrested neurogenesis (TISSUE) → ID ± cortical malformation (ORGANISM)
   3c. Impaired RAR-dependent neuronal differentiation + RhoA/Rac1/GSK3β cytoskeletal gene loss (CELLULAR) → deficient neurite outgrowth/connectivity (CELLULAR) → ID/ASD (ORGANISM)
   3d. Impaired Olig2 induction in OPCs / Notch-crosstalk in astrocytes (CELLULAR) → altered glial development (TISSUE) → contributory NDD (ORGANISM)
   3e. Reduced MSX1/msxb in cranial neural crest derivatives (MOLECULAR/CELLULAR) → failed fusion of facial prominences and palatal shelves (TISSUE) → cleft lip/palate (ORGANISM)

Treatment link for the drug-target pattern: an mTOR-inhibitor node (target_mechanisms with INHIBITS on node 3a) is the natural place to hang rapamycin — but flag it as preclinical-only (mouse, no human data).

Module conformance assessment: no existing kb/modules/ module is a clean fit. pharyngeal_arch_patterning_serial_homology is tempting for the craniofacial arm but is the wrong mechanism — Siderius clefting is a midline fusion failure (lip/palate), not a serially homologous arch-derivative bundle (mandible+maxilla+zygoma+ear), and the lesion is chromatin-level, not an arch-identity code. Recommend no conforms_to rather than a forced fit. A future "chromatinopathy / histone-modifier neurodevelopmental disorder" module would be the right home; note this as a module-creation candidate.

7. Anatomical Structures Affected

7.1 Organ level

Body systems: nervous (primary), craniofacial/musculoskeletal (secondary), and — in a minority — genitourinary. No cardiac, renal, hepatic, hematologic, immune, endocrine or ophthalmologic involvement is reported.

Lateralization: clefting may be unilateral or bilateral (Koivisto's two siblings differed: one unilateral CL/P, one bilateral CL/P ✅) — a nice intrafamilial-variability datapoint. Facial asymmetry is noted as "mild" in Sobering ⚠.

7.2 Tissue and cell level

7.3 Subcellular level

• Nucleus (GO:0005634 ⚠) — primary site of PHF8 action; chromatin (GO:0000785 ⚠); nucleolus (GO:0005730 ⚠) for the rDNA/RNA Pol I arm.
• Pathogenic relevance: F279S and truncating variants that delete the NLSs cause cytoplasmic mislocalization — the subcellular compartment shift is itself part of the molecular pathology ✅ (Loenarz full text).

8. Temporal Development

• Onset:
• Prenatal/congenital: orofacial clefting (detectable on prenatal ultrasound/CMA — Huang 2020 ⚠); brain malformations where present.
• Infancy: feeding difficulties (63%), hypotonia/motor delay; mean age at independent walking 20 months ⚠.
• Early childhood: speech delay (100%), then ID; ASD/ADHD typically recognized in preschool/school years.
• Onset pattern: congenital/insidious developmental, not acute.
• Progression: static (non-progressive). There is no reported regression, neurodegeneration, or organ deterioration. The facial gestalt evolves (long face becomes more marked with age ⚠) — that is a morphologic evolution, not disease progression.
• Course: chronic, lifelong.
• Stages: no staging system exists or is applicable.
• Remission: not applicable.
• Critical periods:
• Weeks 4–10 of gestation — lip and palatal fusion; the only window in which the clefting phenotype could theoretically be modified (and the window in which the hypoxia hypothesis would operate).
• Perinatal-to-early-childhood — the intervention window for cleft repair, feeding support, and early developmental/speech intervention, where outcome is genuinely modifiable.
• Preclinically, the Chen 2018 rapamycin rescue was performed in adult mice and restored LTP and cognition — implying, in mice, a post-developmental window of reversibility. That is a striking and citable claim, but strictly MODEL_ORGANISM evidence.

9. Inheritance and Population

9.1 Epidemiology

• Prevalence: ❌ no formal estimate published. Orphanet classifies it in its rare-disease register; based on ~29 reported individuals worldwide, the appropriate dismech prevalence_class is BELOW_1_IN_1000000 or, more defensibly, NOT_YET_DOCUMENTED with measure_type: CASES_IN_LITERATURE and a count. Recommended structured record: ```yaml prevalence:
• population: Worldwide measure_type: CASES_IN_LITERATURE prevalence_class: NOT_YET_DOCUMENTED notes: >- Approximately 29 affected males from ~15 families reported: 8 individuals in five reports before 2022, plus 16 individuals from 11 families in Sobering et al. 2022 (plus 5 individuals with VUS from 4 families). ```
• Incidence: ❌ unknown.
• Denominator context (do not attribute to this disorder): X-linked ID overall affects ~1–4 per 2,000 males — Qi et al. 2010 ✅: "XLMR affects 1–4 out of 2,000 males, causing intellectual disability (Intelligence Quotient (IQ) <70)". PHF8 is a rare cause within that.
• Screening-based frequency estimate: Koivisto et al. sequenced PHF8 in 18 selected patients drawn from a nationwide cohort of 7,712 cleft-surgery patients and found one family ✅ — i.e., PHF8 is a very rare cause of syndromic clefting.

9.2 Genetic epidemiology

• Inheritance: X-linked recessive (HP:0001419 ✅ in HPO annotations; GenCC/ClinGen submissions use HP:0001417 X-linked inheritance). Suggested dismech inheritance_term: HP:0001419 X-linked recessive inheritance (bind the term: — do not leave preferred_term alone).
• Penetrance: appears complete in hemizygous males for developmental delay (16/16) and near-complete for ID (14/16); incomplete/variable for clefting (3/16 in the unbiased cohort). Carrier females: essentially non-penetrant.
• Expressivity: highly variable, including within families. Abidi et al. ✅ documents intrafamilial variability directly: "One of the truncating mutations was found in the original family with Siderius-Hamel CL/P syndrome where only two of the three affected individuals had mental retardation (MR) with CL/P and one individual had mild MR. The second mutation was present in a family with four affected men, three of whom had MR and CL/P, while the fourth individual had mild MR without clefting."
• Genotype–phenotype correlation: none identified. Sobering et al. ⚠: "The phenotypic variability does not appear to be linked to the variant location in individuals who harbor a null allele." Variants inside vs. outside the JmjC domain did not differ in severity.
• Anticipation: not applicable (not a repeat-expansion disorder).
• Germline mosaicism: ❌ not reported; cannot be excluded (relevant to recurrence counseling after an apparently de novo variant).
• Founder effects: none. The Finnish F279S family is a single family, not a founder population effect.
• Consanguinity: not a factor (X-linked recessive in males).
• Carrier frequency: ❌ unknown; no population carrier-screening data.

9.3 Population demographics

• Sex ratio: affected individuals are essentially all male. Carrier females are typically unaffected (skewed XCI). ❌ No symptomatic heterozygous female has been convincingly reported in the retrieved literature — this is worth stating explicitly and worth flagging as something to re-check as cohorts grow.
• Ethnic/geographic distribution: no predilection. Sobering's 11 families were of diverse ancestry: 9 Western European, 3 Moroccan, 2 Asian-Indian, 1 Afro-Caribbean (family counts as reported) ⚠. Prior families were Dutch (Siderius), French/Italian (Laumonnier), American (Abidi), Finnish (Koivisto).
• Age distribution: lifelong from birth; published individuals span infancy to adulthood.

10. Diagnostics

10.1 Genetic testing (the diagnostic backbone)

There is no biochemical or imaging test that establishes this diagnosis — it is molecular.

Carrier testing: targeted familial variant testing in the mother; X-inactivation studies are informative but not diagnostic.

10.2 Clinical, imaging and functional testing (supportive, not diagnostic)

• Brain MRI: variable and non-specific — normal in some, polymicrogyria/cortical dysplasia, thin corpus callosum, striatal signal change, or increased subarachnoid space in others (6 imaged) ⚠. Justified where seizures or focal findings exist.
• EEG: indicated for the ~31% with seizures ⚠.
• Developmental/neuropsychological assessment: to quantify ID severity and identify ASD/ADHD (formal ASD and ADHD assessment should be routine given ~44% each).
• Audiology and speech assessment: required with clefting (velopharyngeal insufficiency → hypernasal speech, HP:0001611) and given universal speech delay.
• Laboratory tests / biomarkers: ❌ none. No enzyme assay, no metabolite, no circulating biomarker. Histone-methylation levels in patient cells are a research measure, not a clinical test.
• Biopsy / histopathology: not indicated.

10.3 Clinical criteria and differential diagnosis

• Standardized diagnostic criteria: ❌ none published. Diagnosis = compatible phenotype + hemizygous PHF8 LoF variant.
• Differential diagnosis:
• Other XLID chromatinopathies, especially the functionally linked ones: KDM5C/JARID1C (Claes-Jensen; microcephaly, short stature) and ZNF711-related XLID — mechanistically intertwined with PHF8 (Kleine-Kohlbrecher 2010 ✅; Poeta 2019 ⚠).
• Other syndromic clefting + ID disorders: MSX1-related (Witkop/orofacial cleft), IRF6 (Van der Woude — lip pits), TP63 (EEC), 22q11.2 deletion (velocardiofacial — cardiac, immune, hypocalcemia), Kabuki syndrome (KMT2D/KDM6A — another chromatinopathy with clefting).
• Xp11.22 contiguous deletion syndrome (broader phenotype; distinguished by CMA).
• Nonsyndromic cleft lip/palate with coincidental ID — distinguished only by molecular testing.
• Because clefting is present in only ~19%, PHF8 should be considered in males with unexplained DD/ID + ASD/ADHD + subtle dysmorphism even without clefting — arguably the single most actionable clinical message of the 2022 paper.

10.4 Screening

• Newborn screening: ❌ not applicable (no treatable metabolic marker).
• Population carrier screening: ❌ not offered; PHF8 is not on standard expanded carrier-screening panels.
• Cascade testing of at-risk female relatives after a proband diagnosis: standard of care and the main practical screening activity.

11. Outcome / Prognosis

• Survival / life expectancy / mortality: ❌ No mortality or survival data have been published. No deaths attributable to the disorder are reported. Adults are described in multiple families (the original Siderius pedigree and Abidi's four affected men), implying survival to adulthood is expected. Do not assert a numeric life expectancy. The defensible statement is: "no evidence of reduced life expectancy has been reported; formal survival data are absent."
• Morbidity: driven by (1) lifelong intellectual disability with educational and independent-living implications; (2) neurobehavioral comorbidity (ASD ~44%, ADHD ~44%); (3) epilepsy (~31%); (4) surgical and speech burden of clefting where present (~19%); (5) infant feeding difficulty/failure to thrive (~63%).
• Disability outcomes: ID severity spans borderline to severe; most published individuals fall in the mild-to-moderate range. Two individuals in Sobering had no ID (dyscalculia; mild learning difficulty) ✅ — the mild tail of the spectrum is real.
• QoL measures: ❌ none applied.
• Complications: velopharyngeal insufficiency/hypernasal speech, feeding and growth issues in infancy, seizure-related morbidity, behavioral comorbidity. No progressive organ complications.
• Recovery potential: none for the ID itself (static encephalopathy). Cleft repair, speech therapy and educational intervention meaningfully improve function.
• Prognostic factors: ❌ none validated. Notably, variant type and position do not predict severity ⚠ (Sobering), which is itself a prognostically relevant negative finding — counseling cannot be refined by genotype. Sibling pairs were more concordant than the cohort at large ⚠, hinting at genetic-background modifiers.
• Prognostic biomarkers: ❌ none.

12. Treatment

There is no disease-modifying therapy. Management is entirely supportive/symptomatic and follows generic best practice for syndromic ID with clefting. Sobering et al. explicitly offer no management recommendations beyond diagnosis and counseling ⚠.

12.1 Suggested treatment annotations (NCIT — ⚠ all IDs require OAK verification against sqlite:obo:ncit)

12.2 Pharmacogenomics

❌ No PHF8-specific pharmacogenomic interactions are known. Standard CPIC guidance applies to any ASM or psychotropic used.

12.3 Advanced therapeutics

• Gene therapy / gene editing / ASO / siRNA / cell therapy: ❌ none in development. No PHF8 program appears in the retrieved literature or trial registries. Note that ASO strategies are ill-suited here — the lesion is loss of a large multidomain nuclear enzyme, not a splice-correctable or knockdown-amenable target (a few canonical splice-site variants are conceivable splice-modulation targets, but this is speculative).
• Targeted / repurposed pharmacotherapy — the one real lead: rapamycin/sirolimus (CHEBI:9168 ✅), based on Chen et al. 2018 ✅ ("Pharmacological suppression of mTOR signaling with rapamycin in Phf8 knockout mice recovers the weakened LTP and cognitive deficits... provides a potential therapeutic drug target to treat XLID"). Strictly preclinical, mouse-only, and complicated by the Walsh 2017 Phf8-KO line showing no cognitive impairment (§15.4). Curate as evidence_source: MODEL_ORGANISM with an explicit caveat; do not present as a treatment option.
• Immunotherapy: not applicable.

12.4 Clinical trials

❌ No clinical trials for Siderius syndrome / PHF8-XLID were identified. No NCT identifiers to record.

12.5 Treatment outcomes, adverse events, algorithms

❌ No disease-specific response rates, adverse-event data, treatment algorithms, combination regimens, or genotype-guided strategies exist. Cleft repair outcomes follow general cleft-care literature and should not be attributed to this disorder.

13. Prevention

• Primary prevention: not possible for a germline monogenic disorder. Genetic counseling is the operative intervention: a carrier mother has a 50% chance of transmitting the variant to each child — affected sons, carrier daughters. De novo occurrence in ~3/11 families ⚠ means recurrence risk is low but not zero after an apparently de novo event (germline mosaicism cannot be excluded).
• Reproductive options: prenatal diagnosis (CVS/amniocentesis with targeted variant testing), preimplantation genetic testing for monogenic disorders (PGT-M), donor gametes. Prenatal ultrasound may detect clefting; Xp11.22 deletions are detectable by prenatal CMA (PMID:32219840) ⚠.
• Secondary prevention (early detection): cascade carrier testing of maternal female relatives; early developmental surveillance in known at-risk males, enabling early speech/behavioral intervention before formal diagnosis.
• Tertiary prevention (complication avoidance): timely cleft repair to prevent speech and feeding morbidity; audiologic surveillance (middle-ear disease in cleft palate); seizure control; proactive ASD/ADHD assessment (the 2022 paper's practical message is that these are under-recognized in this disorder).
• Immunization / public-health / environmental interventions: not applicable. (General periconceptional folic acid is population cleft-prevention advice; it has no PHF8-specific evidence base.)
• Prophylaxis: none.

14. Other Species / Natural Disease

14.1 Orthologs and taxonomy

Conservation quote (Kleine-Kohlbrecher 2010, cached full text) ✅:

"Two closely related homologs are present in C. elegans showing conservation of the overall domain structure and of the primary amino acid sequence in the PHD and JmjC domains, indicating an evolutionary conserved role for these proteins."

14.2 Naturally occurring disease in other species

❌ No naturally occurring PHF8-related disease is recorded in OMIA or the veterinary literature (no companion-animal or livestock PHF8 disorder was found). No breed associations; no VBO identifiers apply. Veterinary relevance: none.

14.3 Comparative biology

The enzymatic function is deeply conserved (worm→human), and so is the neuronal requirement: the C. elegans homolog is "highly expressed in neurons, and mutant animals show impaired locomotion" ✅, zebrafish morphants show brain-cell apoptosis and jaw defects ✅, and mice show hippocampal LTP/memory deficits ✅ (in one line). The craniofacial arm is conserved from fish to human (jaw/msxb in zebrafish; cleft lip/palate in humans; "neural crest defects" in mice ✅ per Kremp 2024). The notable non-conservation is the mouse craniofacial phenotype: no mouse model reproduces cleft lip/palate (§15.4).

14.4 Transmission

Not applicable — non-infectious, non-zoonotic.

15. Model Organisms

15.1 Mouse (Mus musculus) — MGI:2444341

Two independently generated Phf8 knockout lines, with discordant cognitive phenotypes — this is the central caveat of the mouse literature:

MGI's own summary of the two alleles ⚠: one null allele → impaired learning/memory and impaired hippocampal LTP; another null allele → resiliency to depression-like behavior and decreased anxiety.

Mouse embryonic work (2026) adds a developmental phenotype: PHF8 deficiency halted neurogenesis and brain development in mouse embryos ⚠ (PMID:41714361) — potentially reconciling the adult-behavior discordance by shifting attention to embryonic stages and to background/allele differences.

Available model types: targeted knockouts (≥2 independent lines) ✅; conditional/humanized/knock-in models — ❌ none reported. IMPC data for Phf8 ❌ not retrievable this session (fetch attempt returned the wrong gene page — do not rely on it).

15.2 Zebrafish (Danio rerio)

The best craniofacial model. Qi et al. 2010 ✅: phf8 knockdown → apoptosis/reduced cell survival in the brain and defective jaw development, mediated at least in part by direct regulation of msxb (MSX1); catalytically dead PHF8 fails to rescue, tying the phenotype to demethylase activity. A separate zebrafish study extends the sensory phenotype: He et al. 2020 (PMID:33330448) ⚠ — "PHF8 knockdown significantly disrupted the development of the posterior lateral line system" and caused "severe malformation of the semicircular canal and otoliths." (Inner-ear involvement has not been described in humans — treat as model-only.)

15.3 Invertebrate and cellular models

• C. elegans — F29B9.2/jmjd-1.2: neuronal expression, impaired locomotion, global H3K9me2/H3K27me2 increase in mutants ✅. Also the vehicle for the GenX toxicology study ⚠.
• P19 mouse embryonal carcinoma cells — RA-induced neuronal differentiation assay; PHF8 knockdown impairs it, WT (but not F279S) rescues/drives differentiation ✅ (Qiu 2010). A 2024 follow-up: PMID:38463639 ⚠.
• Primary rodent oligodendroglial cultures + oligodendroglial cell lines ✅ (Kremp 2024).
• Mouse astrocyte cultures ✅ (Iacobucci 2021).
• Human iPSC-derived oligodendrocytes — notably, "generation of human oligodendrocytes from induced pluripotent stem cells did not require PHF8 in a system that relies on forced expression of Olig2" ✅ (Kremp 2024) — a clean epistasis result placing Olig2 downstream.
• HeLa / U2OS / HEK — biochemistry, ChIP-seq, localization ✅.
• Patient-derived iPSC or organoid models: ❌ none published. Given the 2026 neurogenesis/serine finding, a patient-iPSC cortical organoid is the obvious next model and a strong proposed_experiments entry.

15.4 Phenotype recapitulation and limitations — curate as HUMAN_MODEL_MISMATCH

This disorder is a textbook case for dismech's HUMAN_MODEL_MISMATCH discussion kind (evidence exists in models but translational validity is the open question):

1. No mouse model reproduces the human clefting phenotype. Both published Phf8 KO lines are reported without craniofacial malformation ("neither display obvious developmental defects"; "without gross morphological defects"). The cleft phenotype is recapitulated only in zebrafish jaw development. Mechanistic implication: species differences in redundancy (PHF2/KDM7A may compensate in mouse) or in the MSX1-dependence of lip/palate fusion.
2. The two mouse KO lines disagree about cognition — one shows impaired learning/memory + LTP deficits, the other shows no cognitive impairment. Any KB claim that "Phf8 KO mice model the ID phenotype" must be qualified with both citations.
3. The mouse behavioral phenotype that is robust (stress resilience / altered serotonin signaling) has no reported human counterpart — no depression/anxiety-resilience phenotype has been described in PHF8-XLID patients. This is a genuinely testable, unaddressed clinical question.
4. The rapamycin rescue is single-line, mouse-only and rests on the line whose phenotype the other lab did not replicate.

Suggested discussion entries:

discussions:
- kind: HUMAN_MODEL_MISMATCH
  attaches_to: "pathophysiology#Reduced MSX1 Expression in Cranial Neural Crest"
  prompt: >-
    Why do Phf8-null mice fail to develop cleft lip/palate when human PHF8
    loss-of-function causes orofacial clefting and zebrafish phf8 knockdown
    causes jaw malformation?
  rationale: >-
    Both published Phf8 knockout mouse lines are reported without craniofacial
    malformation, while the craniofacial arm of the mechanism rests on zebrafish
    msxb data. Possible paralog compensation (PHF2/KDM7A) or species differences
    in the MSX1 dependence of lip/palate fusion.
- kind: HUMAN_MODEL_MISMATCH
  attaches_to: "pathophysiology#mTOR Hyperactivation via RSK1 De-repression"
  prompt: >-
    Is the mTOR-dependent cognitive phenotype (and its rapamycin rescue) a
    reliable model of human PHF8-XLID, given that a second independent Phf8
    knockout line showed no cognitive impairment?
- kind: KNOWLEDGE_GAP
  attaches_to: "pathophysiology#Loss of PHF8 Histone Demethylase Function"
  prompt: >-
    Does PHF8-XLID have a detectable DNA-methylation episignature that could
    resolve the reported missense variants of uncertain significance?

15.5 Model resources

MGI (informatics.jax.org, Phf8 = MGI:2444341), IMPC/KOMP (status ❌ unverified), ZFIN, WormBase, Alliance of Genome Resources, Cellosaurus (for HeLa/U2OS/P19).

Appendix A — Consolidated citation list

Also relevant, not fetched: an erratum to Sobering et al. exists (HGG Adv. 2022 Dec 20;4(1):100168) ✅ — check it before transcribing any variant nomenclature from Table 1.

Appendix B — Pre-commit checklist for the curator

1. just fetch-reference PMID:X for every PMID marked ❌ in Appendix A that you intend to cite, then re-verify each snippet as an exact substring.
2. Re-verify all ⚠ quotes (Sobering frequencies, Walsh abstract, Qiao, De Wolf, Huang, He, Kim, Fan, Tan, Artés) against freshly cached abstracts — the frequency table in §3.1 came from a summarizing fetch of the PMC full text, not from a verified cache.
3. Read the Sobering erratum (HGG Adv 4(1):100168) before entering variant nomenclature.
4. Run just validate-terms on the file — every ontology ID in this report marked ⚠ is unverified (OAK and OLS were both unavailable in this session; only GO:0032454, GO:0035575, GO:0060021, HP:0410030, HP:0007018, CL:2000029, CL:0002453, CHEBI:9168 were label-verified).
5. Fill the gnomAD constraint gap (pLI / LOEUF / o-e LoF) manually from gnomAD v4.
6. Pull the ClinGen assertion properly: just clingen-refresh + cite the CGGV: id rather than the web page, for the "Definitive" classification.
7. Consider adding an ORPHA:85287 structured-source evidence row via just structured-rebuild-orphanet --id 85287 for prevalence/definition (orpha.net was unreachable directly).
8. Bind inheritance_term to HP:0001419 with a full term: block (the common gap).
9. Do not add a conforms_to module reference — no existing module fits (see §6.10).
10. Record the clefting frequency discordance (§3.4) in notes:, and add the three suggested discussions entries (§15.4).
11. Add a history record: just new-history --kind disorder --slug Siderius_Type_X-Linked_Intellectual_Disability --event CREATE ....

Sources (web-retrieved, non-PubMed): OMIM 300263 · GTR: Syndromic X-linked intellectual disability Siderius type · Orphanet ORPHA:85287 · GenCC PHF8 submissions · ClinGen conditions: MONDO:0010286 · ClinVar Miner: PHF8 / Siderius · MedlinePlus Genetics: PHF8 · GARD: Siderius type · MGI Phf8 (MGI:2444341) · UniProt Q9UPP1 · NCBI Gene 23133 · HPO annotations, OMIM:300263 · EBI OLS4 · PubMed/PMC via NCBI E-utilities (PMIDs cited inline; DOIs listed in the cached reference files).