PHF21A Related Neurodevelopmental Disorder

1. Disease Information

Overview. PHF21A-related neurodevelopmental disorder is a rare monogenic neurodevelopmental syndrome characterized by the triad of intellectual disability/developmental delay, craniofacial dysmorphism, and behavioral abnormalities, with or without seizures. It belongs to the family of chromatin-remodeling ("chromatinopathy") disorders and, because of frequent postnatal overgrowth, overlaps clinically with the overgrowth–intellectual disability (OGID) syndromes (e.g., Sotos, Weaver).

Key identifiers.

Synonyms / alternative names. "PHF21A-related neurodevelopmental disorder(s)"; "IDDBCS"; historically indexed as NEDMS; BHC80-related disorder. Older literature also describes overlap with Potocki–Shaffer syndrome (the 11p11.2 contiguous gene deletion), from which PHF21A was dissected out as the gene responsible for the intellectual disability and craniofacial anomaly components.

Source of information. The knowledge base is derived from aggregated disease-level resources (OMIM, MONDO, ClinVar, gnomAD) and individual patient case series/reports in the primary literature (cohorts of 12–15 patients), rather than from population EHR data. The disorder is ultra-rare, so all epidemiologic inference comes from case aggregation.

The modern clinical entity is defined by Gavilán/Iwase (2025) PMID: 40622422: "PHF21A heterozygosity is associated with intellectual disability, behavioral issues, and craniofacial dysmorphism, with or without seizures (IDDBCS), also known as PHF21A-related neurodevelopmental disorders."

2. Etiology

Primary cause — genetic. IDDBCS is a monogenic disorder caused by heterozygous loss-of-function of PHF21A. Kim et al. (2012) mapped the intellectual disability (ID) and craniofacial anomaly (CFA) phenotypes to single-gene haploinsufficiency of PHF21A using balanced translocation breakpoints and deletion mapping at 11p11.2 (PMID: 22770980): "the ID and CFA phenotypes are both caused by haploinsufficiency of a single gene, PHF21A, at 11p11.2."

Genetic risk factors. The causal genetic events are: - De novo truncating variants (frameshift, nonsense) — the dominant class. - De novo missense variants (rare, e.g., a splice-affecting AT-hook variant). - Balanced translocations disrupting PHF21A. - Intragenic deletions and larger 11p11.2 contiguous-gene deletions encompassing PHF21A.

There are no known common susceptibility loci or modifier genes established for this ultra-rare Mendelian disorder; all pathogenic alleles are private/de novo.

Environmental risk factors. None established. As a de novo dominant Mendelian disorder, there are no recognized environmental, occupational, lifestyle, or infectious contributors. Advanced parental age is a plausible but unquantified contributor to de novo mutation rate (general principle, not specifically demonstrated for PHF21A).

Protective factors. None established (genetic or environmental).

Gene–environment interactions. None established.

3. Phenotypes

The phenotype spectrum has been characterized in three overlapping case series. Core features (ID/DD and craniofacial dysmorphism) are essentially universal; associated features vary in frequency.

Phenotype frequency table (pooled from primary cohorts)

Epilepsy characteristics. Chen et al. (2023) reported (PMID: 37633153): "Seven of the 12 patients (58.33%) had an epileptic phenotype, and the majority (5/7, 71.42%) of affected individuals developed developmental and epileptic encephalopathy (DEE)," often manifesting as infantile epileptic spasms syndrome. The associated features were quantified as: "Overgrowth, ADHD, hypotonia, ASD, and sleep disorders were observed in 100%, 77.78%, 70%, 50%, and 33.33% of patients, respectively."

Wu et al. (2023) independently corroborated core frequencies (PMID: 37528014): "intellectual disability or developmental delay (15 patients), craniofacial anomalies (15 patients), behavioral abnormalities (12 patients), seizures (9 patients), and overgrowth (8 patients)." Poole et al. (2023) placed PHF21A firmly among overgrowth syndromes (PMID: 36876344): "postnatal overgrowth was reported in 5/6 (83%)."

Craniofacial gestalt. Recurrent facial features include a tall, broad/prominent forehead (HP:0000337), sparse eyebrows (HP:0045025), broad nasal bridge/tip (HP:0000431/HP:0000455), anteverted nares (HP:0000463), full cheeks, and a downturned mouth with a tent-shaped upper lip.

Phenotype characteristics. - Age of onset: Congenital/neonatal to early childhood; developmental delay and hypotonia often evident in infancy; overgrowth is postnatal; epileptic spasms typically infantile. - Severity: Variable; ID ranges mild to severe; epilepsy skews toward severe (DEE). - Progression: Largely stable/non-degenerative developmental disability; epilepsy course varies. - Expressivity: Variable, even for identical variants (see recurrent p.Arg580*).

Quality of life impact. No formal EQ-5D/SF-36/PROMIS data exist for this ultra-rare disorder. Impact is inferred from the phenotype: intellectual disability, behavioral comorbidities (ADHD/ASD), and epilepsy (particularly DEE) impose substantial lifelong functional and caregiving burden.

4. Genetic / Molecular Information

Causal gene. PHF21A (PHD finger protein 21A; alias BHC80), gene OMIM 608325, HGNC:24156, NCBI Gene 51317, Ensembl ENSG00000135365, UniProt Q96BD5 (680 aa), at 11p11.2.

Pathogenic variant landscape. The variant spectrum is dominated by de novo loss-of-function alleles. Chen et al. (2023) pooled 12 patients (PMID: 37633153): all variants were de novo heterozygous; "The most common types of variants were frameshift variants (7/12, 58.33%), followed by nonsense variants (4/12, 33.33%) and missense variants (1/12, 8.33%)."

Recurrent hotspot. The nonsense variant p.Arg580* recurs and shows variable expressivity: "Three of the 12 patients (25%) had the same variant (p.Arg580*)."

Functional consequence — haploinsufficiency (loss of function). Hamanaka et al. (2019) reported de novo truncating variants and concluded (PMID: 31649809): "haploinsufficiency is the likely underlying mechanism in the phenotype." Notably, even the only recurrent missense variant (c.1285G>A, at the last nucleotide of exon 13 in the AT-hook motif) acts by reducing splicing efficiency/dosage while preserving DNA binding — Gavilán/Iwase (2025) concluded (PMID: 40622422): "reduced dosage rather than impaired DNA binding likely contributes to the cognitive impairments." This argues against dominant-negative or gain-of-function mechanisms; reduced gene dosage is the unifying pathomechanism. Truncations tend to converge on the AT-hook domain and a C-terminal intrinsically disordered region.

Population constraint (supports haploinsufficiency). gnomAD constraint metrics for PHF21A place it among the most LoF-intolerant genes:

No common LoF alleles exist in the general population; disease-causing variants are private/de novo.

ClinVar summary. ClinVar contains ~427 PHF21A variant records: the large majority (~370) are variants of uncertain significance (VUS), ~21 likely pathogenic, and dozens pathogenic (the pathogenic set includes large 11p11.2 contiguous deletions encompassing PHF21A). The VUS-heavy landscape reflects both the gene's constraint and limited functional annotation.

Allele frequency. Pathogenic alleles are absent/singleton in gnomAD (private, de novo); no recurrent population allele.

Somatic vs germline. Germline (constitutional), typically de novo.

Modifier genes. None established.

Epigenetic information. As a chromatin-reader deficiency, the disorder alters the epigenomic repression state at target loci (loss of LSD1/CoREST/HDAC recruitment reduces H3K4 demethylation and histone deacetylation at target promoters), but a defined patient DNA-methylation "episignature" has not been confirmed here.

Chromosomal abnormalities. Balanced translocations disrupting PHF21A; intragenic deletions; and larger 11p11.2 contiguous-gene deletions (overlapping the Potocki–Shaffer region).

5. Environmental Information

Not applicable. No environmental factors, lifestyle factors, or infectious agents are implicated in this de novo Mendelian chromatinopathy. Disease is fully explained by the germline genetic lesion.

6. Mechanism / Pathophysiology

Molecular reader function

PHF21A/BHC80 is a histone reader. Lan et al. (2007) showed (PMID: 17687328): "the PHD finger of BHC80 binds unmethylated H3K4 (H3K4me0), and this interaction is specifically abrogated by methylation of H3K4." PHF21A also binds DNA via an AT-hook motif. By reading the "unmodified/repressive" H3K4me0 mark, PHF21A anchors a repressor complex to chromatin.

The LSD1/CoREST/HDAC (BHC) complex

PHF21A is a core subunit of the LSD1(KDM1A)/CoREST/HDAC1-2 complex. Shi et al. (2005) established (PMID: 16140033): "LSD1 is associated with HDAC1/2; CoREST, a SANT domain-containing corepressor; and BHC80, a PHD domain-containing protein." Within this complex there is reciprocal dependence: BHC80 (PHF21A) and LSD1 depend on each other for stable chromatin association. Critically, PHF21A is required for repression — Lan et al. showed (PMID: 17687328): "Knockdown of BHC80 by RNA inhibition results in the de-repression of LSD1 target genes."

REST/NRSF neuronal gene silencing

The BHC complex silences neuron-specific genes through the RE1/neural restrictive silencer (NRS) element — the REST/NRSF pathway. Kim et al. (2012) confirmed (PMID: 22770980): "PHF21A, also known as BHC80, is a component of the BRAF-histone deacetylase complex that represses target-gene transcription."

Downstream consequences of haploinsufficiency

1. Target-gene derepression. In translocation-patient lymphoblasts, PHF21A disruption derepressed the neuronal gene SCN3A with reduced LSD1 occupancy — Kim et al. (PMID: 22770980): "we observed derepression of the neuronal gene SCN3A and reduced LSD1 occupancy at the SCN3A promoter."

2. Broad transcriptional dysregulation & impaired cAMP/CREB signaling. RNA-seq of two PHF21A-haploinsufficient patient cell lines identified 1,885 commonly misregulated genes. Porter/Iwase (2018) reported (PMID: 28571721): "The patient cells displayed down-regulation of key pathways relevant to learning and memory, including Cyclic Adenosine Monophosphate (cAMP)-signaling pathway genes," and functionally "PHF21A-deficient patient-derived cells exhibited a delayed induction of immediate early genes following forskolin stimulation" — i.e., impaired activity-dependent (CREB-driven) transcription.

3. Dysregulated synaptogenesis via neuronal microexon splicing. PHF21A and LSD1 both undergo neuron-specific microexon splicing. Nagai/Iwase (2024) showed the PHF21A neuronal microexon (exon 14) interferes with nucleosome binding, producing stepwise deactivation of the LSD1–PHF21A complex during neuronal maturation. Forcing the canonical (non-neuronal) PHF21A isoform in neurons causes excess synapses — PMID: 39395799: "Phf21a neuronal splicing prevents excess synapse formation that otherwise would occur when canonical PHF21A is expressed in neurons."

Causal chain (upstream → downstream)

Heterozygous PHF21A LoF (de novo)          [UPSTREAM]
│  (haploinsufficiency; ~50% protein dose)
▼
Weakened H3K4me0 reading / reduced LSD1-CoREST-HDAC
chromatin anchoring at RE1/NRS neuronal loci
│
▼
Derepression of neuronal target genes (e.g., SCN3A)
+ impaired cAMP/CREB activity-dependent transcription
+ dysregulated synaptogenesis (microexon isoform balance)
│
▼
Aberrant neuronal differentiation, synapse number,
and network excitability; craniofacial developmental defects
│
▼
Intellectual disability, behavioral abnormalities,       [DOWNSTREAM]
craniofacial dysmorphism, overgrowth, epilepsy (DEE)

Cell types & processes. Neurons (CL:0000540) and neural progenitors; craniofacial/neural crest derivatives. GO biological processes: negative regulation of transcription (GO:0000122), histone H3-K4 demethylation (GO:0034720), chromatin organization (GO:0006325), regulation of synapse assembly (GO:0051963), learning or memory (GO:0007611), cAMP-mediated signaling (GO:0019933). GO cellular components: nucleus (GO:0005634), chromatin (GO:0000785), CoREST/LSD1/HDAC complex.

Protein dysfunction. Loss of function via truncation/dosage reduction; the enzymatically inactive neuronal complex interacts with neuron-specific partners including MYT1-family transcription factors and VIRMA.

Metabolic / immune / tissue-damage mechanisms. No primary metabolic, immune, or tissue-necrosis mechanism; this is a developmental transcriptional-regulatory disorder.

7. Anatomical Structures Affected

Organ / system level. - Primary: Central nervous system — brain (UBERON:0000955); nervous system (UBERON:0001016). - Craniofacial skeleton — head/face (UBERON:0000033), reflecting neural crest/craniofacial developmental involvement. - Body systems: Nervous system (cognition, seizures, behavior); musculoskeletal/growth axis (postnatal overgrowth, hypotonia).

Tissue and cell level. Nervous tissue; neurons (CL:0000540) and neural progenitor cells (CL:0011020); neural-crest-derived craniofacial mesenchyme.

Subcellular level. Nucleus (GO:0005634) and chromatin (GO:0000785) — PHF21A is a nuclear chromatin-associated protein.

Localization / lateralization. Craniofacial features are bilateral/symmetric; brain involvement is diffuse rather than focal.

8. Temporal Development

Onset. Congenital to infantile. Hypotonia and developmental delay are often apparent in infancy; craniofacial features are present from birth/early childhood; postnatal overgrowth emerges after birth; epileptic spasms are typically infantile in onset.

Progression. The intellectual disability is a stable, non-degenerative developmental disability rather than a progressive neurodegeneration. Epilepsy course is variable; a substantial subset evolves into developmental and epileptic encephalopathy (DEE) with attendant developmental impact.

Disease duration. Chronic, lifelong.

Critical periods. Neurodevelopmental windows (fetal/infantile neuronal differentiation and synaptogenesis) are the mechanistically relevant vulnerable periods, given PHF21A's role in the LSD1-complex "handoff" during neuronal maturation. Early seizure control (infancy) is the primary time-sensitive intervention opportunity.

9. Inheritance and Population

Epidemiology. Ultra-rare; precise prevalence/incidence are not established. Fewer than ~40 patients are described in aggregate cohorts (12, 15, and 13 patients in the three main series). No population registry estimates exist.

Inheritance pattern. Autosomal dominant, almost always de novo. Because PHF21A is on 11p (autosome), inheritance is not sex-linked.

Penetrance. Effectively complete for the neurodevelopmental phenotype in reported de novo cases; expressivity is variable even for the identical recurrent p.Arg580* variant (present in 3/12 patients with differing severity).

Genetic anticipation. Not applicable (not a repeat-expansion disorder).

Germline mosaicism / founder effects / consanguinity. Not established as relevant; disease is de novo dominant, so consanguinity is not a driver and there are no founder alleles. Recurrence risk for parents of an affected de novo proband is low but non-zero owing to the theoretical possibility of parental gonadal mosaicism.

Carrier frequency. Not applicable (dominant, de novo; gnomAD shows no common LoF carriers).

Population demographics. No ethnic or geographic predilection reported; no established sex ratio skew. Age distribution reflects a pediatric-onset lifelong condition.

10. Diagnostics

Genetic testing is the diagnostic cornerstone.

Clinical evaluation. Developmental/cognitive assessment; EEG and MRI for seizures; growth monitoring (overgrowth); behavioral/psychiatric evaluation (ADHD/ASD, sleep). No specific biochemical biomarker, imaging signature, or laboratory abnormality is diagnostic; there is no metabolic marker.

Diagnostic criteria. No formal consensus criteria; diagnosis is molecular (pathogenic/likely pathogenic PHF21A variant) in a patient with a compatible phenotype (ID/DD, craniofacial dysmorphism ± behavioral abnormalities ± seizures ± overgrowth).

Differential diagnosis. Other overgrowth–intellectual disability (OGID) syndromes (Sotos/NSD1, Weaver/EZH2, Tatton-Brown–Rahman/DNMT3A, Malan/NFIX), other chromatinopathies, and Potocki–Shaffer syndrome (larger 11p11.2 deletion). Molecular testing distinguishes them.

Screening. No newborn or population carrier screening (disorder is de novo, ultra-rare). Cascade testing of parents is chiefly to establish de novo status and recurrence risk.

11. Outcome / Prognosis

Survival / mortality. No specific life-expectancy or mortality data are established. The disorder is not intrinsically lethal; prognosis is dominated by neurodevelopmental disability and seizure burden rather than early mortality. Severe DEE carries the usual associated risks of refractory infantile epilepsy.

Morbidity / function. Lifelong intellectual disability and behavioral comorbidities produce substantial functional impairment and dependency. Hypotonia affects early motor development.

Disease course. Non-progressive developmental disability with chronic, lifelong needs; epilepsy (when present) is a major determinant of outcome, and DEE portends worse cognitive trajectory.

Prognostic factors. Presence and severity of epilepsy (particularly DEE/infantile spasms) is the key prognostic modifier; degree of ID is variable. No molecular prognostic biomarker beyond variant presence is established (variable expressivity limits genotype–phenotype prediction).

12. Treatment

No targeted or disease-modifying therapy exists. Management is supportive and symptom-directed.

Pharmacotherapy. Antiseizure medication is the principal pharmacologic intervention; vigabatrin is highlighted for infantile epileptic spasms syndrome. ADHD and ASD are treated per standard symptomatic approaches.

Advanced therapeutics. No approved gene therapy, cell therapy, RNA-based therapy, or targeted small molecule. Given the dosage-reduction mechanism, dosage-restorative strategies (e.g., approaches that raise residual PHF21A expression) are conceptually attractive but experimental. No pharmacogenomic guidance is established.

Experimental treatments / clinical trials. None identified specific to IDDBCS.

Treatment strategy. Multidisciplinary care (neurology, developmental pediatrics, genetics, therapy services); early aggressive seizure control; individualized developmental and behavioral supports.

13. Prevention

Primary prevention. Not applicable — de novo dominant Mendelian disorder cannot be prevented by risk-factor modification, immunization, or public-health measures.

Secondary prevention / early detection. Early molecular diagnosis (trio exome/genome) enables prompt developmental intervention and early, targeted seizure management (vigabatrin for spasms), which is the most impactful available "preventive" action for downstream disability.

Genetic counseling. Central to family management: for a de novo proband, recurrence risk for future siblings is low but non-zero (parental gonadal mosaicism). Affected individuals who reproduce would have 50% transmission risk (autosomal dominant). Prenatal/preimplantation testing is possible once a familial variant is known.

Screening. No population or newborn screening.

14. Other Species / Natural Disease

Taxonomy / orthologs. - Mouse: Phf21a (NCBI Gene 192285; Ensembl ENSMUSG00000058318), Mus musculus (NCBI Taxon 10090). - Zebrafish: phf21a ortholog, Danio rerio (NCBI Taxon 7955).

Naturally occurring disease. No naturally occurring companion-animal or wildlife disease attributable to PHF21A is documented (no OMIA entry identified here). Relevance is confined to engineered/experimental models.

Comparative biology / evolutionary conservation. The LSD1/CoREST/PHF21A repressor module and the neuron-specific microexon-splicing program are conserved across vertebrates, underpinning the utility of mouse and zebrafish models for studying the disorder's mechanism.

Transmission / zoonotic potential. Not applicable (genetic, non-transmissible).

15. Model Organisms

Phenotype recapitulation. The zebrafish morphant recapitulates the craniofacial dimension and demonstrates neuronal apoptosis — Kim et al. (PMID: 22770980): "suppression of the latter led to both craniofacial abnormalities and neuronal apoptosis." Mouse models illuminate the synaptogenesis-control mechanism via neuronal microexon splicing.

Limitations. Morpholino knockdown has known off-target/transient caveats; mouse "forced-isoform" experiments model a mechanistic axis rather than the exact human haploinsufficient genotype; cognitive/behavioral recapitulation of human ID is inherently limited. Resources: MGI (mouse), ZFIN (zebrafish).

Mechanistic Model / Interpretation

The disorder is best understood as a chromatin-reader haploinsufficiency. A single functional PHF21A allele cannot fully staff the LSD1/CoREST/HDAC repressor complex at neuronal RE1/NRS loci. The resulting partial loss of repression derepresses neuronal genes prematurely/inappropriately (SCN3A being a validated example) and weakens activity-dependent cAMP/CREB transcriptional responses, while the finely tuned neuronal microexon-splicing "handoff" that normally throttles synapse formation is perturbed. Because PHF21A is required at a developmental inflection point — the transition of the LSD1 complex from a proliferative/progenitor configuration to a neuronal one — the phenotype manifests as a fixed developmental disorder (ID, dysmorphism, overgrowth) with a superimposed excitability phenotype (epilepsy/DEE).

The convergence of three independent lines of evidence — (1) gnomAD extreme LoF-intolerance (pLI = 1.0, LOEUF = 0.22), (2) uniformly de novo truncating variants, and (3) the missense-splice variant that reduces dosage without impairing DNA binding — makes haploinsufficiency the unambiguous mechanism, effectively excluding dominant-negative and gain-of-function models.

Evidence Base

Limitations and Knowledge Gaps

• Small cohorts. All clinical data derive from case series totaling ~40 patients; frequency estimates have wide uncertainty and possible ascertainment bias toward severe phenotypes.
• No epidemiologic estimates. Prevalence/incidence, mortality, and life expectancy are unquantified.
• VUS-dominated ClinVar. ~370/427 variants are VUS; functional annotation is limited, complicating classification of new variants.
• No episignature confirmed. A patient DNA-methylation signature (useful for diagnostics) has not been validated here.
• No QoL instruments applied specifically to IDDBCS.
• Genotype–phenotype prediction is weak given variable expressivity (identical p.Arg580* with divergent severity).
• No approved targeted therapy and no clinical trials identified.
• Citation verification. Quotes are drawn from provided abstract snippets; they should be re-verified against source abstracts before formal publication.

Proposed Follow-up Experiments / Actions

1. Reclassify ClinVar VUS via a high-throughput functional assay (e.g., saturation genome editing or splicing/reporter assays) leveraging the established dosage-reduction mechanism, to convert VUS to actionable calls.
2. Define a blood DNA-methylation episignature across a patient cohort to enable a cheap, robust clinical diagnostic and VUS-arbitration tool (mirroring successful approaches for other chromatinopathies).
3. Natural-history / registry study to establish prevalence, developmental trajectories, seizure outcomes (including DEE risk factors), and QoL using standardized instruments (Vineland, PROMIS).
4. iPSC-derived neuron / organoid modeling of patient variants to map the transcriptional derepression program and test whether pharmacologic modulation (e.g., HDAC/LSD1-axis or cAMP/CREB enhancers) rescues activity-dependent transcription.
5. Dosage-restoration proof-of-concept (e.g., ASO- or CRISPRa-mediated upregulation of the residual allele) in mouse/zebrafish, since haploinsufficiency is a favorable target for gene-dosage therapies.
6. Seizure-management evidence synthesis — formally evaluate vigabatrin and other agents for PHF21A-related infantile spasms to build a genotype-informed treatment guideline.

Consensus Answer

PHF21A-related neurodevelopmental disorder (IDDBCS; OMIM #618725; MONDO:0032883) is a rare autosomal-dominant, almost always de novo Mendelian condition caused by heterozygous loss-of-function (haploinsufficiency) of PHF21A/BHC80 at 11p11.2. PHF21A is a chromatin reader of unmethylated H3K4 that anchors the LSD1/CoREST/HDAC repressor complex, so its loss derepresses neuronal target genes (e.g., SCN3A), blunts cAMP/CREB-dependent transcription, and dysregulates synaptogenesis — producing near-universal intellectual disability/developmental delay and craniofacial dysmorphism with frequent postnatal overgrowth, behavioral problems (ADHD/ASD), hypotonia, and epilepsy (often infantile spasms/DEE). Diagnosis is by exome/genome sequencing or CMA/karyotype, and management is supportive (developmental therapies and antiseizure medication, with vigabatrin effective for spasms), as no targeted or curative therapy exists.

Artifacts

• OpenScientist final report
• OpenScientist final report

Reference Validation

Checked with linkml-reference-validator 0.2.1.

All extracted references resolved successfully.

Term Validation

Checked with linkml-term-validator 0.4.5, through the ols: adapter.

Terms the report names something else

These identifiers resolve, so nothing about them looks wrong, and the ontology calls them something unrelated to what the report calls them. That usually means the identifier is not the one the sentence needs:

• HP:0008872 (1 mention) - the report calls it "Postnatal overgrowth"; HP calls it Feeding difficulties in infancy
• HP:0045025 (1 mention) - the report calls it "sparse eyebrows"; HP calls it Narrow palpebral fissure
• UBERON:0000033 (1 mention) - the report calls it "Craniofacial skeleton — head/face"; UBERON calls it head**

Obsolete terms

These terms are real but deprecated. Citing one is not a fabrication; it does mean the report is naming something the ontology has retired:

• GO:0034720 (obsolete histone H3-K4 demethylation) (1 mention)
• GO:0019933 (obsolete cAMP-mediated signaling) (1 mention)

Terms whose name is worth a second look

The report's name for these is recognisably related to the term's own name without being one of them. A loose paraphrase reads the same way as a citation of the wrong sibling term - and so does a related synonym, which the ontology records precisely because it names something adjacent rather than the same thing - so these are listed rather than judged:

• HP:0001999 (1 mention) - the report calls it "Craniofacial dysmorphism"; HP calls it Abnormal facial shape, and lists "Facial dysmorphism" among its other names
• HP:0000708 (1 mention) - the report calls it "Behavioral abnormalities"; HP calls it Atypical behavior, and lists "Behavioral abnormality" among its other names
• HP:0001250 (1 mention) - the report calls it "Epilepsy / seizures"; HP calls it Seizure, and lists "Epileptic seizure" among its other names
• HP:0200134 (1 mention) - the report calls it "Developmental & epileptic encephalopathy (DEE)"; HP calls it Epileptic encephalopathy
• HP:0000729 (1 mention) - the report calls it "Autism spectrum disorder"; HP calls it Autistic behavior, and lists "Autism spectrum disorder" among its other names
• HP:0002360 (1 mention) - the report calls it "Sleep disorder"; HP calls it Sleep disturbance
• HP:0000337 (1 mention) - the report calls it "tall, broad/prominent forehead"; HP calls it Broad forehead
• UBERON:0000955 (1 mention) - the report calls it "Primary: Central nervous system — brain"; UBERON calls it brain**, and lists "suprasegmental levels of nervous system" among its other names