Comprehensive Research Report

Autosomal Recessive Non-Syndromic Intellectual Disability (AR‑NSID / NS‑ARID)

Prepared: 1 August 2026 · Target MONDO: MONDO:0019502 · Category: Genetic

Curation caveat up front (dismech SOP §2a/§2b). This report is a lead document. Several abstracts below were retrieved through a summarizing fetch layer that returned quoted fragments rather than complete verbatim text; those are marked [fragment — re-verify]. Before any snippet here is committed to a kb/disorders/ YAML file, run just fetch-reference PMID:XXXX and just validate-references. Fully verbatim-retrieved abstracts are marked [verbatim].

A second, structural caveat specific to this entry. AR‑NSID is not a disease in the usual dismech sense — it is a nosological class (an OMIM phenotypic series and a MONDO grouping node with 72 descendants and ~56 causal genes). Section 16 at the end discusses what this implies for how the dismech entry should be shaped (strong candidate for a kb/groupings/ entity plus/instead of a Disease).

1. Disease Information

1.1 Overview

Autosomal recessive non-syndromic intellectual disability is the class of intellectual disability (ID) in which (a) biallelic (homozygous or compound heterozygous) variants in an autosomal gene are the cause, and (b) impaired intellectual functioning and adaptive behaviour occur without a consistent, recognizable pattern of associated dysmorphic, malformative, metabolic, or neuroimaging features. It corresponds to the OMIM "intellectual developmental disorder, autosomal recessive" (MRT) numbered series.

Two features define its epistemic character:

1. Extreme genetic heterogeneity. Ropers/Jamra estimate "2500–3000 ARID genes," of which "less than 700 confirmed genes and less than 400 candidate genes have been identified" (PMID:30459488) [fragment — re-verify]. Harripaul et al. write [verbatim]: "Previous studies have indicated high levels of genetic heterogeneity, with estimates of more than 2500 autosomal ID genes, the majority of which are autosomal recessive (AR)." (PMID:28397838)

2. An unstable syndromic/non-syndromic boundary. The "non-syndromic" label is frequently provisional. Jamra states [fragment — re-verify]: "many of the cases were rather unspecific and several ID forms that were reported initially to be nonsyndromic turned out to be syndromic, as other cases with overlapping phenotypes have been identified" (PMID:30459488). This is the single most important curation caveat for this entry — see §3.4 and §16.

1.2 Key identifiers (verified)

Monarch association counts for MONDO:0019502 (retrieved 2026-08-01, Monarch API v3): 72 descendants; 61 correlated genes (54 direct); 56 causal genes; 1131 disease→phenotype associations; 5 disease models.

1.3 Synonyms (from the MONDO record, verified)

• Exact: AR‑NSID; NS‑ARID; autosomal recessive non-syndromic intellectual disability; mental retardation, autosomal recessive; non-syndromic intellectual disability, autosomal recessive
• Broad: autosomal recessive intellectual disability (ARID); intellectual disability, autosomal recessive
• In the literature: NS‑ARMR (non-syndromic autosomal recessive mental retardation); NSID (used loosely for both AD and AR forms); "MRT" series

1.4 Source type

Almost all AR‑NSID knowledge is aggregated disease-level and derives from a specific study design: homozygosity/autozygosity mapping plus exome or genome sequencing in large multiplex consanguineous families, predominantly Iranian, Pakistani, Turkish, and Arab. There is essentially no EHR-derived phenotyping literature for AR‑NSID as a class, and this ascertainment design systematically biases the phenotype spectrum (see §9.4).

2. Etiology

2.1 Primary causal factor

Biallelic pathogenic variants — usually homozygous by descent, less often compound heterozygous — in any one of a very large number of autosomal genes. There is no dominant gene and no dominant pathway. Jamra's conclusion is the key statement [fragment — re-verify]: "there are no prevalent ARID genes, pathways, or protein complexes and the functions of the affected proteins are very diverse and limited not only to neurological aspects" — and therefore "there is no justification for a gene-specific or panel diagnostic when ARID is suspected." (PMID:30459488)

Recurrence across large cohorts is strikingly low. Among 1,131 families aggregated across the major studies, the most frequently implicated genes were [fragment — re-verify]:

That the top four together account for well under 6% of families is the defining epidemiological fact of AR‑NSID.

2.2 Genetic risk factors

Causal architecture. Loss-of-function (nonsense, frameshift, canonical splice, homozygous whole-gene or partial deletion) predominates; missense variants at deeply conserved residues are the second class. Harripaul et al. [verbatim]: "The new ARID genes include nine with loss-of-function mutations (ABI2, MAPK8, MPDZ, PIDD1, SLAIN1, TBC1D23, TRAPPC6B, UBA7 and USP44), and missense mutations include the first reports of variants in BDNF or TET1 associated with ID." (PMID:28397838)

The dominant modifiable "genetic" risk factor is parental consanguinity — a population-structure risk factor rather than a locus. Jamra [fragment — re-verify]: "the total risk for ID is about 2–3 times higher in children of consanguineous families," with "the prevalence of diagnosable monogenic ID in the children of first cousins or closer [being] two times higher" (PMID:30459488). Couples related as second cousins or closer, and their progeny, account for an estimated 10.4% of the global population (Bittles & Black; PMID:19805052 for the PNAS treatment). Consanguinity rates of 20–50% are reported across North Africa, Central and West Asia, and much of South Asia; Saudi Arabia is often quoted at ~56% overall.

Hu et al. frame the population burden [verbatim]: "Autosomal recessive (AR) gene defects are the leading genetic cause of intellectual disability (ID) in countries with frequent parental consanguinity, which account for about 1/7th of the world population." (PMID:29302074)

Modifier loci / oligogenic burden. Multi-locus causation is documented and non-trivial. Riazuddin et al. [verbatim]: "In another eight families segregation of multiple pathogenic variants was observed, affecting 19 genes that were either known or are novel candidates for ID." (PMID:27457812) — i.e. ~6.6% of their 121 families. This is a real, curatable phenomenon and a candidate for the dismech oligogenic-inheritance pattern (HP:0010983).

Genomic imprinting as a dosage modifier. TRAPPC9 is expressed with a maternal bias (~70%) in brain, and heterozygotes lacking the maternal allele phenocopy homozygous nulls in mouse — see §4.5.

2.3 Environmental risk factors

For AR‑NSID sensu stricto, environmental exposure is not causal. Environmental factors are relevant only as (a) confounders in the differential diagnosis (see §10.5) and (b) potential severity modifiers acting on an established genetic lesion. Note that the class-level ID prevalence literature explicitly attributes part of the mild ID burden to environment: "the prevalence of severe ID and mild ID is variable due to the influence of some environmental stressors" (PMC9946902). No AR‑NSID-specific gene–environment interaction has been established.

2.4 Protective factors

• Genetic: None identified. Notably, common variants in NS‑ARID genes are not associated with normal-range intelligence — Davies/Deary et al. [verbatim]: "Gene-based tests indicated that genes implicated in NS-ARID were not significantly enriched for quantitative trait loci (QTL) associated with intelligence. These findings suggest that genes in which mutations can have a large and deleterious effect on intelligence are not associated with variation across the range of intelligence differences." (PMID:26912939). This is an important negative result: AR‑NSID is not the low tail of the polygenic intelligence distribution.
• Environmental/social: Outbreeding is protective at the population level. Early intervention, enriched education, and family support improve adaptive outcome without altering the molecular lesion.

2.5 Gene–environment interactions

Not established for AR‑NSID. The most defensible statement is negative: AR‑NSID is a high-penetrance monogenic class in which environment modulates adaptive functional outcome rather than disease occurrence.

3. Phenotypes

3.1 The defining phenotype

Diagnostic threshold: "an intelligence quotient (IQ) of 70 or below and a deficit in at least two behaviors associated with adaptive functioning" (PMC9946902). DSM-5 requires deficits in intellectual functioning and adaptive functioning with onset during the developmental period; DSM-5 grades severity by adaptive functioning rather than IQ.

3.2 Severity distribution

Consanguineous-cohort ascertainment (multiplex families, often referred to specialist genetics services) is enriched for moderate-to-severe ID. Basel-Vanagaite reported CC2D1A families with "severe autosomal recessive NSMR" (PMID:16033914) [fragment]; NSUN2 families showed "moderate to severe ID" (PMID:22541559) [fragment].

Curation guidance: frequency bands for severity strata are not supportable from the literature at class level (dismech docs/frequency-evidence-guidelines.md — omit rather than fabricate). Record severity as VARIABLE with a note on ascertainment bias.

3.3 The Monarch-propagated annotation set

Terms propagated to MONDO:0019502 across its 72 descendants (retrieved 2026-08-01; no frequency qualifiers present in the source):

HP:0001263 global developmental delay · HP:0031936 delayed ability to walk · HP:0001249 intellectual disability · HP:0002194 motor delay · HP:0001270 delayed gross motor development · HP:0002187 profound intellectual disability · HP:0001257 spasticity · HP:0010864 severe intellectual disability · HP:0100704 cerebral visual impairment · HP:0000400 macrotia · HP:0000598 abnormality of the outer ear · HP:0000377 abnormal pinna morphology · HP:0000356 abnormality of the external ear

Read this set critically. Spasticity, cerebral visual impairment, and the ear-morphology cluster are inherited from syndromic descendants of the grouping node — they are precisely the features whose presence would disqualify a case from being called non-syndromic. Do not transcribe them into the AR‑NSID entry as class-level phenotypes; they are evidence of the leaky syndromic/non-syndromic boundary, not of the class phenotype.

3.4 Features that are commonly present despite the "non-syndromic" label

This is the honest clinical picture, and it should be curated explicitly with per-gene attribution:

3.5 Neuroimaging

Classic AR‑NSID has an unremarkable or minimally abnormal MRI — this is part of the definition. Documented exceptions with mechanistic value:

• TRAPPC9: "MRI analysis of affected patients shows defects in axonal connectivity" (PMID:20004763) [verbatim]; "Brain magnetic resonance imaging of affected individuals indicates the presence of mild cerebral white matter hypoplasia" (PMID:20004765) [verbatim]. HPO: HP:0012429 cerebral white matter hypoplasia (verify label with OAK before use).
• Mild cerebellar atrophy reported in individual families (PMC9946902).

3.6 Temporal characteristics per phenotype

3.7 Quality of life

No AR‑NSID-specific QoL instrument literature was located. Generic ID QoL evidence applies: adaptive-functioning severity, communication ability, presence of epilepsy, and behavioural comorbidity are the dominant determinants of caregiver burden and individual QoL. Instruments used in the broader ID field: WHOQOL-DIS, Quality of Life Questionnaire (QoL-Q), PedsQL, EQ-5D-Y (proxy-reported). Flag as a genuine evidence gap — a candidate discussions: kind: KNOWLEDGE_GAP entry.

4. Genetic / Molecular Information

4.1 Scale of the gene set

• OMIM PS249500 contains on the order of 60–72 numbered MRT entries; roughly a quarter still have no identified gene. (I was unable to retrieve the OMIM table directly — omim.org returned HTTP 403. Pull the authoritative list from https://www.omim.org/phenotypicSeries/PS249500 before finalizing the entry.)
• Monarch reports 56 causal genes and 61 correlated genes for MONDO:0019502.
• SysID (2018 snapshot): "684 genes that, when mutated, would lead to an ARID form and 378 autosomal recessive candidate genes." SysID 2021: "1500 primary ID genes, causing 1797 ID related disorders, and 1248 ID candidate genes" (all ID, all inheritance modes) — cited in PMID:34930158.
• Upper-bound estimate: 2500–3000 ARID genes (PMID:30459488). The DDD study's independent estimate: "903 ARID genes clarify roughly half of the observed excess of damaging biallelic genotypes."

4.2 Landmark and representative genes (HGNC IDs verified locally via OAK, lowercase hgnc: per repo convention)

Additional Harripaul 2018 LoF genes with HGNC to look up if curated: ABI2, PIDD1, SLAIN1, UBA7, USP44; missense: BDNF, TET1.

4.3 Variant classification, type, and frequency

• Classification: ACMG/AMP. For the ultra-heterogeneous AR‑NSID space the practical bottleneck is PS4/PP1 evidence — most novel genes are supported by one or two families, so many variants sit at "likely pathogenic" or VUS pending GeneMatcher-driven case accrual. ClinGen Gene–Disease Validity classifications exist for only a minority of MRT genes; many remain "Limited" or uncurated. Cite CGGV: structured records where available.
• Type: nonsense, frameshift, canonical ±1/2 splice, homozygous intragenic and whole-gene deletions (e.g. "a homozygous deletion of 170.673 Kb which encompassed the TUSC3 gene" — PMID:21513506 [fragment]), and conserved-residue missense.
• Allele frequency: causal alleles are individually ultra-rare; gnomAD homozygote counts of 0 for a given allele are standard supporting evidence. In consanguineous founder populations specific alleles reach appreciable local carrier frequencies with negligible global frequency — a key filtering trap.
• Origin: germline throughout. Somatic mosaicism is not a recognized mechanism for AR‑NSID.
• Functional consequence: loss of function predominates, consistent with recessive inheritance. Gain-of-function and dominant-negative mechanisms are not characteristic. Hypomorphic missense alleles explain part of the severity spectrum within a gene.

4.4 Support for pathogenicity beyond segregation

Riazuddin et al. give the canonical multi-modal argument [verbatim]:

"Transcriptome profiles of normal human brain tissues showed that the novel candidate ID genes formed a network significantly enriched for transcriptional co-expression (P<0.0001) in the frontal cortex during fetal development and in the temporal-parietal and sub-cortex during infancy through adulthood. In addition, proteins encoded by 12 novel ID genes directly interact with previously reported ID proteins in six known pathways essential for cognitive function (P<0.0001)." (PMID:27457812)

Harripaul et al. add [verbatim]: "The genes identified also showed overlap with de novo gene sets for other neuropsychiatric disorders. Transcriptional studies showed prominent expression in the prenatal brain." (PMID:28397838)

4.5 Epigenetics and imprinting

The best-characterized epigenetic dimension is parent-of-origin allelic bias at TRAPPC9 — Wang et al. [verbatim]:

"In an analysis of brain-specific allele-biased expression, we identified that Trappc9, a cellular trafficking factor, was expressed predominantly (~70%) from the maternally inherited allele. … Strikingly, heterozygous mice lacking the maternal allele (70% reduced expression) had pathology similar to homozygous mutants, whereas mice lacking the paternal allele (30% reduction) were phenotypically normal." (PMID:32877400)

This has a direct clinical corollary: for an imprinted-bias ARID gene, a monoallelic maternally-inherited LoF variant may be pathogenic — a genotype that standard recessive filtering would discard.

Separately, TET1 (5mC→5hmC dioxygenase) appearing as a novel ARID gene (PMID:28397838) places DNA demethylation machinery inside the causal set, and differential methylation at TRAPPC9 has been reported in severe childhood obesity.

4.6 Chromosomal abnormalities

Not the primary mechanism — but homozygous CNVs are a genuine and recurrent cause, which is why CMA remains a first-tier test even in a suspected-recessive workflow. Harripaul et al. explicitly combined "microarray genotyping, homozygosity-by-descent (HBD) mapping, copy number variation (CNV) analysis, and whole exome sequencing" and reported "definite or candidate mutations (or CNVs) in 51% of families" (PMID:28397838) [verbatim]. Anwar et al. found "copy number variants in 14% (n=54, 15% are novel)" of their 337-subject ID cohort (PMID:27431290) [verbatim].

5. Environmental Information

Not applicable as an etiologic category. No toxin, occupational exposure, radiation source, lifestyle factor, or infectious agent causes AR‑NSID.

Environmental factors matter in exactly three ways, all of which belong in the differential-diagnosis and prevention sections rather than etiology:

1. Diagnostic confounding. Prenatal alcohol exposure, congenital infection (CMV, Zika, rubella, toxoplasmosis), perinatal hypoxic-ischaemic injury, lead and other heavy-metal exposure, and severe early psychosocial deprivation all produce static ID and must be excluded — especially in consanguineous populations where the prior for a genetic cause is high and a coincidental environmental cause can be missed.
2. Severity modulation. Iodine deficiency, malnutrition, and educational deprivation act additively on adaptive outcome.
3. Population risk structure. Consanguinity is a sociocultural variable with a genetic consequence — arguably the only "environmental" factor with real effect on AR‑NSID incidence.

6. Mechanism / Pathophysiology

6.1 The class-level causal chain

There is no single AR‑NSID mechanism. The defensible class-level model is a convergence architecture:

Biallelic LoF in one of ~2500 autosomal genes  [MOLECULAR]
↓
Loss of a gene-specific molecular function required by developing neurons
(synaptic proteolysis | vesicle trafficking | glycosylation | tRNA/protein
 modification | transcription/Mediator | signalling | metabolism)  [MOLECULAR]
↓
Perturbed neurodevelopmental cellular process
(neurogenesis, neurite outgrowth/branching, synaptogenesis,
 synaptic plasticity, myelination/axonal connectivity)  [CELLULAR]
↓
Altered cortical/hippocampal circuit assembly and function  [TISSUE]
↓
Impaired learning, memory, and adaptive behaviour = intellectual disability  [ORGANISM]

Suggested GO terms (all verified locally via OAK):

⚠️ GO:0006486 ("protein glycosylation") is obsolete in the current GO release — use GO:0006487 (protein N-linked glycosylation). Confirmed via OAK.

Suggested CL terms (verified): CL:0000540 neuron · CL:0000679 glutamatergic neuron · CL:0000617 GABAergic neuron · CL:0000047 neural stem cell · CL:0000127 astrocyte · CL:0000128 oligodendrocyte · CL:0002319 neural cell.

Suggested UBERON terms (verified): UBERON:0000955 brain · UBERON:0000956 cerebral cortex · UBERON:0002421 hippocampal formation · UBERON:0001954 Ammon's horn · UBERON:0002316 white matter · UBERON:0002037 cerebellum.

6.2 Worked mechanism 1 — PRSS12 / neurotrypsin: synaptic proteolysis

Molinari et al. [verbatim]:

"A 4-base pair deletion in the neuronal serine protease neurotrypsin gene was associated with autosomal recessive nonsyndromic mental retardation (MR). In situ hybridization experiments on human fetal brains showed that neurotrypsin was highly expressed in brain structures involved in learning and memory. Immuno-electron microscopy on adult human brain sections revealed that neurotrypsin is located in presynaptic nerve endings, particularly over the presynaptic membrane lining the synaptic cleft. These findings suggest that neurotrypsin-mediated proteolysis is required for normal synaptic function and suggest potential insights into the pathophysiological bases of mental retardation." (PMID:12459588)

Chain: loss of presynaptic neurotrypsin → failure to cleave agrin at the synaptic cleft → loss of the C-terminal agrin-22 fragment that drives dendritic filopodia formation → impaired activity-dependent synaptic remodelling → deficient learning and memory. Scale tags: MOLECULAR → CELLULAR → ORGANISM. Cell type CL:0000540; site UBERON:0002421.

6.3 Worked mechanism 2 — CC2D1A: dual NF‑κB and cAMP–PKA–PDE4D dysregulation

Basel-Vanagaite et al. identified "a protein truncating mutation … in the gene CC2D1A in nine consanguineous families with severe autosomal recessive NSMR," encoding "a putative signal transducer participating in positive regulation of I-kappaB kinase/NFkappaB cascade," expressed most highly "in the cerebral cortex and hippocampus" (PMID:16033914) [fragment — re-verify].

The better-resolved arm is the cAMP branch. CC2D1A co-localizes with PDE4D and, on cAMP stimulation, escorts it to the cell periphery; the human deletion allele (lacking three of four DM14 domains plus the adjacent C2 domain) abolishes this translocation, causing constitutive PKA phosphorylation of PDE4D at Ser126 and PDE4D hyperactivity → excessive cAMP hydrolysis → reduced CREB phosphorylation → memory and social-behaviour deficits. In Cc2d1a KO mice, the PDE4 inhibitor rolipram (CHEBI:104872, verified) rescues spatial memory — in males only (PMID:30732858; see §15.3). Structural work on the human CC2D1A fragment associated with NSID was published in Bioscience Reports 2026 (BSR20253955).

This is the single best druggable-mechanism exemplar in AR‑NSID and the strongest candidate for a treatments entry with target_mechanisms.

6.4 Worked mechanism 3 — TRAPPC9: trafficking + NF‑κB + neurite outgrowth

Both 2009 discovery papers converge [verbatim]:

"Sequence analysis of genes in the candidate interval identified a nonsense nucleotide change in the gene that encodes TRAPPC9 (trafficking protein particle complex 9, also known as NIBP), which has been implicated in NF-kappaB activation and possibly in intracellular protein trafficking. TRAPPC9 is highly expressed in the postmitotic neurons of the cerebral cortex, and MRI analysis of affected patients shows defects in axonal connectivity." (PMID:20004763)

"This gene encodes NIK- and IKK-beta-binding protein (NIBP), which is involved in the NF-kappaB signaling pathway and directly interacts with IKK-beta and MAP3K14." (PMID:20004765)

Downstream cell biology (from the model literature): TRAPPC9 partners TRAPPC10 to direct TRAPPII toward Rab11 activation, governing recycling-endosome traffic; deficiency impairs neurite elongation and branching in both zebrafish and mouse (Int J Biol Sci 2023, PMC10321293), causes disproportionate hippocampal volume loss with Sox2⁺ neural stem/progenitor cell depletion and neuronal lipid-droplet accumulation (bioRxiv 2023), and produces a dopamine D1/D2 neuron imbalance underlying the learning/memory deficit (PMID:33208359). The obesity arm is discussed in §15.2.

Chain: TRAPPC9 LoF → impaired TRAPPII/Rab11 endosomal recycling and reduced NIK/IKKβ-dependent NF‑κB activation → impaired neurite elongation/branching + NSPC depletion → reduced cortical/hippocampal volume, white matter hypoplasia, axonal connectivity defects → ID ± postnatal microcephaly ± obesity.

6.5 Recurrent mechanistic themes across the AR‑NSID gene set

Important: these are post-hoc groupings, not a claim of pathway convergence. Jamra's explicit finding is the opposite — no prevalent pathways or complexes. Curate these as organizing themes with that caveat attached.

6.6 Immune, metabolic, and tissue-damage mechanisms

• Immune: No autoimmunity or immunodeficiency. NF‑κB appears (CC2D1A, TRAPPC9) as a neurodevelopmental signalling node, not an inflammatory one. Do not curate as immune-mediated disease.
• Metabolic: Only in specific genes (TECR — VLCFA elongation; ADK — adenosine/methionine metabolism; IMPA1 — inositol recycling). No class-level metabolic signature. Absence of a metabolic abnormality is part of the non-syndromic definition.
• Tissue damage: None. AR‑NSID is a developmental disorder, not a degenerative one — no oxidative-stress, ischaemia, fibrosis, or necrosis mechanism. The pathology is a mis-built circuit, not a damaged one. This is a load-bearing distinction for the dismech pathograph.

6.7 Molecular profiling

• Transcriptomics: the strongest class-level datum is prenatal-brain co-expression enrichment of AR‑NSID genes (PMID:27457812; PMID:28397838). BrainSpan/GTEx/Allen Brain Atlas are the reference resources.
• Proteomics / metabolomics / lipidomics: no AR‑NSID class-level signature. Gene-specific exceptions: transferrin isoelectric focusing / glycan mass-spec abnormalities in MAN1B1 and TUSC3 (CDG-like); VLCFA profiling in TECR.
• Single-cell / spatial: no AR‑NSID-specific atlas. Human Cell Atlas and developing-brain scRNA-seq resources are used to establish cell-type expression for candidate genes.
• Functional genomics: no AR‑NSID-focused CRISPR screen. DepMap and MorPhiC (see §15.5) are the relevant resources; MorPhiC's iPSC-derived null-allele phenotyping is directly applicable and would license category: Cellular phenotypes with evidence_source: IN_VITRO.

7. Anatomical Structures Affected

Lateralization: bilateral and symmetric throughout — a diffuse developmental process. Focal or asymmetric imaging findings argue against AR‑NSID.

Secondary organ involvement: none by definition. Where present (obesity in TRAPPC9, retinal disease in AHI1 WD40 variants) the case is reclassified as syndromic or as an expanded-phenotype allelic disorder.

8. Temporal Development

• Onset: the molecular lesion is present from conception; the clinical phenotype is congenital. Ascertainment typically occurs at 1–4 years for moderate/severe ID (motor and speech milestones) and at school age for mild ID. Onset pattern is insidious/chronic, never acute.
• Progression: static (non-progressive) encephalopathy. The functional gap versus peers widens with age because the developmental trajectory is shallower — this is apparent rather than true progression, and the distinction matters clinically and for curation.
• Stages: no formal staging system exists. Practical framing: (i) infancy — global developmental delay, HP:0001263; (ii) preschool — speech delay dominant; (iii) school age — formal IQ/adaptive testing establishes ID and severity; (iv) adolescence/adulthood — plateau at an adaptive ceiling, with transition-of-care and supported-living needs.
• Duration: lifelong.
• Remission: none. Spontaneous or treatment-induced remission does not occur.
• Critical periods: (a) prenatal/perinatal — the window in which the gene acts, hence the target for hypothetical molecular therapy; (b) birth to ~5 years — the window of maximal neuroplasticity and the evidence-supported window for early intervention; (c) preconception — the only window for effective prevention (§13).

9. Inheritance and Population

9.1 Epidemiology

Derived class-level estimate for dismech Prevalence, to be recorded with explicit assumptions: if ID prevalence is ~1–2% and ARID is ~10% of diagnosable ID in outbred settings, class prevalence is on the order of 100–200 per 100,000 in outbred populations (prevalence_class: ABOVE_1_IN_1000), and materially higher in highly consanguineous populations. The non-syndromic subset is a fraction of that and cannot be reliably separated with current data. Record measure_type: POINT_PREVALENCE, put the derivation in notes, and do not overstate precision.

9.2 Inheritance mechanics

• Pattern: autosomal recessive, HP:0000007 (verified). Recurrence risk 25% per pregnancy for carrier×carrier couples. Hu et al.: "For autosomal recessive ID (ARID) the recurrence risk is high" (PMID:24176302) [verbatim].
• Multi-locus inheritance: documented in ~6.6% of families (PMID:27457812). Where curated, use HP:0010983 oligogenic inheritance with the genes named in the block description (the Inheritance class has no genes slot) — see the repo's digenic/oligogenic section and the PRPH2-Related_Retinopathy exemplar.
• Penetrance: essentially complete for established LoF genotypes in confirmed genes. Reduced penetrance has not been systematically studied and remains a genuine gap.
• Expressivity: highly variable, both within and between families, and this is arguably the field's central unsolved problem. Jamra [fragment — re-verify]: "there is … accumulating evidence that the spectrum of symptoms due to bi-allelic mutations in established ARID genes may vary enormously" — with the striking example that "compound heterozygous pathogenic variants in the WD40 domain of AHI1 … have been reported to lead to an isolated retinitis pigmentosa" rather than Joubert syndrome (PMID:30459488).
• Anticipation: not applicable — no repeat-expansion mechanism.
• Germline mosaicism: not a recognized feature; irrelevant to recurrence counselling in the recessive setting.
• Founder effects: important and population-specific (Iranian, Pakistani, Arab, Turkish sub-populations). MAN1B1 is a partial exception to the no-recurrence rule — "MAN1B1 is one of the few NS-ARID genes with an elevated mutation frequency in patients with NS-ARID from different populations" (PMID:21763484) [fragment].
• Carrier frequency: no class-level figure is meaningful. Per-gene carrier frequencies are ultra-low globally and locally elevated in founder populations.

9.3 Population demographics

• Highest burden: North Africa, the Middle East, West/Central/South Asia — the consanguinity belt. Study cohorts are overwhelmingly Iranian (Najmabadi/Hu/Kahrizi), Pakistani (Riazuddin, Harripaul, Mir), Saudi/Gulf Arab (Anwar/Alkuraya), and Turkish.
• Outbred populations: AR‑NSID is not rare there either — this is the central corrective message of Musante & Ropers [verbatim]: "Here we review recent progress in this field, show that ARID is not rare even in outbred Western populations, and discuss the prospects for improving its diagnosis and prevention." (PMID:24176302). Harripaul et al. concur [verbatim]: "As with other AR disorders, the relevance will also apply to outbred populations." (PMID:28397838)
• Sex ratio: ~1:1 expected for autosomal recessive inheritance. (Contrast X-linked ID, which is male-biased — the male excess in overall ID cohorts is largely XLID plus ascertainment.) Caveat worth curating: at least one AR‑NSID gene shows sex-dimorphic biology in the model system — Cc2d1a cognitive rescue by PDE4 inhibition is male-specific (§15.3) — and Trappc9 KO obesity is "significantly more severe in females than in males."
• Age distribution: all ages; lifelong condition; prevalence in adults reflects the full lifetime cohort.

9.4 Ascertainment bias — read every number above with this in mind

The AR‑NSID evidence base is derived almost exclusively from multiplex consanguineous families with moderate-to-severe ID. This design systematically over-represents severe phenotypes and LoF alleles, under-represents mild ID and hypomorphic alleles, and under-represents compound heterozygosity (which is the dominant recessive mechanism in outbred populations). Any frequency, severity, or prevalence claim curated from this literature should carry that caveat.

10. Diagnostics

10.1 Genetic testing — the diagnostic backbone

The consensus is unambiguous. Jamra: "a genetic diagnosis of ID in general, and ARID specifically, is better made using large panels or exome sequencing" (PMID:30459488) [fragment], and "there is no justification for a gene-specific or panel diagnostic when ARID is suspected." Anwar et al. conclude [verbatim]: "Our results expand the morbid genome of ID and support the adoption of genomics as a first-tier test for individuals with ID."

Diagnostic yields (all figures verbatim from the cited abstracts unless noted):

Note the isolated-vs-syndromic split (31% vs 53%): non-syndromic ID has a lower ES yield than syndromic ID. This is a curatable, decision-relevant number.

Recommended workflow: 1. Chromosomal microarray — first-tier; detects homozygous deletions and the 14% CNV fraction. Karyotype only for suspected balanced rearrangement or aneuploidy; FISH only for targeted confirmation. 2. Fragile X (FMR1) repeat testing — still first-tier in most guidelines; 1% of the Anwar cohort. Not AR, but a mandatory exclusion. 3. Trio exome sequencing — the workhorse. Trio design is essential: it distinguishes de novo dominant from recessive, and Kahrizi et al. showed that de novo mutations occur in consanguineous families too, which homozygosity-mapping-only designs systematically miss (Clin Genet 2019;95(1):151–159). 4. Genome sequencing — reflex for ES-negative cases; adds deep-intronic, structural, and repeat variants. 5. Homozygosity/autozygosity mapping — high-value adjunct in consanguineous pedigrees; ROH data come free from CMA or can be derived from ES/GS. 6. RNA sequencing on an accessible tissue — reflex for ES/GS-negative cases; resolves splice-altering variants of uncertain significance. 7. Reanalysis at 18–24 months — the DDD experience (27% → 40% on reanalysis) makes this the highest-yield, lowest-cost intervention in the whole pathway.

Gene panels: large "intellectual disability" / neurodevelopmental panels (e.g. Genomics England PanelApp Intellectual disability – microarray and sequencing) are widely used but are explicitly not recommended over ES for suspected ARID because the gene set is too large and too incomplete. Mitochondrial DNA testing and repeat-expansion panels are relevant only for specific differentials, not for AR‑NSID.

10.2 Laboratory and biochemical testing

There are no diagnostic biomarkers for AR‑NSID as a class — the absence of biochemical abnormality is part of the definition. Metabolic testing is performed to exclude treatable IEMs (see §10.5): plasma amino acids, urine organic acids, acylcarnitine profile, ammonia, lactate, homocysteine, copper/ceruloplasmin, creatine metabolites, biotinidase, CDG transferrin isoelectric focusing, thyroid function, lead level.

Gene-specific exceptions worth curating: abnormal transferrin glycoforms in MAN1B1/TUSC3; VLCFA abnormalities in TECR.

10.3 Imaging and electrophysiology

• Brain MRI: normal or near-normal by definition; performed to exclude malformations of cortical development, leukodystrophy, and acquired injury. Positive findings (white-matter hypoplasia, mild cerebellar atrophy) exist for specific genes.
• EEG: indicated only if seizures are suspected.
• Nerve conduction / EMG, ECG: not indicated absent specific features; their necessity would itself argue against a non-syndromic classification.
• Biopsy / histopathology: not indicated. No characteristic histopathology exists.

10.4 Clinical diagnostic criteria

DSM-5-TR, Intellectual Developmental Disorder (317–319 / F70–F73) — all three required: - (A) deficits in intellectual functions confirmed by clinical assessment and standardized testing; - (B) deficits in adaptive functioning across conceptual, social, and practical domains; - (C) onset during the developmental period. Severity is specified by adaptive functioning, not IQ.

ICD-11 6A00 grades by both intellectual functioning and adaptive behaviour (~2 SD below mean; 6A00.0 mild → 6A00.3 profound; 6A00.4 provisional).

"Non-syndromic" designation is a clinical judgement, not a criterion set: ID without a consistent pattern of dysmorphism, congenital malformation, neurological signs, growth abnormality, metabolic derangement, or characteristic neuroimaging. It should always be recorded as provisional pending deep phenotyping and follow-up.

Standard instruments: WISC-V / WPPSI-IV / WAIS-IV / Stanford-Binet-5 (or Bayley-4 / Mullen in infancy); Vineland-3 or ABAS-3 for adaptive behaviour.

10.5 Differential diagnosis

10.6 Screening

• Newborn screening: does not detect AR‑NSID (no biochemical marker). It does detect several key differentials (PKU, congenital hypothyroidism, biotinidase deficiency) — an important negative to state.
• Carrier screening: the operative modality. Expanded carrier screening panels include a growing but incomplete subset of ARID genes; in consanguineous couples, couple-based exome/genome carrier screening outperforms fixed panels because the relevant allele is often population- or family-private.
• Cascade screening: offer to at-risk relatives once the family variant is known — high-value in extended consanguineous kindreds.

11. Outcome / Prognosis

• Survival: for AR‑NSID sensu stricto, life expectancy is near-normal or modestly reduced. Excess mortality in the ID population generally is driven by epilepsy, aspiration, immobility, and reduced access to healthcare — largely features of syndromic ID. No AR‑NSID-specific survival data were located. Genuine evidence gap.
• Mortality: no disease-specific mortality data at class level. Do not fabricate a figure.
• Morbidity and disability: lifelong cognitive and adaptive impairment is the dominant morbidity. GBD captures "idiopathic developmental intellectual disability" as a YLD-generating cause; ICF is the appropriate functional-classification framework.
• Disease course: static; no recovery of the underlying deficit. Adaptive function improves with intervention, education, and support — the developmental trajectory is shallower, not descending.
• Complications: behavioural and psychiatric comorbidity (ADHD, autism, anxiety, self-injury), epilepsy in a subset, communication-related social isolation, and dependency needs in adulthood.
• Prognostic factors: severity of ID at diagnosis; expressive language attainment by age 5 (the strongest practical predictor of adaptive outcome in the broader ID literature); presence of epilepsy; the specific gene and allele; access to early intervention and educational support; family resources.
• Prognostic biomarkers: none. The molecular diagnosis itself is the best available prognostic instrument, via gene-specific natural-history data where it exists.

12. Treatment

There is no disease-modifying therapy for AR‑NSID. Management is supportive, habilitative, and educational, plus targeted treatment of comorbidities. State this plainly in the entry.

12.1 Supportive and rehabilitative care (the mainstay)

Per the repo's mechanical-backfill table, NCIT:C15302, NCIT:C159273, NCIT:C121351, and NCIT:C181743 all map to therapeutic_modality: BEHAVIORAL.

12.2 Pharmacotherapy — symptomatic only

No drug treats ID itself. Comorbidity-directed agents: stimulants and alpha-2 agonists for ADHD; SSRIs for anxiety/OCD; atypical antipsychotics (risperidone, aripiprazole) for severe irritability/aggression; anti-seizure medication where epilepsy is present; melatonin for sleep. Use treatment_term: NCIT:C15986 Pharmacotherapy with a specific therapeutic_agent (CHEBI for small molecules).

12.3 The one mechanism-targeted lead: PDE4 inhibition in CC2D1A

Preclinical (mouse) only — must be curated with evidence_source: MODEL_ORGANISM and must not be presented as a human therapy. In Cc2d1a-deficient mice, PDE4D hyperactivity depletes cAMP and impairs CREB signalling; the PDE4 inhibitor rolipram (CHEBI:104872, verified) rescues spatial-memory deficits — in males only (PMID:30732858, Biol Psychiatry). This is the closest thing AR‑NSID has to a druggable node and is a strong candidate for a target_mechanisms treatment edge pointing at the "PDE4D hyperactivity" pathophysiology node with INHIBITS.

A second, weaker lead: chronic pharmacologic manipulation of dopamine transmission ameliorates the metabolic disturbance in Trappc9-linked syndrome in mice (PMC11383600, 2024) — metabolic, not cognitive, rescue.

12.4 Advanced therapeutics

• Gene therapy / gene editing / ASO / siRNA: none in development for any AR‑NSID gene, and the barriers are structural, not merely technical — (i) ~2500 candidate genes make per-gene development economically infeasible; (ii) the causal window is prenatal/early-postnatal, before diagnosis; (iii) the pathology is a mis-built circuit, not an ongoing degenerative process, so post-hoc restoration of gene function may not restore function. Recessive LoF biology is, in principle, gene-replacement-friendly — the timing problem is the real obstacle.
• Cell therapy, immunotherapy, targeted therapy: not applicable.
• Surgery: not applicable to the ID itself.

12.5 Clinical trials

No AR‑NSID-specific interventional trial was identified. One observational study of relevance: NCT06706934 — Search for Phenotype-modifying Genes in Patients With Intellectual Disabilities (verify status, phase, and sponsor on ClinicalTrials.gov, and fetch with just fetch-reference NCT06706934 before curating).

12.6 Pharmacogenomics

No AR‑NSID-specific PGx. Standard CPIC guidance applies to the psychotropics used for comorbidities (CYP2D6/CYP2C19 for SSRIs and atomoxetine; HLA-B*15:02 for carbamazepine in relevant ancestries).

13. Prevention

Prevention is where AR‑NSID has the most actionable evidence — and it is entirely preconception/reproductive rather than therapeutic. Musante & Ropers frame the field's ambition as improving "its diagnosis and prevention" (PMID:24176302) [verbatim].

• Primary prevention. Genetic counselling about consanguinity risk; preconception and premarital carrier screening; population health education. National Premarital Screening and Genetic Counseling (PMSGC) programmes operate in Saudi Arabia and several Gulf and Middle Eastern states; community survey data show 84.4% of respondents correctly identify consanguinity as increasing autosomal recessive risk and 89.3% recognize that genetic counselling reduces family recurrence (Front Genet 2026;1866894). Note the ethical framing: the objective is informed reproductive choice, not discouragement of consanguineous marriage — this should be stated explicitly in the entry.
• Secondary prevention. For a couple with an affected child and a known molecular diagnosis: prenatal diagnosis (CVS/amniocentesis) or preimplantation genetic testing for monogenic disease (PGT-M), with 25% recurrence risk counselling. Cascade carrier testing across the extended kindred. Early developmental screening for at-risk siblings enables intervention at the point of maximal plasticity.
• Tertiary prevention. Early intervention services; epilepsy control; treatment of hearing and vision impairment (which compound the cognitive deficit); behavioural support; nutritional management (relevant for TRAPPC9-associated obesity); transition planning.
• Immunization: no AR‑NSID-specific vaccine. Routine immunization prevents ID-causing congenital infections (rubella) and is a legitimate public-health adjacency.
• Newborn screening: does not detect AR‑NSID (§10.6).

14. Other Species / Natural Disease

• Taxonomy of relevance: Homo sapiens NCBITaxon:9606. Experimental orthologs: Mus musculus NCBITaxon:10090, Danio rerio NCBITaxon:7955, Drosophila melanogaster NCBITaxon:7227, Caenorhabditis elegans NCBITaxon:6239.
• Naturally occurring disease in other species: None recognized. AR‑NSID is a human-specific nosological construct. Intellectual disability requires a construct (IQ, adaptive functioning) that has no cross-species equivalent, so no OMIA entry corresponds to AR‑NSID. I found no OMIA entry for canine or other companion-animal inherited cognitive impairment matching this class. Curate this section as "not applicable" rather than stretching for an analogue.
• Breed (VBO): not applicable.
• Comparative biology. Evolutionary conservation of the molecular machinery is well demonstrated even though the disease is not: the Drosophila NSUN2 ortholog deletion produces "severe short-term-memory (STM) deficits" rescuable by wild-type re-expression, indicating "an evolutionarily conserved role of RNA methylation in normal cognitive development" (PMID:22541559) [fragment]. The PRSS12/neurotrypsin ortholog tequila regulates long-term memory formation in Drosophila. Trappc9 deficiency impairs neurite elongation and branching in both zebrafish and mice (PMC10321293).
• Zoonotic potential / cross-species transmission: not applicable (non-infectious, non-transmissible).

15. Model Organisms

15.1 Overview

There is no model of "AR‑NSID" — only gene-specific models. Recapitulation is judged by learning/memory assays plus brain morphometry, not by a cognitive construct.

15.2 Trappc9 — the best-characterized model

Wang et al., PLoS Genetics 2020 [verbatim]:

"By studying Trappc9 null mice we discovered that homozygous mutant mice showed a reduction in brain size, exploratory activity and social memory, as well as a marked increase in body weight. A role for Trappc9 in energy balance was further supported by increased ad libitum food intake in a child with TRAPPC9 deficiency. … Taken together, we conclude that Trappc9 deficient mice recapitulate key pathological features of TRAPPC9 mutations in humans and identify a role for Trappc9 and its imprinting in controlling brain development and metabolism." (PMID:32877400)

Additional findings across the Trappc9 model literature: - Cognitive/memory/learning impairment across Morris water maze, Barnes maze, and social learning. - Disproportionate hippocampal volume reduction with loss of Sox2⁺ neural stem/progenitor cells and neuronal lipid-droplet accumulation (bioRxiv 2023.11.20.567859 — preprint, not peer-reviewed; flag as such). - DTI shows reduced white-matter organization/integrity; high-resolution MRI shows multiple regions of reduced volume (ISMRM 2022 abstract — conference abstract, low evidence tier). - Dopamine D1/D2 neuron imbalance as the proximate cause of learning/memory deficits (PMID:33208359). - Obesity with hyperinsulinemia, glucose intolerance, and raised plasma lipids, more severe in females. - Chronic dopaminergic pharmacologic manipulation ameliorates the metabolic disturbance (PMC11383600, 2024). - Zebrafish trappc9 knockdown/knockout: defective neurite elongation and branching (PMC10321293).

Fidelity: good for microcephaly, obesity, and learning/memory. Limitation: the mouse cannot model ID as a construct; and the parent-of-origin effect adds a layer whose human relevance is only partially established (the increased ad libitum food intake in one child is suggestive but n=1).

15.3 Cc2d1a

Constitutive Cc2d1a KO is perinatally lethal, so conditional (cortex/hippocampus) deletion is the workhorse. Phenotype: cognitive and social deficits, hyperactivity, anxiety, and self-injury in males. Mechanism: PDE4D hyperactivity → cAMP depletion → reduced CREB signalling. Rescue: PDE4 inhibition restores spatial memory in males but has no effect in females (PMID:30732858). A companion study dissected Cc2d1a vs its homolog Cc2d1b, which "differentially affect spatial memory, anxiety, and hyperactivity" (PMC5840150, Front Genet 2018). The sex-specificity is a major translational caveat and should be curated explicitly.

15.4 Other gene-specific models

15.5 Model types and resources

• Genetic model types: knockout, conditional (Cre-lox), knock-in of patient missense alleles, humanized alleles, transgenic rescue.
• Cellular models: patient-derived iPSCs and iPSC-derived neurons/cortical organoids — the most human-relevant platform, and the one that can address the cell-autonomy and cell-type-specificity questions the mouse cannot. MorPhiC (null alleles of human genes in iPSC-derived multicellular systems) is directly applicable; MorPhiC-derived phenotypes should be curated as category: Cellular with evidence_source: IN_VITRO.
• Databases: MGI, IMPC, KOMP, IMSR, EMMA, MMRRC (mouse); ZFIN (zebrafish); FlyBase; WormBase; Alliance of Genome Resources; Cellosaurus (lines).

15.6 Shared limitations (candidates for kind: HUMAN_MODEL_MISMATCH)

Per the repo's guidance, these are not generic KNOWLEDGE_GAP items — evidence exists in the model, but translational validity is the open question:

1. ID is not modellable. Rodent learning/memory assays are proxies for a construct (IQ + adaptive functioning) with no animal equivalent.
2. Human-specific cortical biology. Outer radial glia and OSVZ-driven cortical expansion are absent or minimal in mouse — directly relevant to the microcephaly-adjacent AR‑NSID genes.
3. Sex-dimorphic rescue. Cc2d1a PDE4-inhibitor rescue works in males only; the human correlate is unknown.
4. Imprinting divergence. Parent-of-origin Trappc9 bias is established in mouse brain; the extent of the same bias in human brain, and its clinical consequence for maternal-allele heterozygotes, is unresolved.
5. Background and allele-type effects. Most models are null alleles on inbred backgrounds; most human alleles are missense/hypomorphic on outbred backgrounds.

16. Curation guidance for the dismech entry

A few structural recommendations, offered because they materially affect how this entry should be built:

1. Consider a kb/groupings/ entity, not (only) a Disease. AR‑NSID is a MONDO grouping node with 72 descendants and ~56 causal genes and no shared mechanism. It maps cleanly onto the Grouping class: grouping_basis: [SHARED_PHENOTYPE, OTHER] (shared phenotype + shared inheritance mode), with a NECESSARY membership_criteria block combining HAS_PHENOTYPE HP:0001249, HAS_INHERITANCE (autosomal recessive), and a negated leaf for syndromic features. That structure is honest about what the class is: a union defined by a phenotype plus an inheritance mode plus a negation, not a mechanism. The Digenic_and_Oligogenic_Disorders grouping is a close precedent for a criteria-defined union with a HAS_INHERITANCE criterion.

2. If a Disease entry is retained, keep the pathophysiology graph at the architecture level (§6.1) with two or three fully-evidenced worked mechanisms (PRSS12, CC2D1A, TRAPPC9) as exemplars — do not attempt 56 gene-specific chains. Use biological_scale tags per the four-value enum; the chain in §6.1 is already atomized to one scale per node.

3. Do not transcribe the Monarch-propagated HPO set (§3.3). Spasticity, cerebral visual impairment, and the ear-morphology cluster are inherited from syndromic descendants and contradict the class definition.

4. NEC preflight is low-risk here but not zero. "Non-syndromic intellectual disability" has both AD and AR forms with parallel numbered series (MRD/MRT), and MONDO carries both MONDO:0019502 (AR) and an AD counterpart. Any deep-research report used for this entry should be checked for AD/AR series confusion — this falls squarely in the "numbered series" high-NEC-risk class.

5. A curious MONDO artifact worth surfacing rather than propagating: MONDO:0019502 asserts is_a MONDO:0017706 (disorder of carbohydrate transmembrane transport and absorption). That parentage looks like an ontology defect — presumably leaked from a glycosylation-related descendant (TUSC3, MAN1B1, ST3GAL3) — and should not be reflected in the dismech classification. Consider filing it upstream with MONDO.

6. Evidence gaps worth curating as discussions: (a) KNOWLEDGE_GAP — no AR‑NSID-specific survival, mortality, or QoL data; (b) KNOWLEDGE_GAP — penetrance and expressivity systematically unstudied; (c) HUMAN_MODEL_MISMATCH — the five items in §15.6.

7. Every snippet marked [fragment] above must be re-fetched with just fetch-reference and validated before it enters a YAML file. The fully-verbatim set (PMIDs 21937992, 28397838, 27431290, 24176302, 27457812, 29302074, 12459588, 20004763, 20004765, 32877400, 26912939, 34930158) is the safest starting pool.

Reference list

Verbatim-retrieved abstracts (safe snippet pool):

Fragment-retrieved — re-verify before quoting: 30459488 (Abou Jamra R. Genetics of autosomal recessive intellectual disability. Med Genet. 2018;30(3):323-327) · 18452889 (Garshasbi M, et al. Am J Hum Genet. 2008;82(5):1158-64, TUSC3) · 21763484 (Rafiq MA, et al. Am J Hum Genet. 2011;89(1):176-82, MAN1B1) · 21513506 (Khan MA, et al. BMC Med Genet. 2011;12:56, TUSC3) · 22541559 (Khan MA, et al. Am J Hum Genet. 2012, NSUN2) · 22541562 (Abbasi-Moheb L, et al. Am J Hum Genet. 2012, NSUN2) · 16033914 (Basel-Vanagaite L, et al. J Med Genet. 2006, CC2D1A) · 30732858 (Biol Psychiatry, Cc2d1a/PDE4D male-specific rescue) · 33208359 (Trappc9 D1/D2 imbalance) · 19805052 (Bittles & Black, PNAS, consanguinity) · Kahrizi K, et al. Clin Genet. 2019;95(1):151-159 (trio sequencing in consanguineous families) · Ghalamkari S, et al. Am J Med Genet A. 2025 (proband-only ES, Iran).

Databases consulted: MONDO (via OAK sqlite:obo:mondo), HPO / GO / CL / UBERON / NCIT / CHEBI / HGNC (via OAK, all terms in this report verified), Monarch Initiative API v3, MedGen, OMIM PS249500 (inaccessible — HTTP 403; retrieve directly), Orphanet ORPHA:88616 (inaccessible — bot protection; retrieve directly or via just structured-rebuild-orphanet --id 88616), SysID, ClinicalTrials.gov, PubMed/E-utilities.

Sources: - Deep sequencing reveals 50 novel genes for recessive cognitive disorders — PubMed - Mapping autosomal recessive intellectual disability: 26 novel candidate genes in 192 consanguineous families — PubMed - Genetics of autosomal recessive intellectual disability — PMC - Genetics of intellectual disability in consanguineous families — PubMed - Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes — Mol Psychiatry - Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity — PubMed - Male-Specific cAMP Signaling in the Hippocampus Controls Spatial Memory Deficits — PubMed - Loss of Cc2d1a and Cc2d1b Differentially Affect Spatial Memory, Anxiety, Hyperactivity — PMC - Defective neurite elongation and branching in Nibp/Trappc9 deficient zebrafish and mice — PMC - Intellectual disability genomics: current state, pitfalls and future challenges — PMC - Examining NS-ARID genes for an enriched association with intelligence differences — PMC - Non-syndromic Intellectual Disability: An Experimental In-Depth Exploration — PMC - Phosphodiesterase activity is regulated by CC2D1A — PMC - Consanguinity, human evolution, and complex diseases — PNAS - Proband-Only Exome Sequencing for Intellectual Disability in Iran — AJMG A 2025 - Effect of inbreeding on intellectual disability revisited by trio sequencing — Clin Genet - Chronic pharmacologic manipulation of dopamine transmission in Trappc9-linked syndrome — PMC - Community awareness of genetic disorders associated with consanguineous marriage — Frontiers in Genetics - Monarch Initiative — MONDO:0019502 - OMIM Phenotypic Series PS249500