👪

Inheritance

2

Autosomal recessive HP:0000007

Disease requires two damaging alleles at the same autosomal locus. In the consanguineous pedigrees in which most AR-NSID genes were discovered the two alleles are identical by descent (homozygous), so the affected haplotype falls inside a run of homozygosity - the property that makes autozygosity mapping the discovery method of choice. Compound heterozygosity is the corresponding mechanism in outbred families. Heterozygous carriers, including the obligate-carrier parents, are unaffected. Recurrence risk for full siblings is 25%, an order of magnitude higher than for the sporadic de novo dominant class, which is why molecular diagnosis has a far larger counselling and prevention payoff here.

Autosomal recessive inheritance Penetrance: COMPLETE Expressivity: VARIABLE

Show evidence (5 references)

PMID:24176302 SUPPORT Human Clinical

"For autosomal recessive ID (ARID) the recurrence risk is high and, in populations with frequent parental consanguinity, ARID is the most common form of ID."

States both defining features of this class relative to the dominant class: high recurrence risk, and concentration in consanguineous populations.

PMID:29302074 SUPPORT Human Clinical

"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."

Quantifies the population denominator in which autosomal recessive inheritance is the dominant mode for intellectual disability.

PMID:18452889 SUPPORT Human Clinical

"All obligate carriers of this family were heterozygous, but none of 192 unrelated healthy individuals from the same population carried this deletion."

Worked demonstration of the recessive model in a single pedigree: unaffected obligate carriers are heterozygous and the allele is absent from population controls.

+ 2 more references

Oligogenic / multi-locus inheritance HP:0010983

A minority of consanguineous families carry pathogenic variants at more than one locus, so the phenotype is the sum of two or more recessive lesions rather than of a single gene. In the largest systematic look at this, eight of 121 large Pakistani families (about 6.6%) segregated multiple pathogenic variants across 19 genes. This matters practically: stopping at the first convincing homozygous variant can leave part of the phenotype unexplained, and recurrence-risk counselling based on a single locus may be wrong.

Oligogenic inheritance

Show evidence (1 reference)

PMID:27457812 SUPPORT Human Clinical

"In another eight families segregation of multiple pathogenic variants was observed, affecting 19 genes that were either known or are novel candidates for ID."

Direct documentation of multi-locus (oligogenic) causation within a consanguineous recessive intellectual-disability cohort.

◆

Subtypes

10

Intellectual disability, autosomal recessive 1 (PRSS12) MONDO:0009580

PRSS12 hgnc:9477

The first AR-NSID gene identified (2002) and the anchor entry of the OMIM phenotypic series. Biallelic truncation of PRSS12, encoding the neuronal serine protease neurotrypsin, which localises to presynaptic terminals at the synaptic cleft. Places the founding member of the class in the synaptic-proteolysis theme.

Show evidence (1 reference)

PMID:12459588 SUPPORT Human Clinical

"A 4-base pair deletion in the neuronal serine protease neurotrypsin gene was associated with autosomal recessive nonsyndromic mental retardation (MR)."

Establishes the PRSS12/neurotrypsin subtype and its biallelic truncating mechanism.

Intellectual disability, autosomal recessive 2 (CRBN) MONDO:0011828

CRBN hgnc:30185

Homozygous nonsense variant (R419X) in CRBN, encoding cereblon, an ATP-dependent Lon protease, in a large North American sectarian kindred. Distinctive within the class for its MILD severity band (IQ 50-70) and for an explicitly documented absence of dysmorphism, microcephaly, autistic features, and neuroimaging abnormality - making it one of the purest non-syndromic exemplars. Places protein turnover in the theme set.

Show evidence (2 references)

PMID:15557513 SUPPORT Human Clinical

"A nonsense mutation causing a premature stop codon in a novel gene (cereblon; CRBN) was identified that encodes for an ATP-dependent Lon protease."

Identifies the causal CRBN lesion defining the MRT2 subtype.

PMID:15557513 SUPPORT Human Clinical

"Previously, a disease locus was mapped for a mild type of nonsyndromic mental retardation (IQ between 50 and 70) to a 4.2-MB interval on chromosome 3p25-pter in a large kindred."

Documents the mild severity band that distinguishes MRT2 from the moderate-to-severe presentations typical of the class.

Intellectual disability, autosomal recessive 3 (CC2D1A) MONDO:0012037

CC2D1A hgnc:30237

Recurrent protein-truncating founder variant in CC2D1A on 19p13.12, identified in nine consanguineous families with SEVERE non-syndromic intellectual disability. CC2D1A is a positive regulator of the IKK/NF-kB cascade and also restrains PDE4D-dependent cAMP hydrolysis. It is the only subtype in this class with a preclinical pharmacological rescue (PDE4 inhibition in male mice) and is therefore the one druggable-mechanism lead.

Show evidence (2 references)

PMID:16033914 SUPPORT Human Clinical

"A protein truncating mutation was identified in the gene CC2D1A in nine consanguineous families with severe autosomal recessive NSMR."

Establishes the CC2D1A subtype, its founder truncating allele, and its severe phenotype band.

PMID:16033914 SUPPORT In Vitro

"CC2D1A is a putative signal transducer participating in positive regulation of I-kappaB kinase/NFkappaB cascade."

Assigns the subtype to the NF-kB signalling theme shared with TRAPPC9.

Intellectual disability, autosomal recessive 5 (NSUN2) MONDO:0012613

NSUN2 hgnc:25994

Homozygous loss of the tRNA methyltransferase NSUN2 in Iranian and Kurdish consanguineous families, mapping to the previously described MRT5 locus. Places RNA (tRNA) modification in the theme set, with a Drosophila short-term-memory model providing cross-species functional support. Also the clearest example within the class of boundary instability: the authors themselves note accompanying facial dysmorphism and raise the possibility that the entity is syndromic.

Show evidence (2 references)

PMID:22541559 SUPPORT Human Clinical

"that cause a loss of the tRNA-methyltransferase-encoding NSUN2 main transcript in homozygotes"

Identifies the NSUN2 loss-of-function mechanism defining MRT5.

PMID:22541559 SUPPORT Model Organism

"When the Drosophila melanogaster NSUN2 ortholog was deleted, severe short-term-memory (STM) deficits were observed; STM could be rescued by re-expression of the wild-type protein in the nervous system."

Cross-species functional support that NSUN2 loss impairs memory. PARTIAL and MODEL_ORGANISM because the readout is Drosophila short-term memory, not the human cognitive phenotype.

Intellectual disability, autosomal recessive 6 (GRIK2) MONDO:0012614

GRIK2 hgnc:4580

Complex homozygous rearrangement of GRIK2 (GLUR6), the kainate-type ionotropic glutamate receptor subunit, in a large consanguineous Iranian family with moderate-to-severe non-syndromic intellectual disability. The predicted product lacks the first ligand-binding domain, the adjacent transmembrane domain, and the pore loop, giving complete loss of function confirmed electrophysiologically. The most direct link in the class between the disease and excitatory synaptic transmission.

Show evidence (2 references)

PMID:17847003 SUPPORT Human Clinical

"cosegregates with moderate-to-severe nonsyndromic autosomal recessive mental retardation in a large, consanguineous Iranian family"

Establishes the GRIK2 subtype and its moderate-to-severe non-syndromic phenotype in a consanguineous pedigree.

PMID:17847003 SUPPORT In Vitro

"The predicted gene product lacks the first ligand-binding domain, the adjacent transmembrane domain, and the putative pore loop, suggesting a complete loss of function of the GLU(K6) protein, which is supported by electrophysiological data."

Functional confirmation of complete loss of receptor function, anchoring the synaptic-transmission mechanism.

Intellectual disability, autosomal recessive 7 (TUSC3) MONDO:0012615

TUSC3 hgnc:30242

Homozygous deletion of TUSC3 (N33) on chromosome 8, encoding a subunit of the ER-bound oligosaccharyltransferase complex. Mechanistically important because it shows a defect in N-glycosylation - a pathway whose other defects produce florid multisystem congenital disorders of glycosylation - presenting as ISOLATED cognitive impairment, and because it was the first gene in the class with mutations reported in more than one family.

Show evidence (2 references)

PMID:18452889 SUPPORT Human Clinical

"Haplotype analyses and copy-number studies led to the identification of a homozygous deletion partly removing TUSC3 (N33) in all patients."

Establishes the biallelic TUSC3 deletion defining MRT7.

PMID:18452889 SUPPORT Human Clinical

"Our data suggest that in contrast to other genetic defects of glycosylation, inactivation of TUSC3 causes nonsyndromic MR"

Directly supports the non-syndromic designation and contrasts it with the syndromic congenital disorders of glycosylation.

Intellectual disability, autosomal recessive 13 (TRAPPC9) MONDO:0013173

TRAPPC9 hgnc:30832

Biallelic truncating variants in TRAPPC9 (NIBP) on 8q24, reported simultaneously by two groups in 2009 in Pakistani, Iranian, and Israeli Arab pedigrees and in many families since. TRAPPC9 is a vesicle-trafficking (TRAPP complex) subunit that also activates NF-kB via IKK-beta, converging with CC2D1A. It sits at the edge of the non-syndromic definition: variable postnatal microcephaly and mild cerebral white matter hypoplasia are common but not obligate, and obesity is described in the fuller phenotype. It is also the one AR-NSID gene with an established parent-of-origin expression bias (see `mechanistic_hypotheses`).

Show evidence (3 references)

PMID:20004765 SUPPORT Human Clinical

"Within this region, we identified a truncating homozygous mutation, R475X, in exon 7 of the gene TRAPPC9."

Establishes the TRAPPC9 truncating mechanism defining MRT13.

PMID:20004765 SUPPORT Human Clinical

"Brain magnetic resonance imaging of affected individuals indicates the presence of mild cerebral white matter hypoplasia. Microcephaly is present in some but not all affected individuals."

Documents the incompletely penetrant accessory features that make this subtype a borderline case for the non-syndromic designation.

PMID:20004763 SUPPORT Human Clinical

"TRAPPC9 is highly expressed in the postmitotic neurons of the cerebral cortex, and MRI analysis of affected patients shows defects in axonal connectivity."

Independent replication in a separate pedigree, plus the connectivity phenotype linking the molecular lesion to cortical circuit formation.

Intellectual disability, autosomal recessive 15 / Rafiq syndrome (MAN1B1) MONDO:0013624

MAN1B1 hgnc:6823

Biallelic variants in MAN1B1 (alpha-1,2-mannosidase) at 9q34.3, a second glycosylation entry point alongside TUSC3. Notable as one of the few genes in the class with an elevated mutation frequency across populations rather than a single-family report. The MONDO label "Rafiq syndrome" reflects the subsequent recognition of dysmorphic features in some families - within the original report one family already showed additional clinical features while three others were strictly non-syndromic, so this subtype straddles the category boundary.

Show evidence (2 references)

PMID:21763484 SUPPORT Human Clinical

"We have used genome-wide genotyping to identify an overlapping homozygosity-by-descent locus on chromosome 9q34.3 (MRT15) in four consanguineous families affected by nonsyndromic autosomal-recessive intellectual disability (NS-ARID) and one in which the patients show additional clinical features."

Establishes the MRT15 locus and simultaneously documents the mixed syndromic/non-syndromic presentation within the founding cohort.

PMID:21763484 SUPPORT Human Clinical

"MAN1B1 is one of the few NS-ARID genes with an elevated mutation frequency in patients with NS-ARID from different populations."

Supports the claim that this subtype is relatively over-represented in the class, unlike the typical single-family genes.

Intellectual disability, autosomal recessive 18 (MED23) MONDO:0013651

MED23 hgnc:2372

Homozygous missense variant (p.R617Q) in MED23, a subunit of the Mediator coactivator complex. Unusually for the class the lesion is a hypomorphic missense rather than a truncation, and the demonstrated defect is specific - failure of JUN and FOS immediate-early-gene induction rather than global transcriptional collapse. Places activity-dependent transcription in the theme set.

Show evidence (2 references)

PMID:21868677 SUPPORT Human Clinical

"Here, we report a missense mutation (p. R617Q) in MED23 that cosegregates with nonsyndromic autosomal recessive intellectual disability."

Establishes the MED23 subtype and its non-syndromic recessive cosegregation.

PMID:21868677 SUPPORT In Vitro

"This mutation specifically impaired the response of JUN and FOS immediate early genes (IEGs) to serum mitogens by altering the interaction between enhancer-bound transcription factors (TCF4 and ELK1, respectively) and Mediator."

Functional mechanism in patient-derived cells: selective loss of immediate-early-gene induction, the transcriptional arm of activity-dependent plasticity.

Intellectual disability, autosomal recessive 57 (MBOAT7) MONDO:0014962

MBOAT7 hgnc:15505

Homozygous inactivating variants in MBOAT7 (LPIAT1), the acyltransferase that loads arachidonic acid onto lysophosphatidylinositol, identified in six consanguineous families out of a >5,000-family neurodevelopmental cohort. Places membrane phospholipid remodelling in the theme set. Sits at the syndromic boundary in the other direction from CRBN: intellectual disability is frequently accompanied by epilepsy and autistic features.

Show evidence (2 references)

PMID:27616480 SUPPORT Human Clinical

"From a cohort of >5,000 families affected by neurodevelopmental disorders, we identified six consanguineous families harboring homozygous inactivating variants in MBOAT7, encoding lysophosphatidylinositol acyltransferase (LPIAT1)."

Establishes the MBOAT7 subtype, its biallelic inactivating mechanism, and its rarity within a very large neurodevelopmental cohort.

PMID:27616480 SUPPORT Human Clinical

"Subjects presented with ID frequently accompanied by epilepsy and autistic features."

Documents the accompanying epilepsy and autistic features that qualify the non-syndromic designation for this subtype.

◈

Mechanistic Hypotheses

3

The recessive class comprises thousands of loci, each individually ultra-rare, because recessive lesions are not filtered by selection

thousands_of_recessive_id_loci EMERGING

Evidence balance 6 support

Under this model the extreme locus heterogeneity of AR-NSID is not an artefact of incomplete discovery but a structural prediction. A heterozygous carrier is unaffected and reproductively unimpaired, so a damaging recessive allele is under weak purifying selection and persists; consequently essentially any gene required for brain development can contribute, not merely the haploinsufficiency-intolerant minority that dominates the dominant class. The predictions are a long tail of genes each explaining one or a few families worldwide, a total locus count in the low thousands, no prevalent gene or pathway, and continued near-linear discovery of novel genes with each new consanguineous cohort. Successive cohorts have behaved exactly as predicted, but the estimate remains an extrapolation from clinic-ascertained multiplex families rather than a population measurement, which is why the status is EMERGING rather than established.

Pathograph links

Biallelic Loss-of-Function Variant in an Autosomal Neurodevelopmental Gene -> Autozygosity Exposes Recessive Alleles in Consanguineous Pedigrees DIRECT The recessive lesion is silent in the heterozygous state; parental relatedness is what converts a carrier allele into an affected homozygote.

Biallelic Loss-of-Function Variant in an Autosomal Neurodevelopmental Gene -> Loss of a Gene-Specific Molecular Function Required by Developing Neurons DIRECT Absence of the gene product removes one component from one of a limited number of recurring cellular processes required for brain development. Under the long-tail model this single edge stands in for many hundreds to thousands of gene-specific instantiations.

Autozygosity Exposes Recessive Alleles in Consanguineous Pedigrees -> Loss of a Gene-Specific Molecular Function Required by Developing Neurons DIRECT Homozygosity for the damaging allele is what allows the molecular consequence to be expressed at all. This is exactly the step the imprinted-bias model contests: for a gene with strong parent-of-origin expression bias, loss of the dominant (maternal) allele alone may reduce expression enough to cross the threshold, so biallelic loss would not be required.

Show evidence (6 references)

PMID:28397838 SUPPORT Human Clinical

"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)."

States the quantitative form of the hypothesis: >2500 autosomal loci, mostly recessive.

PMID:28397838 SUPPORT Human Clinical

"We identified definite or candidate mutations (or CNVs) in 51% of families in 72 different genes, including 26 not previously reported for ARID."

Observed behaviour matching the prediction: even a large modern cohort yields many genes, a large fraction of them novel.

PMID:29302074 SUPPORT Human Clinical

"In 219 of these, we found likely causative variants, involving 77 known and 77 novel AR-ID (candidate) genes, 21 X-linked genes, as well as 9 genes previously implicated in diseases other than ID."

Novel genes equal known genes one-for-one in the largest cohort to date - the signature of an unsaturated long tail.

+ 3 more references

The non-syndromic label reflects depth of phenotyping and cohort size rather than a mechanistically distinct disease class

nonsyndromic_boundary_is_ascertainment EMERGING

Evidence balance 6 support 1 refute

Under this model AR-NSID is the leading edge of recessive gene discovery rather than a separate biological entity: a gene-disease pair is called non-syndromic while it rests on one or two consanguineous pedigrees and a routine dysmorphology examination, and migrates into a named syndrome once enough patients accumulate to reveal a gestalt. The prediction is that MRT-series entities will be progressively reclassified, and several already have been. Within this entry, MRT5/NSUN2 was reported with accompanying facial dysmorphism and the discovering authors themselves raised the possibility of a syndromic form; MRT15/MAN1B1 was reported with a mixed cohort (three strictly non-syndromic families plus one with additional features) and now carries the eponymous MONDO label "Rafiq syndrome"; and ADAT3 - initially characterised as cognitive impairment with strabismus - became "an important recognizable cause of intellectual disability in Arabia" with dysmorphic features once a second cohort was described. The competing, not mutually exclusive model is that a genuine subset of genes has sufficiently brain-restricted consequence that isolated cognitive impairment is the real phenotype; TUSC3 is the strongest case, since N-glycosylation defects usually produce florid multisystem disease yet TUSC3 loss does not. Distinguishing the two requires prospective deep phenotyping of MRT-series cohorts, which has not been done.

Pathograph links

Impaired Cortical Circuit Formation and Connectivity -> Secondary (Postnatal) Microcephaly DIRECT In a subset of subtypes, notably TRAPPC9, reduced postnatal brain growth manifests as acquired microcephaly. This edge is the pathograph location of the syndromic-boundary problem: it is an accessory, incompletely penetrant feature of the kind whose accumulation with deeper phenotyping progressively reclassifies MRT entities as syndromic.

Show evidence (7 references)

PMID:30459488 SUPPORT Human Clinical

"In addition, in the last few months, evidence has been growing that many ARID genes are pleiotropic and that the resulting phenotypes may have a broad spectrum."

The class-level statement of the hypothesis: ARID gene-phenotype relationships are broader than the initial non-syndromic reports imply.

PMID:22541559 SUPPORT Human Clinical

"suggesting that mutations in this gene might even induce a syndromic form of ID"

The original authors of an MRT-series entity explicitly raise reclassification as syndromic - the predicted migration, stated prospectively.

PMID:22541559 SUPPORT Human Clinical

"The humans homozygous for NSUN2 mutations showed an overlapping phenotype consisting of moderate to severe ID and facial dysmorphism"

Documents the dysmorphic features that undercut the non-syndromic designation for this subtype.

+ 4 more references

Parent-of-origin expression bias makes a subset of recessive ID genes pathogenic in the monoallelic (maternally inherited) state

imprinted_bias_monoallelic_arid EMERGING MRT13

Evidence balance 3 support

Some autosomal genes are not fully biallelically expressed but show a parent-of-origin expression BIAS. For TRAPPC9, brain expression is predominantly (~70%) from the maternally inherited allele. In mouse this produces a striking asymmetry: heterozygotes lacking the maternal allele (~70% loss of expression) phenocopy homozygous nulls, whereas heterozygotes lacking the paternal allele are normal. If the same bias holds in human brain, a SINGLE maternally inherited loss-of-function allele in an imprinted-bias ARID gene could be pathogenic - a genotype that standard recessive filtering (which demands two hits) would discard, and that standard recurrence-risk counselling (25%) would misstate. This would also predict apparent "non-penetrance" or unexplained mild cases in families where only one allele is disrupted. The hypothesis is EMERGING: the allele-bias and the mouse asymmetry are established, but the human clinical correlate rests on a single child with increased ad libitum food intake and has not been tested systematically in monoallelic TRAPPC9 carriers.

Pathograph links

Autozygosity Exposes Recessive Alleles in Consanguineous Pedigrees -> Loss of a Gene-Specific Molecular Function Required by Developing Neurons DIRECT Homozygosity for the damaging allele is what allows the molecular consequence to be expressed at all. This is exactly the step the imprinted-bias model contests: for a gene with strong parent-of-origin expression bias, loss of the dominant (maternal) allele alone may reduce expression enough to cross the threshold, so biallelic loss would not be required.

Show evidence (3 references)

PMID:32877400 SUPPORT Model Organism

"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."

Establishes the parent-of-origin expression bias on which the hypothesis depends.

PMID:32877400 SUPPORT Model Organism

"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."

The core prediction demonstrated in vivo: monoallelic maternal loss is sufficient for the phenotype.

PMID:32877400 SUPPORT Human Clinical

"A role for Trappc9 in energy balance was further supported by increased ad libitum food intake in a child with TRAPPC9 deficiency."

The only human datum bearing on the model's translational validity, and it concerns the metabolic rather than the cognitive arm and rests on n=1 - hence PARTIAL and hence the EMERGING status.

?

Discussions and Knowledge Gaps

4

What fraction of all intellectual disability is attributable to autosomal recessive NON-SYNDROMIC gene defects, in consanguineous and in outbred populations?

KNOWLEDGE GAP OPEN arnsid_no_class_specific_prevalence

Attached to

prevalence#

Every published figure is either the all-cause intellectual disability prevalence (about 1%), a whole-class share (about 10% of affected children in an outbred population, syndromic and non-syndromic combined), or a solve rate within a clinic-ascertained multiplex consanguineous cohort (37-58%). None is a population estimate for the non-syndromic recessive class. Solve rates are inflated by multiplex ascertainment - families were selected for having two or more affected offspring, which enriches for recessive inheritance by construction - and are further confounded by cohorts that mix syndromic and non-syndromic cases; in one large consanguineous series 92% of families had additional features beyond intellectual disability. Without a population-based denominator the burden of this class, and therefore the case for carrier-screening programmes, cannot be quantified.

Proposed experiments

Unselected population-based sequencing of intellectual disability in a high-consanguinity setting

exp_arnsid_population_based_sequencing

Ascertain intellectual disability cases from a defined population rather than from genetics clinics, apply exome or genome sequencing to all, and have a clinical geneticist classify each case as syndromic or non-syndromic blinded to the genotype. This yields the first genuine population numerator and denominator for the non-syndromic recessive class, and simultaneously measures the size of the multiplex ascertainment inflation affecting existing estimates.

Reanalysis of existing consanguineous cohorts stratified by sibship structure

exp_arnsid_multiplex_bias_reanalysis

Re-analyse published large consanguineous cohorts stratifying solve rate by number of affected siblings and by syndromic status, to estimate and correct the ascertainment inflation in the headline yields and derive a bias-adjusted class share.

Show evidence (2 references)

PMID:28097321 SUPPORT Human Clinical

"additional features were present in 140 of the families (92.1%)"

Documents the syndromic-contamination confounder directly: 92% of families in this consanguineous cohort had features beyond intellectual disability, i.e. were not strictly non-syndromic, so the headline yield cannot be read as an AR-NSID-specific estimate.

PMID:30459488 SUPPORT Human Clinical

"The prevalence of ARID in a cohort of affected children of an outbred population is estimated to be about 10%, with an upward tendency in still unclarified cases."

The best available figure, and it illustrates the gap: it is a share of affected children rather than a population rate, and it covers ARID as a whole rather than the non-syndromic subset.

How many of the hundreds of AR-ID "candidate" genes reported from consanguineous cohorts are genuinely causal, given that most rest on a single family with no functional follow-up?

KNOWLEDGE GAP OPEN arnsid_candidate_gene_validation_backlog

Attached to

pathophysiology#Biallelic Loss-of-Function Variant in an Autosomal Neurodevelopmental Gene

The discovery engine for this class - one consanguineous family, one autozygous interval, one homozygous variant - produces gene-disease assertions faster than they can be replicated or functionally validated. In a single-family autozygous interval many rare homozygous variants co-segregate perfectly with disease by construction, so segregation carries much less evidential weight than in an outbred pedigree. The founding papers describe their own findings as "probably disease-causing variants" in "candidate" genes, and the largest cohort reports novel candidates at parity with confirmed genes. The consequence is that gene-disease validity for much of this class is unassessed, which propagates directly into clinical variant interpretation.

Proposed experiments

Systematic gene-disease validity curation of the MRT series

exp_arnsid_clingen_validity_curation

Apply ClinGen-style gene-disease validity classification to every MRT series entry and to candidate genes reported from consanguineous cohorts, producing an auditable Definitive/Moderate/Limited/Disputed assignment per gene and a quantified estimate of how much of the reported gene set is currently below reportable evidence thresholds.

Cross-cohort matchmaking for independent second families

exp_arnsid_matchmaking_second_families

Systematically submit single-family AR-NSID candidate genes to matchmaking platforms to seek independent second families with concordant genotype and phenotype before clinical reporting, and measure the proportion of candidates that do and do not replicate.

Show evidence (3 references)

PMID:21937992 SUPPORT Human Clinical

"This study, the largest published so far, has revealed additional mutations in 23 genes previously implicated in intellectual disability or related neurological disorders, as well as single, probably disease-causing variants in 50 novel candidate genes."

The discovering authors themselves label the findings "single, probably disease-causing variants" in "candidate" genes.

PMID:29302074 SUPPORT Human Clinical

"In 219 of these, we found likely causative variants, involving 77 known and 77 novel AR-ID (candidate) genes, 21 X-linked genes, as well as 9 genes previously implicated in diseases other than ID."

Quantifies the backlog: novel candidates match confirmed genes one-for-one in the largest cohort published.

PMID:30459488 SUPPORT Human Clinical

"For an exhaustive deciphering of the genetics of ARID, we suggest research at the level of single genes rather than large meta-analyses."

The field's own recommendation, which is effectively a call for the per-gene validation work this gap describes.

Why do lesions in themes as different as glutamate receptor signalling, tRNA methylation, N-glycosylation, and phospholipid remodelling all converge on the same clinical phenotype, and does the affected theme predict severity or accessory features?

KNOWLEDGE GAP OPEN arnsid_theme_to_phenotype_mapping

Attached to

pathophysiology#Loss of a Gene-Specific Molecular Function Required by Developing Neurons

Thematic recurrence is documented statistically - the gene set is enriched for co-expression and protein-protein interaction - but it is explicitly NOT pathway convergence, since the class review finds no prevalent pathways or protein complexes. It is therefore unresolved whether the themes act through a shared final common pathway (all ultimately degrading synaptic plasticity) or are genuinely parallel routes to a phenotype coarse enough to absorb them. The question is testable because the class already shows theme-correlated variation: the synaptic member GRIK2 and the transcriptional member MED23 produce clean cognitive phenotypes, whereas the lipid member MBOAT7 carries epilepsy and autistic features and the trafficking member TRAPPC9 carries microcephaly, white-matter change, and obesity. Whether that pattern is real or reflects reporting noise across tiny cohorts is unknown.

Proposed experiments

Multi-gene AR-NSID deep-phenotyping cohort

exp_arnsid_multigene_deep_phenotyping

Assemble a cohort with sufficient numbers per gene across several mechanistic themes and apply uniform deep phenotyping (standardised cognitive and adaptive testing, EEG, longitudinal head circumference, dysmorphology, MRI volumetry), then test whether theme membership predicts severity band, epilepsy risk, microcephaly, and syndromic reclassification.

Comparative profiling of patient-derived neurons across mechanistic themes

exp_arnsid_ipsc_theme_convergence

Generate patient iPSC-derived cortical neurons or organoids for representatives of each theme and compare transcriptomic, morphometric, and electrophysiological readouts, to test whether the themes converge on a shared downstream cellular signature or remain distinguishable.

Show evidence (2 references)

PMID:30459488 SUPPORT Human Clinical

"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."

States the negative finding that makes the convergence question genuinely open rather than already answered by a shared pathway.

PMID:26748517 SUPPORT Computational

"Systematic identification of highly enriched functional themes and phenotypes revealed typical phenotype combinations characterizing process-defined groups of ID disorders, such as chromatin-related disorders and deficiencies in DNA repair."

Demonstrates that theme membership CAN predict phenotype combinations for some processes, making the question tractable for the recessive themes catalogued here.

Do the rodent models used for AR-NSID genes report on human intellectual disability, given that the rescued readouts are sex-restricted rodent memory assays and that a key dosage mechanism (parent-of-origin bias) is established only in mouse brain?

HUMAN MODEL MISMATCH OPEN arnsid_model_translational_validity

Attached to

pathophysiology#Impaired Synaptic Transmission and Plasticity pathophysiology#Impaired Cortical Circuit Formation and Connectivity

Evidence exists in the models; what is open is whether it transfers. Three specific mismatches are documented rather than hypothetical. First, intellectual disability is a construct (IQ plus adaptive functioning) with no animal equivalent, so every model readout is a proxy - Drosophila short-term memory for NSUN2, mouse spatial and social memory for CC2D1A and TRAPPC9. Second, the one pharmacological rescue in the class is MALE-SPECIFIC: the PDE4D signalling deficit is absent in Cc2d1a-deficient female mice and the inhibitor has no effect in them, so a therapy developed on this model could fail in half of patients for reasons the model cannot reveal. Third, the parent-of-origin dosage mechanism for Trappc9 is established in mouse brain, and its human correlate rests on a single child with increased food intake - a metabolic rather than cognitive readout at n=1. This is a HUMAN_MODEL_MISMATCH rather than a KNOWLEDGE_GAP because in each case the model data are solid and it is their translational validity that is uncertain.

Proposed experiments

Sex-stratified analysis of CC2D1A-related disease in humans

exp_arnsid_sex_stratified_human_correlate

Assemble the published and unpublished CC2D1A cohort and test whether cognitive severity, autistic features, and behavioural phenotype differ by sex as the mouse model predicts, before any PDE4-inhibitor trial is contemplated. A negative result would indicate the male-specific mouse mechanism does not transfer.

Allele-specific TRAPPC9 expression in human brain and phenotyping of monoallelic carriers

exp_arnsid_human_brain_trappc9_allelic_bias

Measure allele-specific TRAPPC9 expression in human brain tissue to test whether the ~70% maternal bias seen in mouse holds in humans, and systematically phenotype heterozygous carriers of maternally inherited TRAPPC9 loss-of-function alleles for subclinical cognitive or metabolic effects.

Human iPSC cortical models as a bridge readout

exp_arnsid_human_ipsc_cortical_readouts

Use patient-derived iPSC cortical neurons and organoids, which retain human-specific cortical progenitor biology absent from rodent models, to test whether the cellular deficits identified in mouse (neurite outgrowth, cAMP/CREB signalling, progenitor depletion) are reproduced in human cells of the correct genotype.

Show evidence (4 references)

PMID:30732858 SUPPORT Model Organism

"Restoring PDE4D activity using an inhibitor rescues cognitive deficits in male mice but has no effect on female mice."

The sex-restricted rescue that makes translational validity the open question for the class's only druggable lead.

PMID:30732858 SUPPORT Model Organism

"We propose that male-specific signaling mechanisms are involved in establishing sex bias in neurodevelopmental disorders."

The authors themselves frame the finding as a general claim about sex-specific mechanism, which is precisely what needs human confirmation.

PMID:32877400 SUPPORT Human Clinical

"A role for Trappc9 in energy balance was further supported by increased ad libitum food intake in a child with TRAPPC9 deficiency."

The single human datum supporting the mouse imprinting model, and it concerns metabolism rather than cognition at n=1 - the mismatch this gap describes.

+ 1 more reference

⚙

Pathophysiology

7

Biallelic Loss-of-Function Variant in an Autosomal Neurodevelopmental Gene

The initiating lesion is the presence of two damaging alleles at one autosomal locus - most often a homozygous protein-truncating variant, but also homozygous whole-gene or partial deletion (TUSC3), complex rearrangement (GRIK2), or homozygous hypomorphic missense (MED23, MAN1B1) - so that the gene product is absent or severely reduced. Unlike the dominant class the affected genes are NOT restricted to the haploinsufficiency-intolerant fraction of the genome: because one working copy suffices, a damaging recessive allele is under weak purifying selection and persists in the population, so the recessive class draws on a far larger gene pool. That is the structural reason the number of candidate loci is estimated in the thousands and each individual gene explains only a handful of families. Homozygous copy-number lesions are a genuine and recurrent subset of the trigger, which is why chromosomal microarray remains part of the workup even in a suspected-recessive workflow.

The recessive class comprises thousands of loci, each individually ultra-rare, because recessive lesions are not filtered by selection

nervous system development GO:0007399

brain UBERON:0000955

Downstream

Autozygosity Exposes Recessive Alleles in Consanguineous Pedigrees The recessive lesion is silent in the heterozygous state; parental relatedness is what converts a carrier allele into an affected homozygote.

Loss of a Gene-Specific Molecular Function Required by Developing Neurons Absence of the gene product removes one component from one of a limited number of recurring cellular processes required for brain development. Under the long-tail model this single edge stands in for many hundreds to thousands of gene-specific instantiations.

Show evidence (5 references)

PMID:17718851 SUPPORT Human Clinical

"All of the mutations in the ARNSMR-causing genes are protein truncating, indicating a putative severe loss-of-function effect."

Establishes complete loss of function as the allelic mechanism for the founding members of the class.

PMID:18452889 SUPPORT Human Clinical

"Recent studies have shown that autosomal recessive mental retardation (ARMR) is extremely heterogeneous, and there is reason to believe that the number of underlying gene defects goes into the thousands."

States the locus-heterogeneity consequence of a recessive architecture not confined to constraint-intolerant genes.

PMID:28397838 SUPPORT Human Clinical

"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)."

Quantifies the gene pool and states explicitly that the majority is recessive.

+ 2 more references

Autozygosity Exposes Recessive Alleles in Consanguineous Pedigrees

Parental consanguinity does not create the mutant allele; it raises the probability that a rare carrier allele is inherited in duplicate, identical by descent, so that the two copies lie inside a shared run of homozygosity. This is the population-genetic step that converts an invisible recessive burden into overt disease, and it is why the class is concentrated in populations where cousin marriage is common - the risk of recessive intellectual disability in children of first cousins or closer is an order of magnitude higher than in children of unrelated parents. It is also the methodological lever: because the causal variant must lie inside an autozygous interval shared by all affected siblings, homozygosity mapping reduces the search space enough for exome sequencing to identify the gene in a single family. The same reasoning explains why the class was almost invisible before this method - conventional linkage requires many independent families per locus, which extreme locus heterogeneity makes impossible.

Parent-of-origin expression bias makes a subset of recessive ID genes pathogenic in the monoallelic (maternally inherited) state The recessive class comprises thousands of loci, each individually ultra-rare, because recessive lesions are not filtered by selection

brain UBERON:0000955

Downstream

Loss of a Gene-Specific Molecular Function Required by Developing Neurons Homozygosity for the damaging allele is what allows the molecular consequence to be expressed at all. This is exactly the step the imprinted-bias model contests: for a gene with strong parent-of-origin expression bias, loss of the dominant (maternal) allele alone may reduce expression enough to cross the threshold, so biallelic loss would not be required.

Show evidence (4 references)

PMID:30459488 SUPPORT Human Clinical

"Taken together, it seems that children of related parents are at a 2 to 3 times higher risk for ID."

Quantifies the overall effect of parental relatedness on intellectual disability risk at the population level.

PMID:21937992 SUPPORT Human Clinical

"To expedite the molecular elucidation of autosomal-recessive intellectual disability, we have now performed homozygosity mapping, exon enrichment and next-generation sequencing in 136 consanguineous families with autosomal-recessive intellectual disability from Iran and elsewhere."

The defining methodology: autozygosity mapping plus exome sequencing in consanguineous families, the strategy that opened the class.

PMID:20004765 SUPPORT Human Clinical

"Using Affymetrix 5.0 single nucleotide polymorphism (SNP) microarrays, we identified a 3.2 Mb region on 8q24 with a continuous run of 606 homozygous SNPs shared among all affected members of the family."

Concrete worked example of an autozygous interval - a run of homozygous SNPs shared by affected siblings - delimiting the causal locus.

+ 1 more reference

Loss of a Gene-Specific Molecular Function Required by Developing Neurons

The class has no shared pathway. The authoritative review states plainly that there are no prevalent ARID genes, pathways, or protein complexes, and that the functions of the affected proteins are very diverse. What recurs is a limited set of cellular THEMES, each entered by several genes: synaptic transmission and proteolysis (GRIK2, PRSS12, MPDZ), activity-dependent and general transcriptional regulation (MED23, TET1), RNA and tRNA modification (NSUN2, ADAT3, ELP2, METTL23), protein N-glycosylation and glycoprotein quality control (TUSC3, MAN1B1, ST3GAL3), vesicle trafficking with NF-kB signalling (TRAPPC9, TRAPPC6B, WASHC4, TBC1D23, CC2D1A), regulated protein degradation (CRBN), and membrane phospholipid remodelling (MBOAT7). These are post-hoc organising themes, not a claim of pathway convergence. What IS statistically established at class level is weaker but real: the gene set is enriched for co-expression, for direct protein-protein interaction with products of already-known intellectual disability genes, and for prenatal and infant brain expression.

neuron CL:0000540

NSUN2 hgnc:25994 ADAT3 hgnc:25151 CRBN hgnc:30185

nervous system development GO:0007399 RNA methylation GO:0001510 protein N-linked glycosylation GO:0006487 vesicle-mediated transport GO:0016192 canonical NF-kappaB signal transduction GO:0007249 regulation of DNA-templated transcription GO:0006355 proteolysis GO:0006508 phosphatidylinositol acyl-chain remodeling GO:0036149

brain UBERON:0000955 cerebral cortex UBERON:0000956

Downstream

Impaired Synaptic Transmission and Plasticity Several themes (glutamate receptor function, synaptic proteolysis, immediate-early-gene induction, phosphoinositide remodelling) converge directly on synaptic function.

Impaired Cortical Circuit Formation and Connectivity Trafficking, glycosylation, and transcriptional themes act during prenatal and early postnatal cortical development, affecting neuronal differentiation, neurite outgrowth, and axonal connectivity.

Show evidence (6 references)

PMID:30459488 SUPPORT Human Clinical

"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."

The load-bearing negative claim for this node: the class must be modelled as thematic recurrence, not pathway convergence.

PMID:21937992 SUPPORT Human Clinical

"Proteins encoded by several of these genes interact directly with products of known intellectual disability genes, and many are involved in fundamental cellular processes such as transcription and translation, cell-cycle control, energy metabolism and fatty-acid synthesis, which seem to be..."

Names several of the recurring processes and establishes direct protein-protein interaction with known intellectual disability gene products.

PMID:27457812 SUPPORT Computational

"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."

The strongest class-level statistical evidence for functional coherence: spatiotemporally restricted co-expression in human brain.

+ 3 more references

Impaired Synaptic Transmission and Plasticity

Where the affected theme is synaptic, the immediate cellular consequence is a defect in excitatory neurotransmission or in the activity-dependent changes that encode experience. GRIK2 loss removes a kainate-receptor subunit from the excitatory synapse with electrophysiologically confirmed loss of function; PRSS12/neurotrypsin loss removes a protease acting at the presynaptic membrane lining the synaptic cleft, where its proteolysis is required for normal synaptic function; MED23 loss blocks JUN/FOS immediate-early-gene induction, the transcriptional arm of activity-dependent plasticity; MBOAT7 loss alters the arachidonoyl phosphatidylinositol pool from which synaptic signalling lipids derive; and CC2D1A loss releases PDE4D hyperactivity, depleting hippocampal cAMP and reducing CREB signalling. The resulting deficit is in the machinery of learning and memory rather than in gross brain structure - which is precisely why the phenotype is cognitive and non-syndromic.

neuron CL:0000540 glutamatergic neuron CL:0000679

GRIK2 hgnc:4580 PRSS12 hgnc:9477 MED23 hgnc:2372 MBOAT7 hgnc:15505 CC2D1A hgnc:30237

chemical synaptic transmission GO:0007268 regulation of synaptic plasticity GO:0048167 modulation of chemical synaptic transmission GO:0050804

synapse GO:0045202 presynapse GO:0098793

hippocampal formation UBERON:0002421 cerebral cortex UBERON:0000956

Downstream

Impaired Learning, Memory, and Cognitive Development Loss of synaptic plasticity is the proximate cellular substrate of the learning and memory deficit.

Seizures A subtype-restricted rather than class-level consequence: where the synaptic lesion involves membrane phospholipid remodelling (MBOAT7/MRT57), the altered excitatory signalling substrate also lowers seizure threshold. Seizures are explicitly absent in other subtypes (CRBN/MRT2), so this edge must not be read as a class-wide consequence.

Autistic Features Subtype-restricted in the same way as the seizure edge, and documented in the same MBOAT7/MRT57 cohort; explicitly absent in the CRBN/MRT2 kindred.

Show evidence (6 references)

PMID:12459588 SUPPORT Human Clinical

"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."

Localises the founding gene product to the presynaptic terminal, anchoring the synaptic mechanism in human brain tissue.

PMID:12459588 SUPPORT Human Clinical

"These findings suggest that neurotrypsin-mediated proteolysis is required for normal synaptic function and suggest potential insights into the pathophysiological bases of mental retardation."

Explicitly links the molecular lesion to synaptic function as the pathophysiological basis of the cognitive phenotype.

PMID:17847003 SUPPORT In Vitro

"This finding provides the first proof that GLU(K6) is indispensable for higher brain functions in humans, and future studies of this and other ionotropic kainate receptors will shed more light on the pathophysiology of mental retardation."

Establishes an ionotropic glutamate receptor as necessary for human cognition, the most direct synaptic-transmission link in the class.

+ 3 more references

Impaired Cortical Circuit Formation and Connectivity

Where the affected theme acts during development - vesicle trafficking, N-glycosylation of nascent membrane and secreted proteins, general transcriptional regulation - the consequence is a defect in the BUILDING of cortical circuitry rather than in its moment-to-moment operation. The recessive gene set is preferentially expressed in prenatal brain, and the best-characterised example, TRAPPC9, is highly expressed in postmitotic cortical neurons, with affected individuals showing MRI evidence of impaired axonal connectivity and mild cerebral white matter hypoplasia. Crucially the structural consequence is subtle - reduced connectivity and, in some subtypes, mild postnatal microcephaly - rather than a gross malformation such as lissencephaly or agenesis of the corpus callosum. This quantitative rather than qualitative structural effect is what keeps the class inside the non-syndromic category: routine clinical neuroimaging is usually reported as normal or near-normal. Note also what this node is NOT: there is no degenerative, ischaemic, oxidative, inflammatory, or fibrotic arm. AR-NSID is a mis-built circuit, not a damaged one.

The non-syndromic label reflects depth of phenotyping and cohort size rather than a mechanistically distinct disease class

neuron CL:0000540 pyramidal neuron CL:0000598

TRAPPC9 hgnc:30832 TUSC3 hgnc:30242 MAN1B1 hgnc:6823

neuron differentiation GO:0030182 neurogenesis GO:0022008 nervous system development GO:0007399

cerebral cortex UBERON:0000956 white matter UBERON:0002316

Downstream

Impaired Learning, Memory, and Cognitive Development Reduced cortical connectivity limits the substrate available for cognitive function.

Cerebral White Matter Hypoplasia The imaging correlate of the connectivity defect, reported in TRAPPC9/MRT13. Quantitative and subtle rather than a recognisable malformation, which is why it does not by itself move the case out of the non-syndromic category.

Secondary (Postnatal) Microcephaly In a subset of subtypes, notably TRAPPC9, reduced postnatal brain growth manifests as acquired microcephaly. This edge is the pathograph location of the syndromic-boundary problem: it is an accessory, incompletely penetrant feature of the kind whose accumulation with deeper phenotyping progressively reclassifies MRT entities as syndromic.

Show evidence (5 references)

PMID:20004763 SUPPORT Human Clinical

"TRAPPC9 is highly expressed in the postmitotic neurons of the cerebral cortex, and MRI analysis of affected patients shows defects in axonal connectivity."

Direct human evidence that the molecular lesion produces a cortical connectivity defect rather than a gross malformation.

PMID:20004763 SUPPORT Human Clinical

"This suggests essential roles of TRAPPC9 in human brain development, possibly through its effect on NF-kappaB activation and protein trafficking in the postmitotic neurons of the cerebral cortex."

Links the trafficking/NF-kB theme specifically to postmitotic cortical neuron development.

PMID:20004765 SUPPORT Human Clinical

"Brain magnetic resonance imaging of affected individuals indicates the presence of mild cerebral white matter hypoplasia."

Documents the subtle white-matter imaging correlate, consistent with a connectivity rather than malformation phenotype.

+ 2 more references

Secondary (Postnatal) Microcephaly

An incompletely penetrant consequence in a subset of subtypes, most clearly TRAPPC9-related disease, in which head circumference is normal at birth and falls across centiles postnatally. Its variability is diagnostically important: because microcephaly is present in some but not all affected individuals even within the same gene, its absence does not exclude the diagnosis, and its presence does not by itself reclassify the case as syndromic.

TRAPPC9 hgnc:30832

brain UBERON:0000955

Downstream

Secondary Microcephaly The mechanism node surfaces as the clinically scored phenotype: head circumference normal at birth then falling across centiles. Incompletely penetrant even within one gene.

Show evidence (2 references)

PMID:20004765 SUPPORT Human Clinical

"Microcephaly is present in some but not all affected individuals."

Directly establishes the incomplete penetrance of microcephaly within a single AR-NSID gene.

PMID:20004763 SUPPORT Human Clinical

"We identified a genetic locus for autosomal-recessive nonsyndromic intellectual disability associated with variable postnatal microcephaly through homozygosity mapping of a consanguineous Israeli Arab family."

Independent report of the same variable postnatal microcephaly in a separate pedigree.

Impaired Learning, Memory, and Cognitive Development

The convergent organism-level outcome: significantly impaired intellectual functioning and adaptive behaviour with onset in the developmental period. Severity spans the range from mild (CRBN, IQ 50-70) to severe (CC2D1A), with moderate-to-severe the commonest reported band - though that distribution is inflated by ascertainment through multiplex families referred to specialist genetics services. Because the upstream lesions damage the machinery of synaptic plasticity and cortical connectivity rather than producing a specific structural or metabolic lesion, the deficit is global and STATIC rather than regressive: the developmental trajectory is shallower, so the gap versus peers widens with age, but there is no loss of previously acquired skills. Developmental regression therefore argues against this diagnosis and toward a degenerative or metabolic alternative. Notably, this class is not the low tail of the normal-range polygenic intelligence distribution - common variants in NS-ARID genes are not enriched for intelligence QTLs.

learning or memory GO:0007611 cognition GO:0050890

brain UBERON:0000955

Downstream

Intellectual Disability The obligate clinical expression of the convergent mechanism: significantly impaired intellectual functioning and adaptive behaviour with onset in the developmental period.

Global Developmental Delay The presenting form of the same deficit in infancy and early childhood, before an IQ can be formally assigned; global rather than domain-selective.

Mild Intellectual Disability The mild end of the severity range the mechanism produces, documented in the CRBN/MRT2 kindred.

Moderate Intellectual Disability A documented severity band in the published class literature, exemplified by the GRIK2/MRT6 pedigree.

Severe Intellectual Disability The severe end of the range, documented across the nine consanguineous CC2D1A/MRT3 families.

Impaired Literacy and Academic Attainment The functional expression of the cognitive deficit in adult life, where outcome is reported at attainment rather than IQ granularity.

Show evidence (5 references)

PMID:15557513 SUPPORT Human Clinical

"This finding implicates a role for the ATP-dependent degradation of proteins in memory and learning."

Frames the organism-level outcome of a recessive AR-NSID lesion as a memory and learning deficit.

PMID:17847003 SUPPORT Human Clinical

"cosegregates with moderate-to-severe nonsyndromic autosomal recessive mental retardation in a large, consanguineous Iranian family"

Documents the moderate-to-severe band that is typical for the class.

PMID:16033914 SUPPORT Human Clinical

"A protein truncating mutation was identified in the gene CC2D1A in nine consanguineous families with severe autosomal recessive NSMR."

Documents the severe end of the severity range within the class.

+ 2 more references

⬡

Pathograph

Use the checkboxes to hide or show graph categories. Hover nodes for evidence and cross-linked metadata.

●

Phenotypes

12

Head and Neck 1

Secondary Microcephaly HP:0005484

No frequency band is assigned, and the omission is deliberate rather than an oversight. PMID:32877400 tallies microcephaly at 95% of reported TRAPPC9 cases in the same sentence that gives the 52% obesity figure used to justify the FREQUENT band on `Obesity`, so a VERY_FREQUENT band here would be arithmetically available. It is withheld because that tally counts "microcephaly" without separating primary from secondary (postnatal) forms, whereas this phenotype is specifically bound to HP:0005484 Secondary microcephaly; the sibling PMID:20004765 item is marked PARTIAL for the same reason. Per the repo frequency-evidence guidelines, a band is omitted rather than asserted on a denominator that does not match the phenotype scored.

Show evidence (2 references)

PMID:20004763 SUPPORT Human Clinical

"We identified a genetic locus for autosomal-recessive nonsyndromic intellectual disability associated with variable postnatal microcephaly through homozygosity mapping of a consanguineous Israeli Arab family."

Establishes variable postnatal (secondary) microcephaly in the TRAPPC9 subtype.

PMID:20004765 SUPPORT Human Clinical

"Microcephaly is present in some but not all affected individuals."

Supports the incomplete penetrance. PARTIAL because the report does not distinguish primary from secondary microcephaly in every case.

Nervous System 6

Intellectual Disability OBLIGATE HP:0001249

Show evidence (2 references)

PMID:24176302 SUPPORT Human Clinical

"For autosomal recessive ID (ARID) the recurrence risk is high and, in populations with frequent parental consanguinity, ARID is the most common form of ID."

Intellectual disability is the defining phenotype of the class by construction; this review states the class in those terms.

PMID:12459588 SUPPORT Human Clinical

"A 4-base pair deletion in the neuronal serine protease neurotrypsin gene was associated with autosomal recessive nonsyndromic mental retardation (MR)."

The founding report of the class, associating a biallelic lesion with non-syndromic intellectual disability.

Mild Intellectual Disability HP:0001256

No frequency band is assigned. Per the repo frequency-evidence guidelines, the severity distribution across the class cannot be quantified from a literature ascertained through multiplex families referred for moderate-to-severe intellectual disability.

Show evidence (1 reference)

PMID:15557513 SUPPORT Human Clinical

"Previously, a disease locus was mapped for a mild type of nonsyndromic mental retardation (IQ between 50 and 70) to a 4.2-MB interval on chromosome 3p25-pter in a large kindred."

Documents a mild-severity AR-NSID entity with an explicit IQ range.

Severe Intellectual Disability HP:0010864

Show evidence (1 reference)

PMID:16033914 SUPPORT Human Clinical

"A protein truncating mutation was identified in the gene CC2D1A in nine consanguineous families with severe autosomal recessive NSMR."

Documents severe non-syndromic intellectual disability across nine pedigrees sharing one recessive lesion.

Global Developmental Delay HP:0001263

Show evidence (2 references)

PMID:15557513 SUPPORT Human Clinical

"Developmental milestones were mildly delayed since early childhood."

Documents early-childhood developmental delay as the presenting feature in an AR-NSID kindred.

PMID:31182824 SUPPORT Human Clinical

"We defined NDD as global developmental delay, intellectual disability, and/or autism spectrum disorder."

Confirms that global developmental delay is the standard clinical presentation category under which these cases are ascertained. PARTIAL because the source is a definition within a diagnostic-yield meta-analysis, not an AR-NSID-specific phenotype series.

Seizures HP:0001250

No class-level frequency band is assigned: seizures are subtype-restricted (frequent in MBOAT7, explicitly absent in CRBN) and no cohort quantifies epilepsy across AR-NSID as a whole.

Show evidence (3 references)

PMID:27616480 SUPPORT Human Clinical

"Subjects presented with ID frequently accompanied by epilepsy and autistic features."

Documents epilepsy as a frequent accompaniment in the MBOAT7 subtype.

PMID:27616480 SUPPORT Human Clinical

"The risk of epilepsy among individuals with intellectual disability (ID) is approximately ten times that of the general population."

Provides the base-rate context for the intellectual-disability population within which this subtype-restricted phenotype sits.

PMID:15557513 REFUTE Human Clinical

"None of the affected individuals had peri- or postnatal infections, afebrile seizures, toxic exposures, or significant head trauma."

Explicit documented absence of seizures in a different AR-NSID subtype, confirming this is not a class-level phenotype.

Autistic Features Autism HP:0000717

Show evidence (2 references)

PMID:27616480 SUPPORT Human Clinical

"Subjects presented with ID frequently accompanied by epilepsy and autistic features."

Documents autistic features in the MBOAT7 subtype.

PMID:15557513 REFUTE Human Clinical

"There was good eye contact and no autistic features."

Explicit documented absence of autistic features in a different AR-NSID subtype, confirming that this is not a class-level phenotype.

Growth 1

Obesity FREQUENT HP:0001513

Show evidence (4 references)

PMID:32877400 SUPPORT Human Clinical

"In humans, both homozygous and compound heterozygous mutations in TRAPPC9 (NIBP) associate with developmental delay, microcephaly (95% reported cases), and obesity (52% reported cases)"

Human case-report tally establishing obesity in TRAPPC9-related disease and quantifying it at 52%, which falls in the FREQUENT band (30-79%) and therefore supports the frequency qualifier directly rather than by inference.

PMID:32877400 SUPPORT Human Clinical

"Loss-of-function mutations in human TRAPPC9 cause a rare neurodevelopmental syndrome characterized by microcephaly and obesity."

Second human-anchored statement of the same association, from the abstract.

PMID:32877400 SUPPORT Model Organism

"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."

Animal-model corroboration of the weight phenotype. Kept as a SEPARATE evidence item with evidence_source MODEL_ORGANISM so that the human phenotype is not resting on mouse data.

+ 1 more reference

Other 4

Moderate Intellectual Disability HP:0002342

Show evidence (2 references)

PMID:17847003 SUPPORT Human Clinical

"cosegregates with moderate-to-severe nonsyndromic autosomal recessive mental retardation in a large, consanguineous Iranian family"

Documents the moderate-to-severe band in a founding AR-NSID pedigree.

PMID:22541559 SUPPORT Human Clinical

"The humans homozygous for NSUN2 mutations showed an overlapping phenotype consisting of moderate to severe ID and facial dysmorphism"

Second independent gene with the same moderate-to-severe band.

Impaired Literacy and Academic Attainment

Deliberately left without an ontology `term:` binding. The obvious candidate, HP:0001328 Specific learning disability, is a FALSE match rather than merely an imprecise one: its HPO definition ends "The impairment is not related to a global deficiency of intelligence", which explicitly excludes the case described here, namely attainment limited BY global intellectual disability. HPO currently has no class for functional literacy or academic attainment in the context of global intellectual disability, so this is a needs-term (NTR) candidate; per the term-precision guidance, no term is preferred over a misleading one.

Show evidence (2 references)

PMID:15557513 SUPPORT Human Clinical

"The highest reading ability achieved was at the first-grade level in a 32-year-old affected woman as assessed by the Gray Oral Reading Scale."

Quantifies functional academic attainment in an AR-NSID cohort using a standardised instrument.

PMID:15557513 SUPPORT Human Clinical

"In five affected adult individuals, writing skills were limited to signing their own names."

Further functional-outcome detail for the same cohort.

Cerebral White Matter Hypoplasia Aplasia/Hypoplasia of the cerebral white matter HP:0012429

Show evidence (2 references)

PMID:20004765 SUPPORT Human Clinical

"Brain magnetic resonance imaging of affected individuals indicates the presence of mild cerebral white matter hypoplasia."

Directly documents the white-matter imaging finding in the TRAPPC9 subtype.

PMID:20004763 SUPPORT Human Clinical

"MRI analysis of affected patients shows defects in axonal connectivity."

Independent imaging correlate of the same connectivity defect.

Absence of Dysmorphism and Extra-Neural Involvement

Curated as a phenotype entry because the absence is definitional and diagnostically load-bearing, not incidental; the evidence items are REFUTE because they document the absence of the finding rather than its presence. Deliberately left without an ontology `term:` binding. HP:0000271 Abnormality of the face was considered and rejected: the constellation asserted here also covers absent organomegaly, spasticity, weakness, neuropathy, visual deficits, feeding problems, and metabolic derangement, so a facial term would anchor only a fraction of the claim while implying the rest was scoped to the face. HPO has no single class for "no syndromic gestalt on expert examination", so this is a needs-term (NTR) candidate; the individual absent features are enumerated in the description and evidence snippets instead.

Show evidence (3 references)

PMID:15557513 REFUTE Human Clinical

"Detailed physical and neurologic examinations in all study participants did not show microcephaly, organomegaly, phakomata, spasticity, weakness, neuropathy, visual deficits, feeding problems, or psychiatric disorders."

Documents the absence of extra-neural and additional neurological features. REFUTE is used deliberately: the evidence refutes the presence of dysmorphic/extra-neural abnormality, which is the definitional requirement for this class.

PMID:15557513 REFUTE Human Clinical

"The phenotype does not include congenital anomalies or dysmorphic features."

The explicit non-syndromic statement for an MRT-series entity.

PMID:18452889 REFUTE Human Clinical

"Our data suggest that in contrast to other genetic defects of glycosylation, inactivation of TUSC3 causes nonsyndromic MR"

A second gene in which the absence of the expected multisystem syndromic features is the reported finding.

🧬

Genetic Associations

12

PRSS12

Gene: PRSS12 hgnc:9477 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (2 references)

PMID:12459588 SUPPORT Human Clinical

"A 4-base pair deletion in the neuronal serine protease neurotrypsin gene was associated with autosomal recessive nonsyndromic mental retardation (MR)."

Primary gene-disease evidence for PRSS12 in AR-NSID.

PMID:12459588 SUPPORT Human Clinical

"In situ hybridization experiments on human fetal brains showed that neurotrypsin was highly expressed in brain structures involved in learning and memory."

Human fetal-brain expression pattern linking the gene to the learning and memory substrate.

CRBN

Gene: CRBN hgnc:30185 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (1 reference)

PMID:15557513 SUPPORT Human Clinical

"A nonsense mutation causing a premature stop codon in a novel gene (cereblon; CRBN) was identified that encodes for an ATP-dependent Lon protease."

Primary gene-disease evidence for CRBN in AR-NSID.

CC2D1A

Gene: CC2D1A hgnc:30237 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (2 references)

PMID:16033914 SUPPORT Human Clinical

"A protein truncating mutation was identified in the gene CC2D1A in nine consanguineous families with severe autosomal recessive NSMR."

Primary gene-disease evidence for CC2D1A across nine pedigrees.

PMID:16033914 SUPPORT Model Organism

"Expression of CC2D1A mRNA was shown in the embryonic ventricular zone and developing cortical plate in staged mouse embryos, persisting into adulthood, with highest expression in the cerebral cortex and hippocampus."

Developmental expression pattern consistent with the cortical-circuit mechanism. MODEL_ORGANISM because the expression series is in mouse embryos.

NSUN2

Gene: NSUN2 hgnc:25994 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (1 reference)

PMID:22541559 SUPPORT Human Clinical

"that cause a loss of the tRNA-methyltransferase-encoding NSUN2 main transcript in homozygotes"

Primary gene-disease evidence for NSUN2 in AR-NSID.

GRIK2

Gene: GRIK2 hgnc:4580 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (1 reference)

PMID:17847003 SUPPORT Human Clinical

"cosegregates with moderate-to-severe nonsyndromic autosomal recessive mental retardation in a large, consanguineous Iranian family"

Primary gene-disease evidence for GRIK2 in AR-NSID.

TUSC3

Gene: TUSC3 hgnc:30242 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (2 references)

PMID:18452889 SUPPORT Human Clinical

"Haplotype analyses and copy-number studies led to the identification of a homozygous deletion partly removing TUSC3 (N33) in all patients."

Primary gene-disease evidence for TUSC3 in AR-NSID.

PMID:18452889 SUPPORT Human Clinical

"TUSC3 is only the fifth gene implicated in NS-ARMR and the first for which mutations have been reported in more than one family."

Documents both the historical position of TUSC3 and its multi-family replication.

TRAPPC9

Gene: TRAPPC9 hgnc:30832 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (2 references)

PMID:20004765 SUPPORT Human Clinical

"In a second large NS-ARMR/ID family, previously linked to 8q24 in a study of Iranian families, we identified a 4 bp deletion within exon 14 of TRAPPC9, also segregating with the phenotype and truncating the protein."

Two independent pedigrees with distinct truncating TRAPPC9 alleles within a single report.

PMID:20004763 SUPPORT Human Clinical

"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."

Simultaneous independent discovery by a second group in a different population - unusually strong replication for this class.

MAN1B1

Gene: MAN1B1 hgnc:6823 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (2 references)

PMID:21763484 SUPPORT Human Clinical

"MAN1B1 is one of the few NS-ARID genes with an elevated mutation frequency in patients with NS-ARID from different populations."

Supports the cross-population recurrence claim for MAN1B1.

PMID:21763484 SUPPORT In Vitro

"Both missense mutations are at amino acid residues that are conserved across the animal kingdom, and they either reduce k(cat) by ∼1300-fold or disrupt stable protein expression in mammalian cells."

Functional demonstration that the missense alleles are severe loss-of-function, satisfying the recessive model.

MED23

Gene: MED23 hgnc:2372 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (1 reference)

PMID:21868677 SUPPORT Human Clinical

"Here, we report a missense mutation (p. R617Q) in MED23 that cosegregates with nonsyndromic autosomal recessive intellectual disability."

Primary gene-disease evidence for MED23 in AR-NSID.

MBOAT7

Gene: MBOAT7 hgnc:15505 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (2 references)

PMID:27616480 SUPPORT Human Clinical

"From a cohort of >5,000 families affected by neurodevelopmental disorders, we identified six consanguineous families harboring homozygous inactivating variants in MBOAT7, encoding lysophosphatidylinositol acyltransferase (LPIAT1)."

Primary gene-disease evidence for MBOAT7 in AR-NSID, with cohort denominator.

PMID:27616480 SUPPORT In Vitro

"LPIAT1 is a membrane-bound phospholipid-remodeling enzyme that transfers arachidonic acid (AA) to lysophosphatidylinositol to produce AA-containing phosphatidylinositol."

Defines the biochemical function placing this gene in the membrane-lipid theme.

ADAT3

Gene: ADAT3 hgnc:25151 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (2 references)

PMID:26842963 SUPPORT Human Clinical

"A single homozygous founder mutation (c.382G>A:p.(V128M)) in the ADAT3 gene, which encodes a protein that functions in tRNA editing, was identified in all affected individuals."

Documents the recurrent founder allele - the population-genetic pattern that makes a few AR-ID genes locally common despite global rarity.

PMID:26842963 SUPPORT Human Clinical

"In this report, we present additional 15 individuals from 11 families (10 Saudis and 1 Emirati) who are homozygous for the same founder mutation."

Replication of the founder allele. PARTIAL with respect to THIS entry because the expanded phenotype places ADAT3-related disease at or beyond the non-syndromic boundary.

Homozygous copy-number variants

Gene: TUSC3 hgnc:30242 relationship_type: CAUSATIVE variant_origin: GERMLINE

Show evidence (2 references)

PMID:27431290 SUPPORT Human Clinical

"These included copy number variants in 14% (n=54, 15% are novel), and point mutations revealed by multi-gene panel and exome sequencing in the remaining 43% (1% were found to have Fragile-X)."

Quantifies the copy-number share of solved cases in a highly consanguineous intellectual-disability cohort.

PMID:18452889 SUPPORT Human Clinical

"Haplotype analyses and copy-number studies led to the identification of a homozygous deletion partly removing TUSC3 (N33) in all patients."

The archetypal homozygous-deletion lesion within this disease class.

💊

Medical Actions

6

Genetic Counselling and Recurrence-Risk Assessment

Action: genetic counseling Ontology label: Genetic Counseling NCIT:C15240

The highest-value intervention in this class, and the point at which it differs most sharply from the dominant class. Once biallelic causation is confirmed, sibling recurrence risk is 25% and carrier testing can be offered to at-risk relatives; in consanguineous kindreds where the allele is identical by descent, extended-family cascade testing and preconception or prenatal testing become possible. Counselling must be delivered with care for the cultural context in which consanguineous marriage occurs; the objective is informed reproductive choice. Two technical caveats: where multi-locus (oligogenic) causation is present, single-locus recurrence figures may be wrong; and for a gene with parent-of-origin expression bias the standard 25% figure may not apply.

Show evidence (5 references)

PMID:17718851 SUPPORT Human Clinical

"The future objective will be the development of diagnostic kits for molecular diagnosis in mentally retarded individuals in order to offer at-risk families pre-natal diagnosis to detect affected offspring."

States the counselling and prenatal-diagnosis rationale that molecular diagnosis in this class enables.

PMID:24176302 SUPPORT Human Clinical

"For autosomal recessive ID (ARID) the recurrence risk is high and, in populations with frequent parental consanguinity, ARID is the most common form of ID."

Establishes the high recurrence risk that makes counselling the primary actionable output of diagnosis.

PMID:24176302 SUPPORT Human Clinical

"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."

Frames diagnosis-enabled prevention as the field's stated objective for this disease class.

+ 2 more references

Early Intervention, Rehabilitation, and Special Education

Action: early intervention and special education Ontology label: Rehabilitation NCIT:C15315

No disease-modifying therapy exists for any AR-NSID subtype. Management is the standard, non-gene-specific developmental package: early intervention services, individualised special education, and habilitative therapies, directed at maximising adaptive function rather than altering the underlying lesion. The functional-outcome data available for this class (for example reading limited to first-grade level in the CRBN kindred) underline how much is at stake in delivering this well.

Target Phenotypes: Global developmental delay HP:0001263 Intellectual disability HP:0001249

Show evidence (2 references)

PMID:26503795 SUPPORT Human Clinical

"This approach has provided novel insights into the biological pathways underlying ID, improved the diagnostic process and offered new targets for therapy."

Supports the framing that therapy for intellectual disability remains prospective - genetics has provided TARGETS, not treatments. PARTIAL because the review does not evaluate educational or rehabilitative interventions.

PMID:15557513 SUPPORT Human Clinical

"In five affected adult individuals, writing skills were limited to signing their own names."

Documents the adult functional attainment that educational and habilitative intervention aims to improve.

Speech and Language Therapy

Action: speech therapy Ontology label: Speech Language Therapy NCIT:C159273

Directed at the delayed speech and language development that accompanies global developmental delay in this class. Symptomatic, with no subtype-specific modification.

Target Phenotypes: Global developmental delay HP:0001263

Show evidence (1 reference)

PMID:15557513 SUPPORT Human Clinical

"Developmental milestones were mildly delayed since early childhood."

Establishes the developmental delay that speech and language therapy addresses. PARTIAL because the source documents the indication, not the efficacy of the intervention.

Antiseizure Pharmacotherapy for Comorbid Epilepsy

Action: anticonvulsant therapy Ontology label: Anticonvulsant Therapy NCIT:C64172

Agent: anticonvulsant agent NCIT:C264

Indicated only for the subset of subtypes in which seizures occur, most clearly MBOAT7/MRT57. Choice of agent follows standard epilepsy practice; no AR-NSID subtype has an established gene-directed antiseizure strategy, so the agent is recorded at drug-class rather than individual-drug granularity.

Target Phenotypes: Seizure HP:0001250

Show evidence (1 reference)

PMID:27616480 SUPPORT Human Clinical

"Subjects presented with ID frequently accompanied by epilepsy and autistic features."

Establishes the epilepsy indication in this subtype. PARTIAL because the report documents the phenotype, not antiseizure treatment response.

PDE4 Inhibition in CC2D1A-Related Disease (preclinical)

Action: Pharmacotherapy NCIT:C15986

Agent: rolipram CHEBI:104872

The single mechanism-targeted therapeutic lead in this disease class, and PRECLINICAL ONLY - no human data exist and it must not be presented as a therapy. In Cc2d1a-deficient mice, loss of the CC2D1A restraint on PDE4D causes PDE4D hyperactivity, hippocampal cAMP depletion, and reduced CREB signalling; a PDE4 inhibitor restores spatial memory. Two hard caveats: the rescue is MALE-SPECIFIC (the underlying molecular deficit is absent in female mice, and the inhibitor has no effect in them), and the rescued readout is rodent spatial memory rather than the human cognitive construct.

Mechanism Target:

INHIBITS Impaired Synaptic Transmission and Plasticity — PDE4 inhibition restores hippocampal cAMP/CREB signalling that is depleted by PDE4D hyperactivity in CC2D1A deficiency, targeting the synaptic plasticity node rather than the upstream genetic lesion.

Show evidence (1 reference)

PMID:30732858 SUPPORT Model Organism

"Restoring PDE4D activity using an inhibitor rescues cognitive deficits in male mice but has no effect on female mice."

Demonstrates target engagement at the plasticity node with cognitive rescue. PARTIAL and MODEL_ORGANISM: mouse only, and male-specific.

Show evidence (2 references)

PMID:30732858 SUPPORT Model Organism

"We found that in Cc2d1a-deficient male mice PDE4D is hyperactive, leading to a reduction in cAMP response element binding protein signaling, but this molecular deficit is not present in female mice."

Establishes the drug target and the sex restriction. PARTIAL and MODEL_ORGANISM because the entire result is in mice.

PMID:30732858 SUPPORT Model Organism

"Cc2d1a-deficient male mice show a deficit in spatial memory, which is not present in Cc2d1a-deficient female mice."

The rescued phenotype is itself sex-restricted in the model, compounding the translational uncertainty.

Disease-Specific Metabolic Therapy After Reclassification

Action: dietary intervention Ontology label: Dietary Intervention NCIT:C15447

Initiate the specific metabolic therapy - dietary restriction, cofactor or vitamin supplementation, or substrate-lowering pharmacotherapy as dictated by the gene - in any child whose provisional AR-NSID label is overturned by sequencing in favour of a treatable inborn error of metabolism. In a large consanguineous cohort this applied to 5 of 152 families (PAH, CBS, MTHFR, CYP27A1, HIBCH). The therapy is disease-modifying for those children, which is what distinguishes this from every other entry in this section: AR-NSID proper has no disease-modifying treatment, so the single largest therapeutic gain available in this clinical population comes from correctly identifying the minority who do not actually have it. The complementary diagnostic step is curated under `diagnosis` as "Exclusion of Recognised Syndromic, Metabolic, and Treatable Causes"; this entry records the therapeutic action that follows.

Show evidence (1 reference)

PMID:28097321 SUPPORT Human Clinical

"In 5 of these families, potentially treatable disorders were diagnosed (mutations in PAH, CBS, MTHFR, CYP27A1, and HIBCH), and in 1 family, 2 disease-causing homozygous variants in different genes were identified."

Direct evidence that treatable metabolic disorders are found within cohorts ascertained as recessive intellectual disability, and therefore that this therapeutic action has a real indication in this population.

🔬

Diagnosis

3

Exome or Genome Sequencing with Autozygosity Mapping

Exome (increasingly genome) sequencing is the definitive diagnostic modality and is recommended as a first-tier test for unexplained neurodevelopmental disorders. Because there are no prevalent ARID genes, pathways, or protein complexes, targeted gene selection is explicitly not justified - broad sequencing rather than a small panel is the correct first step. In consanguineous families the yield is substantially raised by combining sequencing with homozygosity-by-descent mapping from SNP genotyping: filtering candidate variants to those inside runs of homozygosity shared by all affected siblings converts an unmanageable variant list into a short one. Trio or family-based testing remains valuable, because de novo dominant lesions occur in consanguineous families too and a homozygosity-only filtering strategy will miss them.

molecular genetic testing NCIT:C19770

Show evidence (8 references)

PMID:30459488 SUPPORT Human Clinical

"Thus, in a regular case, there is no reasoning for picking a few genes for a first diagnostic step, and a genetic diagnosis of ID in general, and ARID specifically, is better made using large panels or exome sequencing."

The explicit recommendation against targeted gene selection in this disease class, and for broad sequencing instead.

PMID:31182824 SUPPORT Human Clinical

"Our review demonstrates that ES consistently outperforms CMA for evaluation of unexplained NDDs. We propose a diagnostic algorithm placing ES at the beginning of the evaluation of unexplained NDDs."

Consensus recommendation placing exome sequencing first-tier for unexplained neurodevelopmental disorders.

PMID:31182824 SUPPORT Human Clinical

"Yield of ES was 36% overall, 31% for isolated NDD, and 53% for the NDD plus associated conditions."

Quantifies yield, including the ISOLATED (non-syndromic) figure of 31% that is most relevant here and is notably LOWER than for syndromic cases. PARTIAL because the meta-analysis is not restricted to consanguineous or recessive cases.

+ 5 more references

Chromosomal Microarray

Chromosomal microarray remains a standard early test for unexplained developmental delay or intellectual disability and is not made redundant by sequencing in this class: several AR-NSID lesions are homozygous deletions (TUSC3 is the archetype) that a copy-number-naive exome pipeline can miss, copy-number variants accounted for 14% of solved cases in a large consanguineous cohort, and in consanguineous families the array simultaneously supplies the genome-wide homozygosity map used to prioritise sequence variants.

chromosomal microarray NCIT:C19770

Show evidence (3 references)

PMID:20466091 SUPPORT Human Clinical

"Available evidence strongly supports the use of CMA in place of G-banded karyotyping as the first-tier cytogenetic diagnostic test for patients with DD/ID, ASD, or MCA."

Establishes chromosomal microarray as the first-tier cytogenetic test in this clinical population.

PMID:27431290 SUPPORT Human Clinical

"These included copy number variants in 14% (n=54, 15% are novel), and point mutations revealed by multi-gene panel and exome sequencing in the remaining 43% (1% were found to have Fragile-X)."

Quantifies the copy-number contribution that justifies retaining microarray alongside sequencing.

PMID:18452889 SUPPORT Human Clinical

"Haplotype analyses and copy-number studies led to the identification of a homozygous deletion partly removing TUSC3 (N33) in all patients."

Worked example of an AR-NSID lesion that is a homozygous deletion, requiring copy-number analysis to detect.

Exclusion of Recognised Syndromic, Metabolic, and Treatable Causes

A non-syndromic designation is a diagnosis of exclusion and requires expert dysmorphology assessment plus exclusion of recognised aetiologies (fragile X, chromosomal disorders, congenital infection, prenatal exposures, acquired perinatal injury, and inborn errors of metabolism). This is not merely nosological housekeeping: exome sequencing in consanguineous intellectual-disability cohorts uncovers TREATABLE inborn errors of metabolism in a small but clinically decisive minority, so a case provisionally labelled non-syndromic must not bypass metabolic evaluation. Developmental regression, in particular, argues against this diagnosis and toward a degenerative or metabolic alternative.

disease screening NCIT:C15419

Show evidence (4 references)

PMID:28097321 SUPPORT Human Clinical

"In 5 of these families, potentially treatable disorders were diagnosed (mutations in PAH, CBS, MTHFR, CYP27A1, and HIBCH), and in 1 family, 2 disease-causing homozygous variants in different genes were identified."

Directly documents treatable metabolic disorders surfacing within a consanguineous intellectual-disability cohort - the clinical reason this exclusion step matters.

PMID:15557513 SUPPORT Human Clinical

"Affected individuals had normal laboratory studies on the following: MRI of the brain, high-resolution (ISCN G-banding 550 resolution) cytogenetic studies, DNA testing for a CCG repeat expansion in fragile X, fasting plasma amino acids, urine amino acids and organic acids, pyruvate, lactate, and ammonia."

Specifies the exclusion workup that operationally assigns a family to the non-syndromic category.

PMID:15557513 SUPPORT Human Clinical

"All individuals had normal newborn screening examinations for hypothyroidism, galactosemia, maple syrup urine disease, phenylketonuria, and biotinidase."

Documents the metabolic exclusions performed before assigning a non-syndromic recessive label.

+ 1 more reference

📊

Prevalence

3

Worldwide (all-cause intellectual disability, the denominator class)

Point Prevalence 1037.0 per 100,000 >1 in 1,000

Prevalence of intellectual disability from all causes, NOT of the autosomal recessive non-syndromic subset specifically. AR-NSID is a large but unquantified fraction of this denominator; no direct population prevalence estimate for the class exists, because ascertainment requires a molecular diagnosis that most affected individuals never receive. See the `arnsid_no_class_specific_prevalence` discussion.

Show evidence (2 references)

PMID:21236634 SUPPORT Human Clinical

"The prevalence of intellectual disability across all 52 studies included in the meta-analysis was 10.37/1000 population."

Provides the all-cause intellectual disability denominator from which the normalised rate is derived (10.37/1000 = 1037 per 100,000).

PMID:28397838 SUPPORT Human Clinical

"Approximately 1% of the global population is affected by intellectual disability (ID), and the majority receive no molecular diagnosis."

Independent statement of the same denominator plus the ascertainment caveat that prevents a class-specific prevalence figure.

Consanguineous populations (Iran, Pakistan, Arabian peninsula and comparable settings)

Unknown Common

Qualitative class only. In populations with frequent parental consanguinity recessive gene defects are the leading genetic cause of intellectual disability, and in a highly consanguineous clinical cohort 81% of identified point mutations were recessive. No numerator/denominator estimate specific to the NON-SYNDROMIC recessive subset has been published; the supporting cohorts are clinic-ascertained multiplex families, not population samples.

Show evidence (3 references)

PMID:29302074 SUPPORT Human Clinical

"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."

Supports the COMMON qualitative class in the consanguineous stratum.

PMID:27431290 SUPPORT Human Clinical

"The identified point mutations were mostly recessive (n=117, 81%), consistent with the high consanguinity of the study cohort"

Quantifies the recessive share of solved cases in a highly consanguineous clinical cohort.

PMID:24176302 SUPPORT Human Clinical

"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."

Important qualifier: the class is not confined to consanguineous populations. PARTIAL because the claim is a review conclusion rather than a population prevalence survey.

🔀

Differential Diagnoses

5

Conditions with similar clinical presentations that must be differentiated from Autosomal Recessive Non-Syndromic Intellectual Disability:

Autosomal Dominant Non-Syndromic Intellectual Disability MONDO:0015802

Overlapping Features The sibling class and the principal differential in a simplex case. Both present as isolated intellectual disability without a syndromic gestalt, and the two cannot be distinguished clinically - only by the molecular result and the pedigree. Distinguishing them matters enormously for counselling: recurrence risk is 25% for recessive versus low (mosaicism-limited) for a de novo dominant lesion.

Distinguishing Features

Dominant cases are typically simplex with a heterozygous de novo variant in a constraint-intolerant gene and no consanguinity.
Recessive cases are more often multiplex with affected siblings, homozygous variants inside a run of homozygosity, and frequently consanguineous parents.
Only trio or family-based molecular testing resolves the two; de novo dominant lesions also occur within consanguineous families, so a consanguineous pedigree does not settle the question.

Show evidence (2 references)

PMID:29302074 SUPPORT Human Clinical

"Yet, compared to autosomal dominant de novo mutations, which are the predominant cause of ID in Western countries, the identification of AR-ID genes has lagged behind."

States the population-dependent split between the recessive and de novo dominant classes that drives the differential.

PMID:28097321 SUPPORT Human Clinical

"50 genes already established in neurodevelopmental disorders (46 autosomal recessive, 2 X-linked, and 2 de novo)"

Shows de novo dominant diagnoses arising inside a consanguineous cohort - the practical reason a recessive assumption must not be allowed to filter them out.

Syndromic Intellectual Disability

Overlapping Features Any recognised syndrome in which intellectual disability is accompanied by a characteristic pattern of dysmorphic, malformative, metabolic, or neuroimaging features. The boundary is genuinely unstable in this class: several MRT-series genes (NSUN2, MAN1B1, ADAT3) have been reclassified as syndromic once larger series were phenotyped, and the class review notes that many ARID genes are pleiotropic with broad phenotypic spectra. A non-syndromic label assigned on a small cohort should be treated as provisional.

Distinguishing Features

Presence of a consistent extra-neural gestalt on expert dysmorphology examination.
A recognisable neuroimaging malformation rather than subtle volume or connectivity change.
A diagnostic metabolic or biochemical abnormality.
Absence of all of these on adequate assessment defines the non-syndromic category, but adequacy of assessment scales with cohort size and follow-up duration.

Show evidence (3 references)

PMID:30459488 SUPPORT Human Clinical

"In addition, in the last few months, evidence has been growing that many ARID genes are pleiotropic and that the resulting phenotypes may have a broad spectrum."

Class-level statement that the syndromic/non-syndromic boundary is unreliable for recessive intellectual-disability genes.

PMID:26842963 SUPPORT Human Clinical

"ADAT3-related intellectual disability is an important recognizable cause of intellectual disability in Arabia."

Worked example of a recessive intellectual-disability gene becoming a recognisable (syndromic) entity after further phenotyping.

PMID:22541559 SUPPORT Human Clinical

"suggesting that mutations in this gene might even induce a syndromic form of ID"

Shows the same boundary problem raised prospectively by the discovering authors.

Treatable Inborn Errors of Metabolism

Overlapping Features Metabolic disorders such as phenylketonuria, homocystinuria, MTHFR deficiency, cerebrotendinous xanthomatosis, and HIBCH deficiency are themselves autosomal recessive and are enriched in exactly the same consanguineous populations. When mild or late-treated they can present as apparently isolated intellectual disability. This is the one part of the differential where the distinction changes management immediately.

Distinguishing Features

Diagnostic biochemical abnormality on plasma amino acids, urine organic acids, homocysteine, acylcarnitines, or sterol profiling.
Often a history of episodic decompensation or a progressive rather than static course.
A specific dietary or pharmacological treatment exists, so the distinction is management-changing.
Newborn screening excludes several but is not universally available in the populations where AR-NSID is concentrated.

Show evidence (2 references)

PMID:28097321 SUPPORT Human Clinical

"In 5 of these families, potentially treatable disorders were diagnosed (mutations in PAH, CBS, MTHFR, CYP27A1, and HIBCH), and in 1 family, 2 disease-causing homozygous variants in different genes were identified."

Quantifies how often treatable metabolic disease is found in cohorts ascertained as recessive intellectual disability.

PMID:15557513 SUPPORT Human Clinical

"All individuals had normal newborn screening examinations for hypothyroidism, galactosemia, maple syrup urine disease, phenylketonuria, and biotinidase."

Illustrates the metabolic exclusions performed before assigning a non-syndromic recessive label.

X-Linked Intellectual Disability

Overlapping Features The historically better-characterised class of inherited non-syndromic intellectual disability, with more than 90 genes known well before the autosomal recessive class was opened up. In a family with affected males only, X-linkage and autosomal recessive inheritance can be indistinguishable on the pedigree, and fragile X remains a mandatory exclusion.

Distinguishing Features

Male-only affectation with transmission through unaffected carrier females and absence of male-to-male transmission favours X-linkage.
Affected siblings of both sexes with consanguineous parents favours autosomal recessive inheritance.
Large consanguineous cohort studies still recover X-linked genes alongside autosomal recessive ones, so an assumption of autosomal recessive inheritance should not restrict variant filtering.
FMR1 repeat testing is a standard first-tier exclusion regardless of pedigree structure.

Show evidence (2 references)

PMID:21937992 SUPPORT Human Clinical

"More than 90 different gene defects have been identified for X-chromosome-linked intellectual disability alone, but research into the more frequent autosomal forms of intellectual disability is still in its infancy."

Establishes the relative maturity of the X-linked class and the size of its gene set.

PMID:29302074 SUPPORT Human Clinical

"In 219 of these, we found likely causative variants, involving 77 known and 77 novel AR-ID (candidate) genes, 21 X-linked genes, as well as 9 genes previously implicated in diseases other than ID."

Shows X-linked genes being recovered within a consanguineous cohort ascertained for recessive disease - the practical reason not to pre-filter on inheritance mode.

Progressive or Degenerative Neurological Disease

Overlapping Features Neurodegenerative and neurometabolic disorders can present in early childhood with apparent developmental delay before the progressive nature is evident. Because AR-NSID is a static developmental encephalopathy - a mis-built rather than a damaged circuit, with no degenerative, ischaemic, or inflammatory arm - loss of previously acquired skills is the single cleanest discriminator against this diagnosis.

Distinguishing Features

Developmental regression (loss of acquired skills) effectively excludes AR-NSID.
Progressive neurological signs, evolving neuroimaging abnormality, or a deteriorating clinical course argue for a degenerative alternative.
In AR-NSID the gap versus peers widens with age because the developmental trajectory is shallower, not because skills are lost.

Show evidence (1 reference)

PMID:15557513 SUPPORT Human Clinical

"Detailed physical and neurologic examinations in all study participants did not show microcephaly, organomegaly, phakomata, spasticity, weakness, neuropathy, visual deficits, feeding problems, or psychiatric disorders."

Documents the absence of the progressive neurological signs that would point to a degenerative alternative, in the best-characterised AR-NSID kindred.

{ }

Source YAML

click to show

name: Autosomal Recessive Non-Syndromic Intellectual Disability
creation_date: "2026-08-01T00:00:00Z"
description: >-
  Autosomal recessive non-syndromic intellectual disability (AR-NSID / NS-ARID;
  the OMIM "mental retardation / intellectual developmental disorder, autosomal
  recessive" MRT series) is a genetically extremely heterogeneous class of
  intellectual disability in which impaired intellectual functioning and
  adaptive behaviour arise from biallelic (usually homozygous) loss-of-function
  variants in an autosomal gene, without a consistent pattern of associated
  dysmorphic, malformative, metabolic, or neuroimaging features that would
  define a clinically recognisable syndrome. It is the mechanistic and
  epidemiological mirror image of the autosomal dominant class: where AD-NSID is
  dominated by sporadic de novo heterozygous lesions in constraint-intolerant
  genes, AR-NSID is inherited, carries a 25% sibling recurrence risk, and is
  concentrated in populations with frequent parental consanguinity - roughly a
  seventh of the world population, in which recessive gene defects are the
  leading genetic cause of intellectual disability. Because each gene accounts
  for only a handful of families worldwide, the class was intractable to
  conventional linkage and was opened up only by homozygosity (autozygosity)
  mapping in large consanguineous pedigrees combined with exome sequencing: the
  first gene, PRSS12/neurotrypsin, was reported in 2002, only four genes were
  known by 2008, and large consanguineous cohorts have since implicated many
  hundreds, against an estimated total of more than 2,500 autosomal
  intellectual-disability genes, most of them recessive. Critically, the causal
  genes do NOT converge on a prevalent pathway or protein complex - the
  authoritative review of the class states there are no prevalent ARID genes,
  pathways, or protein complexes - but they do fall repeatedly into a limited
  set of recurring cellular themes: synaptic transmission and proteolysis
  (GRIK2, PRSS12), transcriptional regulation (MED23), RNA modification
  (NSUN2, ADAT3), protein glycosylation (TUSC3, MAN1B1), vesicle trafficking
  with NF-kB signalling (TRAPPC9, CC2D1A), protein turnover (CRBN), and membrane
  lipid remodelling (MBOAT7). The disorder is a static (non-progressive)
  developmental encephalopathy - a mis-built rather than a damaged circuit - so
  developmental regression argues against the diagnosis. As in the dominant
  class the non-syndromic label is unstable: several MRT-series entities (NSUN2,
  MAN1B1, ADAT3) acquired recognisable dysmorphic or systemic features once
  larger series were phenotyped.
category: Genetic
disease_term:
  preferred_term: autosomal recessive non-syndromic intellectual disability
  term:
    id: MONDO:0019502
    label: autosomal recessive non-syndromic intellectual disability
parents:
- Non-Syndromic Intellectual Disability
- Intellectual Disability, Autosomal Recessive
synonyms:
- non-syndromic intellectual disability, autosomal recessive
- autosomal recessive non-syndromic mental retardation
- mental retardation, autosomal recessive (MRT)
- AR-NSID
- NS-ARID
- ARNSMR

references:
- reference: PMID:30459488
  title: "Genetics of autosomal recessive intellectual disability."
- reference: PMID:24176302
  title: "Genetics of recessive cognitive disorders."
- reference: PMID:21937992
  title: "Deep sequencing reveals 50 novel genes for recessive cognitive disorders."
- reference: PMID:29302074
  title: "Genetics of intellectual disability in consanguineous families."
- reference: PMID:28397838
  title: "Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families."
- reference: PMID:27457812
  title: "Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes for recessive intellectual disability."
- reference: PMID:17718851
  title: "Genetics of autosomal recessive non-syndromic mental retardation: recent advances."
- reference: PMID:26503795
  title: "Genetic studies in intellectual disability and related disorders."

inheritance:
- name: Autosomal recessive
  description: >-
    Disease requires two damaging alleles at the same autosomal locus. In the
    consanguineous pedigrees in which most AR-NSID genes were discovered the two
    alleles are identical by descent (homozygous), so the affected haplotype
    falls inside a run of homozygosity - the property that makes autozygosity
    mapping the discovery method of choice. Compound heterozygosity is the
    corresponding mechanism in outbred families. Heterozygous carriers,
    including the obligate-carrier parents, are unaffected. Recurrence risk for
    full siblings is 25%, an order of magnitude higher than for the sporadic de
    novo dominant class, which is why molecular diagnosis has a far larger
    counselling and prevention payoff here.
  inheritance_term:
    preferred_term: Autosomal recessive inheritance
    term:
      id: HP:0000007
      label: Autosomal recessive inheritance
  penetrance: COMPLETE
  expressivity: VARIABLE
  evidence:
  - reference: PMID:24176302
    reference_title: "Genetics of recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      For autosomal recessive ID (ARID) the recurrence risk is high and, in
      populations with frequent parental consanguinity, ARID is the most common
      form of ID.
    explanation: >-
      States both defining features of this class relative to the dominant
      class: high recurrence risk, and concentration in consanguineous
      populations.
  - reference: PMID:29302074
    reference_title: "Genetics of intellectual disability in consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      Quantifies the population denominator in which autosomal recessive
      inheritance is the dominant mode for intellectual disability.
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All obligate carriers of this family were heterozygous, but none of 192
      unrelated healthy individuals from the same population carried this
      deletion.
    explanation: >-
      Worked demonstration of the recessive model in a single pedigree:
      unaffected obligate carriers are heterozygous and the allele is absent
      from population controls.
  - reference: PMID:17718851
    reference_title: "Genetics of autosomal recessive non-syndromic mental retardation: recent advances."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All of the mutations in the ARNSMR-causing genes are protein truncating,
      indicating a putative severe loss-of-function effect.
    explanation: >-
      Characterises the allelic mechanism as biallelic loss of function rather
      than dominant-negative or gain-of-function.
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The risk for ARID in children of first cousins or closer is a magnitude
      higher than for children of unrelated parents.
    explanation: >-
      Quantifies the effect of parental relatedness on recessive
      intellectual-disability risk.
- name: Oligogenic / multi-locus inheritance
  description: >-
    A minority of consanguineous families carry pathogenic variants at more than
    one locus, so the phenotype is the sum of two or more recessive lesions
    rather than of a single gene. In the largest systematic look at this, eight
    of 121 large Pakistani families (about 6.6%) segregated multiple pathogenic
    variants across 19 genes. This matters practically: stopping at the first
    convincing homozygous variant can leave part of the phenotype unexplained,
    and recurrence-risk counselling based on a single locus may be wrong.
  inheritance_term:
    preferred_term: Oligogenic inheritance
    term:
      id: HP:0010983
      label: Oligogenic inheritance
  evidence:
  - reference: PMID:27457812
    reference_title: "Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes for recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In another eight families segregation of multiple pathogenic variants was
      observed, affecting 19 genes that were either known or are novel
      candidates for ID.
    explanation: >-
      Direct documentation of multi-locus (oligogenic) causation within a
      consanguineous recessive intellectual-disability cohort.

prevalence:
- population: Worldwide (all-cause intellectual disability, the denominator class)
  measure_type: POINT_PREVALENCE
  prevalence_class: ABOVE_1_IN_1000
  rate_per_100000: 1037.0
  notes: >-
    Prevalence of intellectual disability from all causes, NOT of the autosomal
    recessive non-syndromic subset specifically. AR-NSID is a large but
    unquantified fraction of this denominator; no direct population prevalence
    estimate for the class exists, because ascertainment requires a molecular
    diagnosis that most affected individuals never receive. See the
    `arnsid_no_class_specific_prevalence` discussion.
  evidence:
  - reference: PMID:21236634
    reference_title: "Prevalence of intellectual disability: a meta-analysis of population-based studies."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The prevalence of intellectual disability across all 52 studies included
      in the meta-analysis was 10.37/1000 population.
    explanation: >-
      Provides the all-cause intellectual disability denominator from which the
      normalised rate is derived (10.37/1000 = 1037 per 100,000).
  - reference: PMID:28397838
    reference_title: "Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Approximately 1% of the global population is affected by intellectual
      disability (ID), and the majority receive no molecular diagnosis.
    explanation: >-
      Independent statement of the same denominator plus the ascertainment
      caveat that prevents a class-specific prevalence figure.
- population: Outbred (non-consanguineous) populations, as a share of affected children
  measure_type: UNKNOWN
  prevalence_class: UNKNOWN
  notes: >-
    Not a population rate but a share of the affected-child cohort: autosomal
    recessive intellectual disability (ARID, syndromic and non-syndromic
    combined) is estimated at about 10% of affected children in an outbred
    population, with an upward tendency as unsolved cases are resolved. Recorded
    as a share rather than a rate because the source expresses it that way;
    the non-syndromic subset is a fraction of this and cannot be separated.
  evidence:
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The prevalence of ARID in a cohort of affected children of an outbred
      population is estimated to be about 10%, with an upward tendency in still
      unclarified cases.
    explanation: >-
      The only published quantitative estimate of the recessive share in an
      outbred setting.
- population: Consanguineous populations (Iran, Pakistan, Arabian peninsula and comparable settings)
  measure_type: UNKNOWN
  prevalence_class: COMMON
  notes: >-
    Qualitative class only. In populations with frequent parental consanguinity
    recessive gene defects are the leading genetic cause of intellectual
    disability, and in a highly consanguineous clinical cohort 81% of identified
    point mutations were recessive. No numerator/denominator estimate specific
    to the NON-SYNDROMIC recessive subset has been published; the supporting
    cohorts are clinic-ascertained multiplex families, not population samples.
  evidence:
  - reference: PMID:29302074
    reference_title: "Genetics of intellectual disability in consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      Supports the COMMON qualitative class in the consanguineous stratum.
  - reference: PMID:27431290
    reference_title: "Clinical genomics expands the morbid genome of intellectual disability and offers a high diagnostic yield."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The identified point mutations were mostly recessive (n=117, 81%),
      consistent with the high consanguinity of the study cohort
    explanation: >-
      Quantifies the recessive share of solved cases in a highly consanguineous
      clinical cohort.
  - reference: PMID:24176302
    reference_title: "Genetics of recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      Important qualifier: the class is not confined to consanguineous
      populations. PARTIAL because the claim is a review conclusion rather than
      a population prevalence survey.

has_subtypes:
- name: MRT1
  display_name: Intellectual disability, autosomal recessive 1 (PRSS12)
  classification: molecular
  subtype_term:
    preferred_term: intellectual disability, autosomal recessive 1
    term:
      id: MONDO:0009580
      label: intellectual disability, autosomal recessive 1
  genes:
  - preferred_term: PRSS12
    term:
      id: hgnc:9477
      label: PRSS12
  description: >-
    The first AR-NSID gene identified (2002) and the anchor entry of the OMIM
    phenotypic series. Biallelic truncation of PRSS12, encoding the neuronal
    serine protease neurotrypsin, which localises to presynaptic terminals at
    the synaptic cleft. Places the founding member of the class in the
    synaptic-proteolysis theme.
  evidence:
  - reference: PMID:12459588
    reference_title: "Truncating neurotrypsin mutation in autosomal recessive nonsyndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A 4-base pair deletion in the neuronal serine protease neurotrypsin gene
      was associated with autosomal recessive nonsyndromic mental retardation
      (MR).
    explanation: >-
      Establishes the PRSS12/neurotrypsin subtype and its biallelic truncating
      mechanism.
- name: MRT2
  display_name: Intellectual disability, autosomal recessive 2 (CRBN)
  classification: molecular
  subtype_term:
    preferred_term: intellectual disability, autosomal recessive 2
    term:
      id: MONDO:0011828
      label: intellectual disability, autosomal recessive 2
  genes:
  - preferred_term: CRBN
    term:
      id: hgnc:30185
      label: CRBN
  description: >-
    Homozygous nonsense variant (R419X) in CRBN, encoding cereblon, an
    ATP-dependent Lon protease, in a large North American sectarian kindred.
    Distinctive within the class for its MILD severity band (IQ 50-70) and for
    an explicitly documented absence of dysmorphism, microcephaly, autistic
    features, and neuroimaging abnormality - making it one of the purest
    non-syndromic exemplars. Places protein turnover in the theme set.
  evidence:
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A nonsense mutation causing a premature stop codon in a novel gene
      (cereblon; CRBN) was identified that encodes for an ATP-dependent Lon
      protease.
    explanation: >-
      Identifies the causal CRBN lesion defining the MRT2 subtype.
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Previously, a disease locus was mapped for a mild type of nonsyndromic
      mental retardation (IQ between 50 and 70) to a 4.2-MB interval on
      chromosome 3p25-pter in a large kindred.
    explanation: >-
      Documents the mild severity band that distinguishes MRT2 from the
      moderate-to-severe presentations typical of the class.
- name: MRT3
  display_name: Intellectual disability, autosomal recessive 3 (CC2D1A)
  classification: molecular
  subtype_term:
    preferred_term: intellectual disability, autosomal recessive 3
    term:
      id: MONDO:0012037
      label: intellectual disability, autosomal recessive 3
  genes:
  - preferred_term: CC2D1A
    term:
      id: hgnc:30237
      label: CC2D1A
  description: >-
    Recurrent protein-truncating founder variant in CC2D1A on 19p13.12,
    identified in nine consanguineous families with SEVERE non-syndromic
    intellectual disability. CC2D1A is a positive regulator of the IKK/NF-kB
    cascade and also restrains PDE4D-dependent cAMP hydrolysis. It is the only
    subtype in this class with a preclinical pharmacological rescue (PDE4
    inhibition in male mice) and is therefore the one druggable-mechanism lead.
  evidence:
  - reference: PMID:16033914
    reference_title: "The CC2D1A, a member of a new gene family with C2 domains, is involved in autosomal recessive non-syndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A protein truncating mutation was identified in the gene CC2D1A in nine
      consanguineous families with severe autosomal recessive NSMR.
    explanation: >-
      Establishes the CC2D1A subtype, its founder truncating allele, and its
      severe phenotype band.
  - reference: PMID:16033914
    reference_title: "The CC2D1A, a member of a new gene family with C2 domains, is involved in autosomal recessive non-syndromic mental retardation."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      CC2D1A is a putative signal transducer participating in positive
      regulation of I-kappaB kinase/NFkappaB cascade.
    explanation: >-
      Assigns the subtype to the NF-kB signalling theme shared with TRAPPC9.
- name: MRT5
  display_name: Intellectual disability, autosomal recessive 5 (NSUN2)
  classification: molecular
  subtype_term:
    preferred_term: intellectual disability, autosomal recessive 5
    term:
      id: MONDO:0012613
      label: intellectual disability, autosomal recessive 5
  genes:
  - preferred_term: NSUN2
    term:
      id: hgnc:25994
      label: NSUN2
  description: >-
    Homozygous loss of the tRNA methyltransferase NSUN2 in Iranian and Kurdish
    consanguineous families, mapping to the previously described MRT5 locus.
    Places RNA (tRNA) modification in the theme set, with a Drosophila
    short-term-memory model providing cross-species functional support. Also the
    clearest example within the class of boundary instability: the authors
    themselves note accompanying facial dysmorphism and raise the possibility
    that the entity is syndromic.
  evidence:
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      that cause a loss of the tRNA-methyltransferase-encoding NSUN2 main
      transcript in homozygotes
    explanation: >-
      Identifies the NSUN2 loss-of-function mechanism defining MRT5.
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      When the Drosophila melanogaster NSUN2 ortholog was deleted, severe
      short-term-memory (STM) deficits were observed; STM could be rescued by
      re-expression of the wild-type protein in the nervous system.
    explanation: >-
      Cross-species functional support that NSUN2 loss impairs memory. PARTIAL
      and MODEL_ORGANISM because the readout is Drosophila short-term memory,
      not the human cognitive phenotype.
- name: MRT6
  display_name: Intellectual disability, autosomal recessive 6 (GRIK2)
  classification: molecular
  subtype_term:
    preferred_term: intellectual disability, autosomal recessive 6
    term:
      id: MONDO:0012614
      label: intellectual disability, autosomal recessive 6
  genes:
  - preferred_term: GRIK2
    term:
      id: hgnc:4580
      label: GRIK2
  description: >-
    Complex homozygous rearrangement of GRIK2 (GLUR6), the kainate-type
    ionotropic glutamate receptor subunit, in a large consanguineous Iranian
    family with moderate-to-severe non-syndromic intellectual disability. The
    predicted product lacks the first ligand-binding domain, the adjacent
    transmembrane domain, and the pore loop, giving complete loss of function
    confirmed electrophysiologically. The most direct link in the class between
    the disease and excitatory synaptic transmission.
  evidence:
  - reference: PMID:17847003
    reference_title: "A defect in the ionotropic glutamate receptor 6 gene (GRIK2) is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      cosegregates with moderate-to-severe nonsyndromic autosomal recessive
      mental retardation in a large, consanguineous Iranian family
    explanation: >-
      Establishes the GRIK2 subtype and its moderate-to-severe non-syndromic
      phenotype in a consanguineous pedigree.
  - reference: PMID:17847003
    reference_title: "A defect in the ionotropic glutamate receptor 6 gene (GRIK2) is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      The predicted gene product lacks the first ligand-binding domain, the
      adjacent transmembrane domain, and the putative pore loop, suggesting a
      complete loss of function of the GLU(K6) protein, which is supported by
      electrophysiological data.
    explanation: >-
      Functional confirmation of complete loss of receptor function, anchoring
      the synaptic-transmission mechanism.
- name: MRT7
  display_name: Intellectual disability, autosomal recessive 7 (TUSC3)
  classification: molecular
  subtype_term:
    preferred_term: intellectual disability, autosomal recessive 7
    term:
      id: MONDO:0012615
      label: intellectual disability, autosomal recessive 7
  genes:
  - preferred_term: TUSC3
    term:
      id: hgnc:30242
      label: TUSC3
  description: >-
    Homozygous deletion of TUSC3 (N33) on chromosome 8, encoding a subunit of
    the ER-bound oligosaccharyltransferase complex. Mechanistically important
    because it shows a defect in N-glycosylation - a pathway whose other defects
    produce florid multisystem congenital disorders of glycosylation -
    presenting as ISOLATED cognitive impairment, and because it was the first
    gene in the class with mutations reported in more than one family.
  evidence:
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Haplotype analyses and copy-number studies led to the identification of a
      homozygous deletion partly removing TUSC3 (N33) in all patients.
    explanation: >-
      Establishes the biallelic TUSC3 deletion defining MRT7.
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Our data suggest that in contrast to other genetic defects of
      glycosylation, inactivation of TUSC3 causes nonsyndromic MR
    explanation: >-
      Directly supports the non-syndromic designation and contrasts it with the
      syndromic congenital disorders of glycosylation.
- name: MRT13
  display_name: Intellectual disability, autosomal recessive 13 (TRAPPC9)
  classification: molecular
  subtype_term:
    preferred_term: intellectual disability, autosomal recessive 13
    term:
      id: MONDO:0013173
      label: intellectual disability, autosomal recessive 13
  genes:
  - preferred_term: TRAPPC9
    term:
      id: hgnc:30832
      label: TRAPPC9
  description: >-
    Biallelic truncating variants in TRAPPC9 (NIBP) on 8q24, reported
    simultaneously by two groups in 2009 in Pakistani, Iranian, and Israeli Arab
    pedigrees and in many families since. TRAPPC9 is a vesicle-trafficking
    (TRAPP complex) subunit that also activates NF-kB via IKK-beta, converging
    with CC2D1A. It sits at the edge of the non-syndromic definition: variable
    postnatal microcephaly and mild cerebral white matter hypoplasia are common
    but not obligate, and obesity is described in the fuller phenotype. It is
    also the one AR-NSID gene with an established parent-of-origin expression
    bias (see `mechanistic_hypotheses`).
  evidence:
  - reference: PMID:20004765
    reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Within this region, we identified a truncating homozygous mutation, R475X,
      in exon 7 of the gene TRAPPC9.
    explanation: >-
      Establishes the TRAPPC9 truncating mechanism defining MRT13.
  - reference: PMID:20004765
    reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Brain magnetic resonance imaging of affected individuals indicates the
      presence of mild cerebral white matter hypoplasia. Microcephaly is present
      in some but not all affected individuals.
    explanation: >-
      Documents the incompletely penetrant accessory features that make this
      subtype a borderline case for the non-syndromic designation.
  - reference: PMID:20004763
    reference_title: "A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      TRAPPC9 is highly expressed in the postmitotic neurons of the cerebral
      cortex, and MRI analysis of affected patients shows defects in axonal
      connectivity.
    explanation: >-
      Independent replication in a separate pedigree, plus the connectivity
      phenotype linking the molecular lesion to cortical circuit formation.
- name: MRT15
  display_name: Intellectual disability, autosomal recessive 15 / Rafiq syndrome (MAN1B1)
  classification: molecular
  subtype_term:
    preferred_term: Rafiq syndrome
    term:
      id: MONDO:0013624
      label: Rafiq syndrome
  genes:
  - preferred_term: MAN1B1
    term:
      id: hgnc:6823
      label: MAN1B1
  description: >-
    Biallelic variants in MAN1B1 (alpha-1,2-mannosidase) at 9q34.3, a second
    glycosylation entry point alongside TUSC3. Notable as one of the few genes
    in the class with an elevated mutation frequency across populations rather
    than a single-family report. The MONDO label "Rafiq syndrome" reflects the
    subsequent recognition of dysmorphic features in some families - within the
    original report one family already showed additional clinical features while
    three others were strictly non-syndromic, so this subtype straddles the
    category boundary.
  evidence:
  - reference: PMID:21763484
    reference_title: "Mutations in the alpha 1,2-mannosidase gene, MAN1B1, cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      We have used genome-wide genotyping to identify an overlapping
      homozygosity-by-descent locus on chromosome 9q34.3 (MRT15) in four
      consanguineous families affected by nonsyndromic autosomal-recessive
      intellectual disability (NS-ARID) and one in which the patients show
      additional clinical features.
    explanation: >-
      Establishes the MRT15 locus and simultaneously documents the mixed
      syndromic/non-syndromic presentation within the founding cohort.
  - reference: PMID:21763484
    reference_title: "Mutations in the alpha 1,2-mannosidase gene, MAN1B1, cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      MAN1B1 is one of the few NS-ARID genes with an elevated mutation frequency
      in patients with NS-ARID from different populations.
    explanation: >-
      Supports the claim that this subtype is relatively over-represented in the
      class, unlike the typical single-family genes.
- name: MRT18
  display_name: Intellectual disability, autosomal recessive 18 (MED23)
  classification: molecular
  subtype_term:
    preferred_term: intellectual disability, autosomal recessive 18
    term:
      id: MONDO:0013651
      label: intellectual disability, autosomal recessive 18
  genes:
  - preferred_term: MED23
    term:
      id: hgnc:2372
      label: MED23
  description: >-
    Homozygous missense variant (p.R617Q) in MED23, a subunit of the Mediator
    coactivator complex. Unusually for the class the lesion is a hypomorphic
    missense rather than a truncation, and the demonstrated defect is specific -
    failure of JUN and FOS immediate-early-gene induction rather than global
    transcriptional collapse. Places activity-dependent transcription in the
    theme set.
  evidence:
  - reference: PMID:21868677
    reference_title: "MED23 mutation links intellectual disability to dysregulation of immediate early gene expression."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Here, we report a missense mutation (p. R617Q) in MED23 that cosegregates
      with nonsyndromic autosomal recessive intellectual disability.
    explanation: >-
      Establishes the MED23 subtype and its non-syndromic recessive
      cosegregation.
  - reference: PMID:21868677
    reference_title: "MED23 mutation links intellectual disability to dysregulation of immediate early gene expression."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      This mutation specifically impaired the response of JUN and FOS immediate
      early genes (IEGs) to serum mitogens by altering the interaction between
      enhancer-bound transcription factors (TCF4 and ELK1, respectively) and
      Mediator.
    explanation: >-
      Functional mechanism in patient-derived cells: selective loss of
      immediate-early-gene induction, the transcriptional arm of
      activity-dependent plasticity.
- name: MRT57
  display_name: Intellectual disability, autosomal recessive 57 (MBOAT7)
  classification: molecular
  subtype_term:
    preferred_term: intellectual disability, autosomal recessive 57
    term:
      id: MONDO:0014962
      label: intellectual disability, autosomal recessive 57
  genes:
  - preferred_term: MBOAT7
    term:
      id: hgnc:15505
      label: MBOAT7
  description: >-
    Homozygous inactivating variants in MBOAT7 (LPIAT1), the acyltransferase
    that loads arachidonic acid onto lysophosphatidylinositol, identified in six
    consanguineous families out of a >5,000-family neurodevelopmental cohort.
    Places membrane phospholipid remodelling in the theme set. Sits at the
    syndromic boundary in the other direction from CRBN: intellectual disability
    is frequently accompanied by epilepsy and autistic features.
  evidence:
  - reference: PMID:27616480
    reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      From a cohort of >5,000 families affected by neurodevelopmental disorders,
      we identified six consanguineous families harboring homozygous
      inactivating variants in MBOAT7, encoding lysophosphatidylinositol
      acyltransferase (LPIAT1).
    explanation: >-
      Establishes the MBOAT7 subtype, its biallelic inactivating mechanism, and
      its rarity within a very large neurodevelopmental cohort.
  - reference: PMID:27616480
    reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Subjects presented with ID frequently accompanied by epilepsy and autistic features."
    explanation: >-
      Documents the accompanying epilepsy and autistic features that qualify the
      non-syndromic designation for this subtype.

pathophysiology:
- name: Biallelic Loss-of-Function Variant in an Autosomal Neurodevelopmental Gene
  biological_scale: MOLECULAR
  role: trigger
  description: >-
    The initiating lesion is the presence of two damaging alleles at one
    autosomal locus - most often a homozygous protein-truncating variant, but
    also homozygous whole-gene or partial deletion (TUSC3), complex
    rearrangement (GRIK2), or homozygous hypomorphic missense (MED23, MAN1B1) -
    so that the gene product is absent or severely reduced. Unlike the dominant
    class the affected genes are NOT restricted to the
    haploinsufficiency-intolerant fraction of the genome: because one working
    copy suffices, a damaging recessive allele is under weak purifying selection
    and persists in the population, so the recessive class draws on a far larger
    gene pool. That is the structural reason the number of candidate loci is
    estimated in the thousands and each individual gene explains only a handful
    of families. Homozygous copy-number lesions are a genuine and recurrent
    subset of the trigger, which is why chromosomal microarray remains part of
    the workup even in a suspected-recessive workflow.
  locations:
  - preferred_term: brain
    term:
      id: UBERON:0000955
      label: brain
  biological_processes:
  - preferred_term: nervous system development
    term:
      id: GO:0007399
      label: nervous system development
  downstream:
  - target: Autozygosity Exposes Recessive Alleles in Consanguineous Pedigrees
    description: >-
      The recessive lesion is silent in the heterozygous state; parental
      relatedness is what converts a carrier allele into an affected homozygote.
    causal_link_type: DIRECT
    hypothesis_groups:
    - thousands_of_recessive_id_loci
    evidence:
    - reference: PMID:30459488
      reference_title: "Genetics of autosomal recessive intellectual disability."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        The risk for ARID in children of first cousins or closer is a magnitude
        higher than for children of unrelated parents.
      explanation: >-
        Supports the edge itself: it is parental relatedness, not the allele
        alone, that converts recessive carrier burden into affected offspring.
  - target: Loss of a Gene-Specific Molecular Function Required by Developing Neurons
    description: >-
      Absence of the gene product removes one component from one of a limited
      number of recurring cellular processes required for brain development.
      Under the long-tail model this single edge stands in for many hundreds to
      thousands of gene-specific instantiations.
    causal_link_type: DIRECT
    hypothesis_groups:
    - thousands_of_recessive_id_loci
    evidence:
    - reference: PMID:17718851
      reference_title: "Genetics of autosomal recessive non-syndromic mental retardation: recent advances."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        All of the mutations in the ARNSMR-causing genes are protein truncating,
        indicating a putative severe loss-of-function effect.
      explanation: >-
        Supports the edge: the allelic class is protein-truncating loss of
        function, so the biallelic lesion removes the gene-specific function.
  evidence:
  - reference: PMID:17718851
    reference_title: "Genetics of autosomal recessive non-syndromic mental retardation: recent advances."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All of the mutations in the ARNSMR-causing genes are protein truncating,
      indicating a putative severe loss-of-function effect.
    explanation: >-
      Establishes complete loss of function as the allelic mechanism for the
      founding members of the class.
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Recent studies have shown that autosomal recessive mental retardation
      (ARMR) is extremely heterogeneous, and there is reason to believe that the
      number of underlying gene defects goes into the thousands.
    explanation: >-
      States the locus-heterogeneity consequence of a recessive architecture not
      confined to constraint-intolerant genes.
  - reference: PMID:28397838
    reference_title: "Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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).
    explanation: >-
      Quantifies the gene pool and states explicitly that the majority is
      recessive.
  - reference: PMID:28397838
    reference_title: "Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      Illustrates the allelic spectrum of the trigger - predominantly
      loss-of-function with a minority of conserved-residue missense.
  - reference: PMID:21937992
    reference_title: "Deep sequencing reveals 50 novel genes for recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      More than 90 different gene defects have been identified for
      X-chromosome-linked intellectual disability alone, but research into the
      more frequent autosomal forms of intellectual disability is still in its
      infancy.
    explanation: >-
      Contextualises the autosomal recessive class as the larger but
      historically under-characterised fraction of inherited intellectual
      disability.

- name: Autozygosity Exposes Recessive Alleles in Consanguineous Pedigrees
  biological_scale: ORGANISM
  role: modifier
  description: >-
    Parental consanguinity does not create the mutant allele; it raises the
    probability that a rare carrier allele is inherited in duplicate, identical
    by descent, so that the two copies lie inside a shared run of homozygosity.
    This is the population-genetic step that converts an invisible recessive
    burden into overt disease, and it is why the class is concentrated in
    populations where cousin marriage is common - the risk of recessive
    intellectual disability in children of first cousins or closer is an order
    of magnitude higher than in children of unrelated parents. It is also the
    methodological lever: because the causal variant must lie inside an
    autozygous interval shared by all affected siblings, homozygosity mapping
    reduces the search space enough for exome sequencing to identify the gene in
    a single family. The same reasoning explains why the class was almost
    invisible before this method - conventional linkage requires many
    independent families per locus, which extreme locus heterogeneity makes
    impossible.
  locations:
  - preferred_term: brain
    term:
      id: UBERON:0000955
      label: brain
  downstream:
  - target: Loss of a Gene-Specific Molecular Function Required by Developing Neurons
    description: >-
      Homozygosity for the damaging allele is what allows the molecular
      consequence to be expressed at all. This is exactly the step the
      imprinted-bias model contests: for a gene with strong parent-of-origin
      expression bias, loss of the dominant (maternal) allele alone may reduce
      expression enough to cross the threshold, so biallelic loss would not be
      required.
    causal_link_type: DIRECT
    hypothesis_groups:
    - thousands_of_recessive_id_loci
    - imprinted_bias_monoallelic_arid
    evidence:
    - reference: PMID:20004765
      reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Using Affymetrix 5.0 single nucleotide polymorphism (SNP) microarrays, we
        identified a 3.2 Mb region on 8q24 with a continuous run of 606 homozygous
        SNPs shared among all affected members of the family.
      explanation: >-
        Worked instance of the edge: the autozygous interval shared by affected
        siblings is what localises and exposes the biallelic functional loss.
  evidence:
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Taken together, it seems that children of related parents are at a 2 to 3
      times higher risk for ID.
    explanation: >-
      Quantifies the overall effect of parental relatedness on intellectual
      disability risk at the population level.
  - reference: PMID:21937992
    reference_title: "Deep sequencing reveals 50 novel genes for recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      To expedite the molecular elucidation of autosomal-recessive intellectual
      disability, we have now performed homozygosity mapping, exon enrichment
      and next-generation sequencing in 136 consanguineous families with
      autosomal-recessive intellectual disability from Iran and elsewhere.
    explanation: >-
      The defining methodology: autozygosity mapping plus exome sequencing in
      consanguineous families, the strategy that opened the class.
  - reference: PMID:20004765
    reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Using Affymetrix 5.0 single nucleotide polymorphism (SNP) microarrays, we
      identified a 3.2 Mb region on 8q24 with a continuous run of 606 homozygous
      SNPs shared among all affected members of the family.
    explanation: >-
      Concrete worked example of an autozygous interval - a run of homozygous
      SNPs shared by affected siblings - delimiting the causal locus.
  - reference: PMID:16033914
    reference_title: "The CC2D1A, a member of a new gene family with C2 domains, is involved in autosomal recessive non-syndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The molecular basis of autosomal recessive non-syndromic mental
      retardation (NSMR) is poorly understood, mostly owing to heterogeneity and
      absence of clinical criteria for grouping families for linkage analysis.
    explanation: >-
      States the pre-autozygosity-mapping obstacle: locus heterogeneity plus
      absence of a clinical gestalt makes multi-family linkage impossible.

- name: Loss of a Gene-Specific Molecular Function Required by Developing Neurons
  biological_scale: MOLECULAR
  description: >-
    The class has no shared pathway. The authoritative review states plainly
    that there are no prevalent ARID genes, pathways, or protein complexes, and
    that the functions of the affected proteins are very diverse. What recurs is
    a limited set of cellular THEMES, each entered by several genes: synaptic
    transmission and proteolysis (GRIK2, PRSS12, MPDZ), activity-dependent and
    general transcriptional regulation (MED23, TET1), RNA and tRNA modification
    (NSUN2, ADAT3, ELP2, METTL23), protein N-glycosylation and glycoprotein
    quality control (TUSC3, MAN1B1, ST3GAL3), vesicle trafficking with NF-kB
    signalling (TRAPPC9, TRAPPC6B, WASHC4, TBC1D23, CC2D1A), regulated protein
    degradation (CRBN), and membrane phospholipid remodelling (MBOAT7). These
    are post-hoc organising themes, not a claim of pathway convergence. What IS
    statistically established at class level is weaker but real: the gene set is
    enriched for co-expression, for direct protein-protein interaction with
    products of already-known intellectual disability genes, and for prenatal
    and infant brain expression.
  locations:
  - preferred_term: brain
    term:
      id: UBERON:0000955
      label: brain
  - preferred_term: cerebral cortex
    term:
      id: UBERON:0000956
      label: cerebral cortex
  cell_types:
  - preferred_term: neuron
    term:
      id: CL:0000540
      label: neuron
  biological_processes:
  - preferred_term: nervous system development
    term:
      id: GO:0007399
      label: nervous system development
  - preferred_term: RNA methylation
    term:
      id: GO:0001510
      label: RNA methylation
  - preferred_term: protein N-linked glycosylation
    term:
      id: GO:0006487
      label: protein N-linked glycosylation
  - preferred_term: vesicle-mediated transport
    term:
      id: GO:0016192
      label: vesicle-mediated transport
  - preferred_term: canonical NF-kappaB signal transduction
    term:
      id: GO:0007249
      label: canonical NF-kappaB signal transduction
  - preferred_term: regulation of DNA-templated transcription
    term:
      id: GO:0006355
      label: regulation of DNA-templated transcription
  - preferred_term: proteolysis
    term:
      id: GO:0006508
      label: proteolysis
  - preferred_term: phosphatidylinositol acyl-chain remodeling
    term:
      id: GO:0036149
      label: phosphatidylinositol acyl-chain remodeling
  genes:
  - preferred_term: NSUN2
    term:
      id: hgnc:25994
      label: NSUN2
  - preferred_term: ADAT3
    term:
      id: hgnc:25151
      label: ADAT3
  - preferred_term: CRBN
    term:
      id: hgnc:30185
      label: CRBN
  downstream:
  - target: Impaired Synaptic Transmission and Plasticity
    description: >-
      Several themes (glutamate receptor function, synaptic proteolysis,
      immediate-early-gene induction, phosphoinositide remodelling) converge
      directly on synaptic function.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:12459588
      reference_title: "Truncating neurotrypsin mutation in autosomal recessive nonsyndromic mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        These findings suggest that neurotrypsin-mediated proteolysis is required
        for normal synaptic function and suggest potential insights into the
        pathophysiological bases of mental retardation.
      explanation: >-
        Supports the edge for the synaptic-proteolysis theme: loss of the
        gene-specific molecular function degrades synaptic function.
  - target: Impaired Cortical Circuit Formation and Connectivity
    description: >-
      Trafficking, glycosylation, and transcriptional themes act during prenatal
      and early postnatal cortical development, affecting neuronal
      differentiation, neurite outgrowth, and axonal connectivity.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:20004763
      reference_title: "A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        TRAPPC9 is highly expressed in the postmitotic neurons of the cerebral
        cortex, and MRI analysis of affected patients shows defects in axonal
        connectivity.
      explanation: >-
        Supports the edge for the trafficking theme: loss of the gene-specific
        function in cortical neurons yields an axonal-connectivity defect.
  evidence:
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      The load-bearing negative claim for this node: the class must be modelled
      as thematic recurrence, not pathway convergence.
  - reference: PMID:21937992
    reference_title: "Deep sequencing reveals 50 novel genes for recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Proteins encoded by several of these genes interact directly with products
      of known intellectual disability genes, and many are involved in
      fundamental cellular processes such as transcription and translation,
      cell-cycle control, energy metabolism and fatty-acid synthesis, which seem
      to be pivotal for normal brain development and function.
    explanation: >-
      Names several of the recurring processes and establishes direct
      protein-protein interaction with known intellectual disability gene
      products.
  - reference: PMID:27457812
    reference_title: "Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes for recessive intellectual disability."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      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.
    explanation: >-
      The strongest class-level statistical evidence for functional coherence:
      spatiotemporally restricted co-expression in human brain.
  - reference: PMID:27457812
    reference_title: "Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes for recessive intellectual disability."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      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).
    explanation: >-
      Quantified protein-interaction enrichment supporting thematic (not
      single-pathway) coherence.
  - reference: PMID:26748517
    reference_title: "Systematic Phenomics Analysis Deconvolutes Genes Mutated in Intellectual Disability into Biologically Coherent Modules."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      Using this integrated resource, we show that ID-AGs are substantially
      enriched with co-expression, protein-protein interactions, and specific
      biological functions.
    explanation: >-
      Systematic evidence that intellectual disability genes as a set are
      functionally coherent rather than randomly distributed.
  - reference: PMID:28397838
    reference_title: "Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Transcriptional studies showed prominent expression in the prenatal brain."
    explanation: >-
      Establishes the shared expression signature - prenatal brain - of the
      recessive gene set, constraining when the themes matter.

- name: Impaired Synaptic Transmission and Plasticity
  biological_scale: CELLULAR
  description: >-
    Where the affected theme is synaptic, the immediate cellular consequence is
    a defect in excitatory neurotransmission or in the activity-dependent
    changes that encode experience. GRIK2 loss removes a kainate-receptor
    subunit from the excitatory synapse with electrophysiologically confirmed
    loss of function; PRSS12/neurotrypsin loss removes a protease acting at the
    presynaptic membrane lining the synaptic cleft, where its proteolysis is
    required for normal synaptic function; MED23 loss blocks JUN/FOS
    immediate-early-gene induction, the transcriptional arm of
    activity-dependent plasticity; MBOAT7 loss alters the arachidonoyl
    phosphatidylinositol pool from which synaptic signalling lipids derive; and
    CC2D1A loss releases PDE4D hyperactivity, depleting hippocampal cAMP and
    reducing CREB signalling. The resulting deficit is in the machinery of
    learning and memory rather than in gross brain structure - which is
    precisely why the phenotype is cognitive and non-syndromic.
  cell_types:
  - preferred_term: neuron
    term:
      id: CL:0000540
      label: neuron
  - preferred_term: glutamatergic neuron
    term:
      id: CL:0000679
      label: glutamatergic neuron
  locations:
  - preferred_term: hippocampal formation
    term:
      id: UBERON:0002421
      label: hippocampal formation
  - preferred_term: cerebral cortex
    term:
      id: UBERON:0000956
      label: cerebral cortex
  cellular_components:
  - preferred_term: synapse
    term:
      id: GO:0045202
      label: synapse
  - preferred_term: presynapse
    term:
      id: GO:0098793
      label: presynapse
  biological_processes:
  - preferred_term: chemical synaptic transmission
    term:
      id: GO:0007268
      label: chemical synaptic transmission
  - preferred_term: regulation of synaptic plasticity
    term:
      id: GO:0048167
      label: regulation of synaptic plasticity
  - preferred_term: modulation of chemical synaptic transmission
    term:
      id: GO:0050804
      label: modulation of chemical synaptic transmission
  genes:
  - preferred_term: GRIK2
    term:
      id: hgnc:4580
      label: GRIK2
  - preferred_term: PRSS12
    term:
      id: hgnc:9477
      label: PRSS12
  - preferred_term: MED23
    term:
      id: hgnc:2372
      label: MED23
  - preferred_term: MBOAT7
    term:
      id: hgnc:15505
      label: MBOAT7
  - preferred_term: CC2D1A
    term:
      id: hgnc:30237
      label: CC2D1A
  downstream:
  - target: Impaired Learning, Memory, and Cognitive Development
    description: >-
      Loss of synaptic plasticity is the proximate cellular substrate of the
      learning and memory deficit.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:17847003
      reference_title: "A defect in the ionotropic glutamate receptor 6 gene (GRIK2) is associated with autosomal recessive mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        This finding provides the first proof that GLU(K6) is indispensable for
        higher brain functions in humans, and future studies of this and other
        ionotropic kainate receptors will shed more light on the pathophysiology
        of mental retardation.
      explanation: >-
        Supports the edge: loss of an ionotropic glutamate receptor subunit at
        the excitatory synapse is sufficient to impair human higher cognition.
  - target: Seizures
    description: >-
      A subtype-restricted rather than class-level consequence: where the
      synaptic lesion involves membrane phospholipid remodelling (MBOAT7/MRT57),
      the altered excitatory signalling substrate also lowers seizure threshold.
      Seizures are explicitly absent in other subtypes (CRBN/MRT2), so this edge
      must not be read as a class-wide consequence.
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    evidence:
    - reference: PMID:27616480
      reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: "Subjects presented with ID frequently accompanied by epilepsy and autistic features."
      explanation: >-
        Documents epilepsy as a frequent accompaniment where the synaptic lesion
        is MBOAT7-mediated phospholipid remodelling.
  - target: Autistic Features
    description: >-
      Subtype-restricted in the same way as the seizure edge, and documented in
      the same MBOAT7/MRT57 cohort; explicitly absent in the CRBN/MRT2 kindred.
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    evidence:
    - reference: PMID:27616480
      reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: "Subjects presented with ID frequently accompanied by epilepsy and autistic features."
      explanation: >-
        Documents autistic features alongside epilepsy in the same
        synaptic-lipid subtype.
  evidence:
  - reference: PMID:12459588
    reference_title: "Truncating neurotrypsin mutation in autosomal recessive nonsyndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      Localises the founding gene product to the presynaptic terminal, anchoring
      the synaptic mechanism in human brain tissue.
  - reference: PMID:12459588
    reference_title: "Truncating neurotrypsin mutation in autosomal recessive nonsyndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      These findings suggest that neurotrypsin-mediated proteolysis is required
      for normal synaptic function and suggest potential insights into the
      pathophysiological bases of mental retardation.
    explanation: >-
      Explicitly links the molecular lesion to synaptic function as the
      pathophysiological basis of the cognitive phenotype.
  - reference: PMID:17847003
    reference_title: "A defect in the ionotropic glutamate receptor 6 gene (GRIK2) is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      This finding provides the first proof that GLU(K6) is indispensable for
      higher brain functions in humans, and future studies of this and other
      ionotropic kainate receptors will shed more light on the pathophysiology
      of mental retardation.
    explanation: >-
      Establishes an ionotropic glutamate receptor as necessary for human
      cognition, the most direct synaptic-transmission link in the class.
  - reference: PMID:21868677
    reference_title: "MED23 mutation links intellectual disability to dysregulation of immediate early gene expression."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      These findings highlight the crucial role of Mediator in brain development
      and functioning and suggest that altered IEG expression might be a common
      molecular hallmark of cognitive deficit.
    explanation: >-
      Connects the transcriptional theme to activity-dependent plasticity via
      immediate-early-gene induction.
  - reference: PMID:30732858
    reference_title: "Male-Specific cAMP Signaling in the Hippocampus Controls Spatial Memory Deficits in a Mouse Model of Autism and Intellectual Disability."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      We found that in Cc2d1a-deficient male mice PDE4D is hyperactive, leading
      to a reduction in cAMP response element binding protein signaling, but
      this molecular deficit is not present in female mice.
    explanation: >-
      Mechanistic detail for the CC2D1A subtype's cAMP arm. PARTIAL and
      MODEL_ORGANISM because the signalling defect is shown in mouse hippocampus
      and is male-specific, so its human generality is unestablished.
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      Moreover, our observations from the Drosophila model point toward an
      evolutionarily conserved role of RNA methylation in normal cognitive
      development.
    explanation: >-
      Supports the plasticity/memory readout for the RNA-modification theme.
      PARTIAL and MODEL_ORGANISM because the functional demonstration is in
      Drosophila, not human neurons.

- name: Impaired Cortical Circuit Formation and Connectivity
  biological_scale: TISSUE
  description: >-
    Where the affected theme acts during development - vesicle trafficking,
    N-glycosylation of nascent membrane and secreted proteins, general
    transcriptional regulation - the consequence is a defect in the BUILDING of
    cortical circuitry rather than in its moment-to-moment operation. The
    recessive gene set is preferentially expressed in prenatal brain, and the
    best-characterised example, TRAPPC9, is highly expressed in postmitotic
    cortical neurons, with affected individuals showing MRI evidence of impaired
    axonal connectivity and mild cerebral white matter hypoplasia. Crucially the
    structural consequence is subtle - reduced connectivity and, in some
    subtypes, mild postnatal microcephaly - rather than a gross malformation
    such as lissencephaly or agenesis of the corpus callosum. This quantitative
    rather than qualitative structural effect is what keeps the class inside the
    non-syndromic category: routine clinical neuroimaging is usually reported as
    normal or near-normal. Note also what this node is NOT: there is no
    degenerative, ischaemic, oxidative, inflammatory, or fibrotic arm. AR-NSID
    is a mis-built circuit, not a damaged one.
  cell_types:
  - preferred_term: neuron
    term:
      id: CL:0000540
      label: neuron
  - preferred_term: pyramidal neuron
    term:
      id: CL:0000598
      label: pyramidal neuron
  locations:
  - preferred_term: cerebral cortex
    term:
      id: UBERON:0000956
      label: cerebral cortex
  - preferred_term: white matter
    term:
      id: UBERON:0002316
      label: white matter
  biological_processes:
  - preferred_term: neuron differentiation
    term:
      id: GO:0030182
      label: neuron differentiation
  - preferred_term: neurogenesis
    term:
      id: GO:0022008
      label: neurogenesis
  - preferred_term: nervous system development
    term:
      id: GO:0007399
      label: nervous system development
  genes:
  - preferred_term: TRAPPC9
    term:
      id: hgnc:30832
      label: TRAPPC9
  - preferred_term: TUSC3
    term:
      id: hgnc:30242
      label: TUSC3
  - preferred_term: MAN1B1
    term:
      id: hgnc:6823
      label: MAN1B1
  downstream:
  - target: Impaired Learning, Memory, and Cognitive Development
    description: >-
      Reduced cortical connectivity limits the substrate available for cognitive
      function.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:27457812
      reference_title: "Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes for recessive intellectual disability."
      supports: SUPPORT
      evidence_source: COMPUTATIONAL
      snippet: >-
        These results suggest that disruptions of temporal parietal and
        sub-cortical neurogenesis during infancy are critical to the
        pathophysiology of ID.
      explanation: >-
        Supports the edge: disrupted cortical/subcortical circuit formation in
        the infant brain is the asserted route to the cognitive phenotype.
  - target: Cerebral White Matter Hypoplasia
    description: >-
      The imaging correlate of the connectivity defect, reported in
      TRAPPC9/MRT13. Quantitative and subtle rather than a recognisable
      malformation, which is why it does not by itself move the case out of the
      non-syndromic category.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:20004765
      reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Brain magnetic resonance imaging of affected individuals indicates the
        presence of mild cerebral white matter hypoplasia.
      explanation: >-
        Directly documents the white-matter imaging consequence of the
        connectivity defect in an AR-NSID subtype.
  - target: Secondary (Postnatal) Microcephaly
    description: >-
      In a subset of subtypes, notably TRAPPC9, reduced postnatal brain growth
      manifests as acquired microcephaly. This edge is the pathograph location
      of the syndromic-boundary problem: it is an accessory, incompletely
      penetrant feature of the kind whose accumulation with deeper phenotyping
      progressively reclassifies MRT entities as syndromic.
    causal_link_type: DIRECT
    hypothesis_groups:
    - nonsyndromic_boundary_is_ascertainment
    evidence:
    - reference: PMID:20004763
      reference_title: "A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        We identified a genetic locus for autosomal-recessive nonsyndromic
        intellectual disability associated with variable postnatal microcephaly
        through homozygosity mapping of a consanguineous Israeli Arab family.
      explanation: >-
        Supports the edge and its variability: postnatal microcephaly
        accompanies the connectivity defect in only part of the cohort.
  evidence:
  - reference: PMID:20004763
    reference_title: "A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      TRAPPC9 is highly expressed in the postmitotic neurons of the cerebral
      cortex, and MRI analysis of affected patients shows defects in axonal
      connectivity.
    explanation: >-
      Direct human evidence that the molecular lesion produces a cortical
      connectivity defect rather than a gross malformation.
  - reference: PMID:20004763
    reference_title: "A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This suggests essential roles of TRAPPC9 in human brain development,
      possibly through its effect on NF-kappaB activation and protein
      trafficking in the postmitotic neurons of the cerebral cortex.
    explanation: >-
      Links the trafficking/NF-kB theme specifically to postmitotic cortical
      neuron development.
  - reference: PMID:20004765
    reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Brain magnetic resonance imaging of affected individuals indicates the
      presence of mild cerebral white matter hypoplasia.
    explanation: >-
      Documents the subtle white-matter imaging correlate, consistent with a
      connectivity rather than malformation phenotype.
  - reference: PMID:27457812
    reference_title: "Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes for recessive intellectual disability."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      These results suggest that disruptions of temporal parietal and
      sub-cortical neurogenesis during infancy are critical to the
      pathophysiology of ID.
    explanation: >-
      Places the pathophysiology in a specific developmental window and set of
      regions, on the basis of human brain co-expression data.
  - reference: PMID:32877400
    reference_title: "Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      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.
    explanation: >-
      Animal-model support that biallelic loss of an AR-NSID gene reduces brain
      size and impairs memory. PARTIAL and MODEL_ORGANISM: mouse brain size and
      social memory are proxies for the human construct.

- name: Secondary (Postnatal) Microcephaly
  biological_scale: ORGANISM
  description: >-
    An incompletely penetrant consequence in a subset of subtypes, most clearly
    TRAPPC9-related disease, in which head circumference is normal at birth and
    falls across centiles postnatally. Its variability is diagnostically
    important: because microcephaly is present in some but not all affected
    individuals even within the same gene, its absence does not exclude the
    diagnosis, and its presence does not by itself reclassify the case as
    syndromic.
  locations:
  - preferred_term: brain
    term:
      id: UBERON:0000955
      label: brain
  genes:
  - preferred_term: TRAPPC9
    term:
      id: hgnc:30832
      label: TRAPPC9
  downstream:
  - target: Secondary Microcephaly
    description: >-
      The mechanism node surfaces as the clinically scored phenotype: head
      circumference normal at birth then falling across centiles. Incompletely
      penetrant even within one gene.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:20004765
      reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: "Microcephaly is present in some but not all affected individuals."
      explanation: >-
        Supports both the edge and its incomplete penetrance within a single
        AR-NSID gene.
  evidence:
  - reference: PMID:20004765
    reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Microcephaly is present in some but not all affected individuals."
    explanation: >-
      Directly establishes the incomplete penetrance of microcephaly within a
      single AR-NSID gene.
  - reference: PMID:20004763
    reference_title: "A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      We identified a genetic locus for autosomal-recessive nonsyndromic
      intellectual disability associated with variable postnatal microcephaly
      through homozygosity mapping of a consanguineous Israeli Arab family.
    explanation: >-
      Independent report of the same variable postnatal microcephaly in a
      separate pedigree.

- name: Impaired Learning, Memory, and Cognitive Development
  biological_scale: ORGANISM
  role: consequence
  description: >-
    The convergent organism-level outcome: significantly impaired intellectual
    functioning and adaptive behaviour with onset in the developmental period.
    Severity spans the range from mild (CRBN, IQ 50-70) to severe (CC2D1A), with
    moderate-to-severe the commonest reported band - though that distribution is
    inflated by ascertainment through multiplex families referred to specialist
    genetics services. Because the upstream lesions damage the machinery of
    synaptic plasticity and cortical connectivity rather than producing a
    specific structural or metabolic lesion, the deficit is global and STATIC
    rather than regressive: the developmental trajectory is shallower, so the
    gap versus peers widens with age, but there is no loss of previously
    acquired skills. Developmental regression therefore argues against this
    diagnosis and toward a degenerative or metabolic alternative. Notably, this
    class is not the low tail of the normal-range polygenic intelligence
    distribution - common variants in NS-ARID genes are not enriched for
    intelligence QTLs.
  locations:
  - preferred_term: brain
    term:
      id: UBERON:0000955
      label: brain
  biological_processes:
  - preferred_term: learning or memory
    term:
      id: GO:0007611
      label: learning or memory
  - preferred_term: cognition
    term:
      id: GO:0050890
      label: cognition
  downstream:
  - target: Intellectual Disability
    description: >-
      The obligate clinical expression of the convergent mechanism: significantly
      impaired intellectual functioning and adaptive behaviour with onset in the
      developmental period.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:15557513
      reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        A gene on chromosome 3p that is associated with mild mental retardation
        in a large kindred is reported. This finding implicates a role for the
        ATP-dependent degradation of proteins in memory and learning.
      explanation: >-
        Supports the edge from the memory/learning mechanism to the scored
        intellectual-disability phenotype in a worked AR-NSID kindred.
  - target: Global Developmental Delay
    description: >-
      The presenting form of the same deficit in infancy and early childhood,
      before an IQ can be formally assigned; global rather than domain-selective.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:15557513
      reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: "Developmental milestones were mildly delayed since early childhood."
      explanation: >-
        Documents the early-childhood developmental presentation of the
        cognitive mechanism.
  - target: Mild Intellectual Disability
    description: >-
      The mild end of the severity range the mechanism produces, documented in
      the CRBN/MRT2 kindred.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:15557513
      reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Previously, a disease locus was mapped for a mild type of nonsyndromic
        mental retardation (IQ between 50 and 70) to a 4.2-MB interval on
        chromosome 3p25-pter in a large kindred.
      explanation: >-
        Anchors the mild severity band with an explicit IQ range.
  - target: Moderate Intellectual Disability
    description: >-
      A documented severity band in the published class literature, exemplified
      by the GRIK2/MRT6 pedigree.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:17847003
      reference_title: "A defect in the ionotropic glutamate receptor 6 gene (GRIK2) is associated with autosomal recessive mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        cosegregates with moderate-to-severe nonsyndromic autosomal recessive
        mental retardation in a large, consanguineous Iranian family
      explanation: >-
        Anchors the moderate-to-severe band in a founding AR-NSID pedigree.
  - target: Severe Intellectual Disability
    description: >-
      The severe end of the range, documented across the nine consanguineous
      CC2D1A/MRT3 families.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:16033914
      reference_title: "The CC2D1A, a member of a new gene family with C2 domains, is involved in autosomal recessive non-syndromic mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        A protein truncating mutation was identified in the gene CC2D1A in nine
        consanguineous families with severe autosomal recessive NSMR.
      explanation: >-
        Anchors the severe band across nine pedigrees sharing one recessive
        lesion.
  - target: Impaired Literacy and Academic Attainment
    description: >-
      The functional expression of the cognitive deficit in adult life, where
      outcome is reported at attainment rather than IQ granularity.
    causal_link_type: DIRECT
    evidence:
    - reference: PMID:15557513
      reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        The highest reading ability achieved was at the first-grade level in a
        32-year-old affected woman as assessed by the Gray Oral Reading Scale.
      explanation: >-
        Quantifies the functional attainment ceiling produced by the mechanism,
        using a standardised instrument.
  evidence:
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This finding implicates a role for the ATP-dependent degradation of
      proteins in memory and learning.
    explanation: >-
      Frames the organism-level outcome of a recessive AR-NSID lesion as a
      memory and learning deficit.
  - reference: PMID:17847003
    reference_title: "A defect in the ionotropic glutamate receptor 6 gene (GRIK2) is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      cosegregates with moderate-to-severe nonsyndromic autosomal recessive
      mental retardation in a large, consanguineous Iranian family
    explanation: >-
      Documents the moderate-to-severe band that is typical for the class.
  - reference: PMID:16033914
    reference_title: "The CC2D1A, a member of a new gene family with C2 domains, is involved in autosomal recessive non-syndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A protein truncating mutation was identified in the gene CC2D1A in nine
      consanguineous families with severe autosomal recessive NSMR.
    explanation: >-
      Documents the severe end of the severity range within the class.
  - reference: PMID:26912939
    reference_title: "Examining non-syndromic autosomal recessive intellectual disability (NS-ARID) genes for an enriched association with intelligence differences."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      Gene-based tests indicated that genes implicated in NS-ARID were not
      significantly enriched for quantitative trait loci (QTL) associated with
      intelligence.
    explanation: >-
      Key negative result: the AR-NSID gene set is genetically distinct from the
      polygenic architecture of normal-range intelligence.
  - reference: PMID:26912939
    reference_title: "Examining non-syndromic autosomal recessive intellectual disability (NS-ARID) genes for an enriched association with intelligence differences."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      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.
    explanation: >-
      States the conclusion in the form directly relevant here: AR-NSID is a
      large-effect monogenic class, not the extreme of a continuum.

mechanistic_hypotheses:
- hypothesis_group_id: thousands_of_recessive_id_loci
  hypothesis_label: >-
    The recessive class comprises thousands of loci, each individually
    ultra-rare, because recessive lesions are not filtered by selection
  status: EMERGING
  description: >-
    Under this model the extreme locus heterogeneity of AR-NSID is not an
    artefact of incomplete discovery but a structural prediction. A heterozygous
    carrier is unaffected and reproductively unimpaired, so a damaging recessive
    allele is under weak purifying selection and persists; consequently
    essentially any gene required for brain development can contribute, not
    merely the haploinsufficiency-intolerant minority that dominates the
    dominant class. The predictions are a long tail of genes each explaining one
    or a few families worldwide, a total locus count in the low thousands, no
    prevalent gene or pathway, and continued near-linear discovery of novel
    genes with each new consanguineous cohort. Successive cohorts have behaved
    exactly as predicted, but the estimate remains an extrapolation from
    clinic-ascertained multiplex families rather than a population measurement,
    which is why the status is EMERGING rather than established.
  evidence:
  - reference: PMID:28397838
    reference_title: "Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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).
    explanation: >-
      States the quantitative form of the hypothesis: >2500 autosomal loci,
      mostly recessive.
  - reference: PMID:28397838
    reference_title: "Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      We identified definite or candidate mutations (or CNVs) in 51% of families
      in 72 different genes, including 26 not previously reported for ARID.
    explanation: >-
      Observed behaviour matching the prediction: even a large modern cohort
      yields many genes, a large fraction of them novel.
  - reference: PMID:29302074
    reference_title: "Genetics of intellectual disability in consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In 219 of these, we found likely causative variants, involving 77 known
      and 77 novel AR-ID (candidate) genes, 21 X-linked genes, as well as 9
      genes previously implicated in diseases other than ID.
    explanation: >-
      Novel genes equal known genes one-for-one in the largest cohort to date -
      the signature of an unsaturated long tail.
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      Directly confirms the "no prevalent locus" corollary of the long-tail
      model.
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Particularly, autosomal-recessive forms of the disorder have a very
      heterogeneous molecular basis, and genes with an increased number of
      disease-causing mutations are not common.
    explanation: >-
      States the corollary that recurrently mutated genes are the exception, not
      the rule.
  - reference: PMID:28097321
    reference_title: "Diagnostic Yield and Novel Candidate Genes by Exome Sequencing in 152 Consanguineous Families With Neurodevelopmental Disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Autosomal recessive inherited neurodevelopmental disorders are highly
      heterogeneous, and many, possibly most, of the disease genes are still
      unknown.
    explanation: >-
      Independent statement that the gene set remains largely undiscovered, as
      the hypothesis predicts.

- hypothesis_group_id: nonsyndromic_boundary_is_ascertainment
  hypothesis_label: >-
    The non-syndromic label reflects depth of phenotyping and cohort size rather
    than a mechanistically distinct disease class
  status: EMERGING
  description: >-
    Under this model AR-NSID is the leading edge of recessive gene discovery
    rather than a separate biological entity: a gene-disease pair is called
    non-syndromic while it rests on one or two consanguineous pedigrees and a
    routine dysmorphology examination, and migrates into a named syndrome once
    enough patients accumulate to reveal a gestalt. The prediction is that
    MRT-series entities will be progressively reclassified, and several already
    have been. Within this entry, MRT5/NSUN2 was reported with accompanying
    facial dysmorphism and the discovering authors themselves raised the
    possibility of a syndromic form; MRT15/MAN1B1 was reported with a mixed
    cohort (three strictly non-syndromic families plus one with additional
    features) and now carries the eponymous MONDO label "Rafiq syndrome"; and
    ADAT3 - initially characterised as cognitive impairment with strabismus -
    became "an important recognizable cause of intellectual disability in
    Arabia" with dysmorphic features once a second cohort was described. The
    competing, not mutually exclusive model is that a genuine subset of genes
    has sufficiently brain-restricted consequence that isolated cognitive
    impairment is the real phenotype; TUSC3 is the strongest case, since
    N-glycosylation defects usually produce florid multisystem disease yet TUSC3
    loss does not. Distinguishing the two requires prospective deep phenotyping
    of MRT-series cohorts, which has not been done.
  evidence:
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In addition, in the last few months, evidence has been growing that many
      ARID genes are pleiotropic and that the resulting phenotypes may have a
      broad spectrum.
    explanation: >-
      The class-level statement of the hypothesis: ARID gene-phenotype
      relationships are broader than the initial non-syndromic reports imply.
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      suggesting that mutations in this gene might even induce a syndromic form
      of ID
    explanation: >-
      The original authors of an MRT-series entity explicitly raise
      reclassification as syndromic - the predicted migration, stated
      prospectively.
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The humans homozygous for NSUN2 mutations showed an overlapping phenotype
      consisting of moderate to severe ID and facial dysmorphism
    explanation: >-
      Documents the dysmorphic features that undercut the non-syndromic
      designation for this subtype.
  - reference: PMID:26842963
    reference_title: "ADAT3-related intellectual disability: Further delineation of the phenotype."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      ADAT3-related intellectual disability is an important recognizable cause
      of intellectual disability in Arabia.
    explanation: >-
      Worked example of completed migration: a recessive tRNA-modification
      intellectual-disability gene becomes a RECOGNISABLE (i.e. syndromic)
      entity once a second cohort is phenotyped.
  - reference: PMID:26842963
    reference_title: "ADAT3-related intellectual disability: Further delineation of the phenotype."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      ADAT3-related intellectual disability has been recently described in 24
      individuals from eight Saudi families who had cognitive impairment and
      strabismus.
    explanation: >-
      Establishes the starting point of that migration - a narrowly described
      cognitive phenotype - against which the later expansion is measured.
  - reference: PMID:21763484
    reference_title: "Mutations in the alpha 1,2-mannosidase gene, MAN1B1, cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      four consanguineous families affected by nonsyndromic autosomal-recessive
      intellectual disability (NS-ARID) and one in which the patients show
      additional clinical features
    explanation: >-
      Shows the boundary being crossed inside a single founding report, with
      syndromic and non-syndromic families for the same gene.
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: REFUTE
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Our data suggest that in contrast to other genetic defects of
      glycosylation, inactivation of TUSC3 causes nonsyndromic MR
    explanation: >-
      Counter-evidence for the competing model: a gene in a pathway whose other
      defects are floridly syndromic nevertheless produces isolated cognitive
      impairment, arguing that non-syndromic can be a real biological outcome
      rather than shallow phenotyping.

- hypothesis_group_id: imprinted_bias_monoallelic_arid
  hypothesis_label: >-
    Parent-of-origin expression bias makes a subset of recessive ID genes
    pathogenic in the monoallelic (maternally inherited) state
  status: EMERGING
  description: >-
    Some autosomal genes are not fully biallelically expressed but show a
    parent-of-origin expression BIAS. For TRAPPC9, brain expression is
    predominantly (~70%) from the maternally inherited allele. In mouse this
    produces a striking asymmetry: heterozygotes lacking the maternal allele
    (~70% loss of expression) phenocopy homozygous nulls, whereas heterozygotes
    lacking the paternal allele are normal. If the same bias holds in human
    brain, a SINGLE maternally inherited loss-of-function allele in an
    imprinted-bias ARID gene could be pathogenic - a genotype that standard
    recessive filtering (which demands two hits) would discard, and that
    standard recurrence-risk counselling (25%) would misstate. This would also
    predict apparent "non-penetrance" or unexplained mild cases in families
    where only one allele is disrupted. The hypothesis is EMERGING: the
    allele-bias and the mouse asymmetry are established, but the human clinical
    correlate rests on a single child with increased ad libitum food intake and
    has not been tested systematically in monoallelic TRAPPC9 carriers.
  applies_to_subtypes:
  - MRT13
  evidence:
  - reference: PMID:32877400
    reference_title: "Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      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.
    explanation: >-
      Establishes the parent-of-origin expression bias on which the hypothesis
      depends.
  - reference: PMID:32877400
    reference_title: "Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      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.
    explanation: >-
      The core prediction demonstrated in vivo: monoallelic maternal loss is
      sufficient for the phenotype.
  - reference: PMID:32877400
    reference_title: "Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A role for Trappc9 in energy balance was further supported by increased ad
      libitum food intake in a child with TRAPPC9 deficiency.
    explanation: >-
      The only human datum bearing on the model's translational validity, and it
      concerns the metabolic rather than the cognitive arm and rests on n=1 -
      hence PARTIAL and hence the EMERGING status.

phenotypes:
- category: Neurologic
  name: Intellectual Disability
  description: >-
    The defining and universal phenotype: significantly impaired intellectual
    functioning and adaptive behaviour with onset in the developmental period.
    Severity is gene- and family-dependent rather than uniform across the class.
    The course is static (non-progressive).
  phenotype_term:
    preferred_term: Intellectual disability
    term:
      id: HP:0001249
      label: Intellectual disability
  frequency: OBLIGATE
  diagnostic: true
  evidence:
  - reference: PMID:24176302
    reference_title: "Genetics of recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      For autosomal recessive ID (ARID) the recurrence risk is high and, in
      populations with frequent parental consanguinity, ARID is the most common
      form of ID.
    explanation: >-
      Intellectual disability is the defining phenotype of the class by
      construction; this review states the class in those terms.
  - reference: PMID:12459588
    reference_title: "Truncating neurotrypsin mutation in autosomal recessive nonsyndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A 4-base pair deletion in the neuronal serine protease neurotrypsin gene
      was associated with autosomal recessive nonsyndromic mental retardation
      (MR).
    explanation: >-
      The founding report of the class, associating a biallelic lesion with
      non-syndromic intellectual disability.

- category: Neurologic
  name: Mild Intellectual Disability
  description: >-
    The mild end of the severity spectrum (IQ approximately 50-70), documented
    most clearly in the CRBN/MRT2 kindred, in which affected adults achieved
    first-grade reading at best. Mild cases are systematically under-ascertained
    in this class because they are less likely to reach clinical genetics, so
    the published severity distribution understates them.
  phenotype_term:
    preferred_term: Mild intellectual disability
    term:
      id: HP:0001256
      label: Mild intellectual disability
  subtype: MRT2
  evidence:
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Previously, a disease locus was mapped for a mild type of nonsyndromic
      mental retardation (IQ between 50 and 70) to a 4.2-MB interval on
      chromosome 3p25-pter in a large kindred.
    explanation: >-
      Documents a mild-severity AR-NSID entity with an explicit IQ range.
  notes: >-
    No frequency band is assigned. Per the repo frequency-evidence guidelines,
    the severity distribution across the class cannot be quantified from a
    literature ascertained through multiplex families referred for
    moderate-to-severe intellectual disability.

- category: Neurologic
  name: Moderate Intellectual Disability
  description: >-
    Together with severe intellectual disability, the commonest reported
    severity band in the published AR-NSID literature, typified by the
    GRIK2/MRT6 and NSUN2/MRT5 pedigrees, which were reported as
    moderate-to-severe.
  phenotype_term:
    preferred_term: Moderate intellectual disability
    term:
      id: HP:0002342
      label: Moderate intellectual disability
  evidence:
  - reference: PMID:17847003
    reference_title: "A defect in the ionotropic glutamate receptor 6 gene (GRIK2) is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      cosegregates with moderate-to-severe nonsyndromic autosomal recessive
      mental retardation in a large, consanguineous Iranian family
    explanation: >-
      Documents the moderate-to-severe band in a founding AR-NSID pedigree.
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The humans homozygous for NSUN2 mutations showed an overlapping phenotype
      consisting of moderate to severe ID and facial dysmorphism
    explanation: >-
      Second independent gene with the same moderate-to-severe band.

- category: Neurologic
  name: Severe Intellectual Disability
  description: >-
    The severe end of the spectrum, documented in the nine consanguineous
    CC2D1A/MRT3 families. Severe cases are over-represented in published AR-NSID
    cohorts because they are the ones referred for molecular study.
  phenotype_term:
    preferred_term: Severe intellectual disability
    term:
      id: HP:0010864
      label: Severe intellectual disability
  subtype: MRT3
  evidence:
  - reference: PMID:16033914
    reference_title: "The CC2D1A, a member of a new gene family with C2 domains, is involved in autosomal recessive non-syndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A protein truncating mutation was identified in the gene CC2D1A in nine
      consanguineous families with severe autosomal recessive NSMR.
    explanation: >-
      Documents severe non-syndromic intellectual disability across nine
      pedigrees sharing one recessive lesion.

- category: Neurologic
  name: Global Developmental Delay
  description: >-
    The presenting phenotype in infancy and early childhood, before an IQ can be
    formally assigned. Delay is global rather than domain-selective, consistent
    with a deficit in general cortical circuit function rather than a focal
    lesion. In the CRBN kindred, developmental milestones were described as
    mildly delayed from early childhood.
  phenotype_term:
    preferred_term: Global developmental delay
    term:
      id: HP:0001263
      label: Global developmental delay
  evidence:
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Developmental milestones were mildly delayed since early childhood."
    explanation: >-
      Documents early-childhood developmental delay as the presenting feature in
      an AR-NSID kindred.
  - reference: PMID:31182824
    reference_title: "Meta-analysis and multidisciplinary consensus statement: exome sequencing is a first-tier clinical diagnostic test for individuals with neurodevelopmental disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      We defined NDD as global developmental delay, intellectual disability,
      and/or autism spectrum disorder.
    explanation: >-
      Confirms that global developmental delay is the standard clinical
      presentation category under which these cases are ascertained. PARTIAL
      because the source is a definition within a diagnostic-yield
      meta-analysis, not an AR-NSID-specific phenotype series.

- category: Neurologic
  name: Impaired Literacy and Academic Attainment
  description: >-
    The functional expression of the cognitive deficit. In the well-documented
    CRBN/MRT2 kindred the highest reading ability achieved by any affected
    individual was first-grade level, and most affected adults could not write
    at all. This level of functional detail is unusual in AR-NSID reports, where
    outcome is more often summarised as an IQ band.
  phenotype_term:
    preferred_term: Functional literacy and academic attainment ceiling
  subtype: MRT2
  evidence:
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The highest reading ability achieved was at the first-grade level in a
      32-year-old affected woman as assessed by the Gray Oral Reading Scale.
    explanation: >-
      Quantifies functional academic attainment in an AR-NSID cohort using a
      standardised instrument.
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In five affected adult individuals, writing skills were limited to signing
      their own names.
    explanation: >-
      Further functional-outcome detail for the same cohort.
  notes: >-
    Deliberately left without an ontology `term:` binding. The obvious
    candidate, HP:0001328 Specific learning disability, is a FALSE match rather
    than merely an imprecise one: its HPO definition ends "The impairment is not
    related to a global deficiency of intelligence", which explicitly excludes
    the case described here, namely attainment limited BY global intellectual
    disability. HPO currently has no class for functional literacy or academic
    attainment in the context of global intellectual disability, so this is a
    needs-term (NTR) candidate; per the term-precision guidance, no term is
    preferred over a misleading one.

- category: Neurologic
  name: Seizures
  description: >-
    An accessory, subtype-dependent phenotype rather than a feature of the class
    as a whole. Epilepsy is markedly over-represented among people with
    intellectual disability generally (approximately ten-fold), and is
    particularly characteristic of the MBOAT7/MRT57 subtype, while explicitly
    absent in the CRBN kindred. Its presence should prompt review of whether the
    case is genuinely non-syndromic.
  phenotype_term:
    preferred_term: Seizure
    term:
      id: HP:0001250
      label: Seizure
  subtype: MRT57
  evidence:
  - reference: PMID:27616480
    reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Subjects presented with ID frequently accompanied by epilepsy and autistic features."
    explanation: >-
      Documents epilepsy as a frequent accompaniment in the MBOAT7 subtype.
  - reference: PMID:27616480
    reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The risk of epilepsy among individuals with intellectual disability (ID)
      is approximately ten times that of the general population.
    explanation: >-
      Provides the base-rate context for the intellectual-disability population
      within which this subtype-restricted phenotype sits.
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: REFUTE
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      None of the affected individuals had peri- or postnatal infections,
      afebrile seizures, toxic exposures, or significant head trauma.
    explanation: >-
      Explicit documented absence of seizures in a different AR-NSID subtype,
      confirming this is not a class-level phenotype.
  notes: >-
    No class-level frequency band is assigned: seizures are subtype-restricted
    (frequent in MBOAT7, explicitly absent in CRBN) and no cohort quantifies
    epilepsy across AR-NSID as a whole.

- category: Neurologic
  name: Autistic Features
  description: >-
    Autistic features occur in a subset of subtypes, notably MBOAT7/MRT57.
    Conversely their explicit absence was recorded in the CRBN/MRT2 kindred,
    illustrating that autism is a subtype-level rather than class-level feature.
  phenotype_term:
    preferred_term: Autism
    term:
      id: HP:0000717
      label: Autism
  subtype: MRT57
  evidence:
  - reference: PMID:27616480
    reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Subjects presented with ID frequently accompanied by epilepsy and autistic features."
    explanation: >-
      Documents autistic features in the MBOAT7 subtype.
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: REFUTE
    evidence_source: HUMAN_CLINICAL
    snippet: "There was good eye contact and no autistic features."
    explanation: >-
      Explicit documented absence of autistic features in a different AR-NSID
      subtype, confirming that this is not a class-level phenotype.

- category: Neurologic
  name: Secondary Microcephaly
  description: >-
    Postnatal-onset (acquired) microcephaly, best documented in TRAPPC9/MRT13,
    where head circumference falls across centiles after a normal birth head
    size. Variable even within a single gene, so neither its presence nor its
    absence is decisive. Explicitly absent on examination in the CRBN kindred.
  phenotype_term:
    preferred_term: Secondary microcephaly
    term:
      id: HP:0005484
      label: Secondary microcephaly
  subtype: MRT13
  evidence:
  - reference: PMID:20004763
    reference_title: "A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      We identified a genetic locus for autosomal-recessive nonsyndromic
      intellectual disability associated with variable postnatal microcephaly
      through homozygosity mapping of a consanguineous Israeli Arab family.
    explanation: >-
      Establishes variable postnatal (secondary) microcephaly in the TRAPPC9
      subtype.
  - reference: PMID:20004765
    reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Microcephaly is present in some but not all affected individuals."
    explanation: >-
      Supports the incomplete penetrance. PARTIAL because the report does not
      distinguish primary from secondary microcephaly in every case.
  notes: >-
    No frequency band is assigned, and the omission is deliberate rather than an
    oversight. PMID:32877400 tallies microcephaly at 95% of reported TRAPPC9
    cases in the same sentence that gives the 52% obesity figure used to justify
    the FREQUENT band on `Obesity`, so a VERY_FREQUENT band here would be
    arithmetically available. It is withheld because that tally counts
    "microcephaly" without separating primary from secondary (postnatal) forms,
    whereas this phenotype is specifically bound to HP:0005484 Secondary
    microcephaly; the sibling PMID:20004765 item is marked PARTIAL for the same
    reason. Per the repo frequency-evidence guidelines, a band is omitted rather
    than asserted on a denominator that does not match the phenotype scored.

- category: Endocrine
  name: Obesity
  description: >-
    Obesity is a well-documented, gene-specific accompaniment of TRAPPC9/MRT13
    disease, reported in 52% of published human cases alongside microcephaly in
    95%. It is the clearest example in this entry of an accessory feature that
    is neither propagation noise nor a class-level phenotype: it is a directly
    reported finding for one subtype, and it is mechanistically anchored, since
    TRAPPC9 is highly expressed in hypothalamic regions controlling energy
    balance and the null mouse shows a marked increase in body weight. Its
    presence is a further reason MRT13 sits at the edge of the non-syndromic
    definition.
  phenotype_term:
    preferred_term: Obesity
    term:
      id: HP:0001513
      label: Obesity
  frequency: FREQUENT
  subtype: MRT13
  evidence:
  - reference: PMID:32877400
    reference_title: "Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In humans, both homozygous and compound heterozygous mutations in TRAPPC9
      (NIBP) associate with developmental delay, microcephaly (95% reported
      cases), and obesity (52% reported cases)
    explanation: >-
      Human case-report tally establishing obesity in TRAPPC9-related disease
      and quantifying it at 52%, which falls in the FREQUENT band (30-79%) and
      therefore supports the frequency qualifier directly rather than by
      inference.
  - reference: PMID:32877400
    reference_title: "Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Loss-of-function mutations in human TRAPPC9 cause a rare neurodevelopmental
      syndrome characterized by microcephaly and obesity.
    explanation: >-
      Second human-anchored statement of the same association, from the
      abstract.
  - reference: PMID:32877400
    reference_title: "Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      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.
    explanation: >-
      Animal-model corroboration of the weight phenotype. Kept as a SEPARATE
      evidence item with evidence_source MODEL_ORGANISM so that the human
      phenotype is not resting on mouse data.
  - reference: PMID:32877400
    reference_title: "Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A role for Trappc9 in energy balance was further supported by increased ad
      libitum food intake in a child with TRAPPC9 deficiency.
    explanation: >-
      Suggests a hyperphagic mechanism for the weight gain. PARTIAL because it
      rests on a single child.

- category: Neurologic
  name: Cerebral White Matter Hypoplasia
  description: >-
    Mild reduction in cerebral white matter volume on MRI, reported in the
    TRAPPC9/MRT13 subtype and consistent with the axonal-connectivity defect
    seen on imaging. This is a subtle quantitative finding rather than a
    recognisable malformation, which is why it does not by itself move the case
    out of the non-syndromic category. Routine neuroimaging is reported as
    normal in other subtypes - brain MRI was normal throughout the CRBN kindred.
  phenotype_term:
    preferred_term: Aplasia/Hypoplasia of the cerebral white matter
    term:
      id: HP:0012429
      label: Aplasia/Hypoplasia of the cerebral white matter
  subtype: MRT13
  evidence:
  - reference: PMID:20004765
    reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Brain magnetic resonance imaging of affected individuals indicates the
      presence of mild cerebral white matter hypoplasia.
    explanation: >-
      Directly documents the white-matter imaging finding in the TRAPPC9
      subtype.
  - reference: PMID:20004763
    reference_title: "A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "MRI analysis of affected patients shows defects in axonal connectivity."
    explanation: >-
      Independent imaging correlate of the same connectivity defect.

- category: Neurologic
  name: Absence of Dysmorphism and Extra-Neural Involvement
  description: >-
    The negative phenotype that DEFINES the class and is the single most
    important discriminator from syndromic recessive intellectual disability.
    Operationally it means that expert dysmorphology and general physical
    examination find no consistent extra-neural gestalt, brain imaging is normal
    or near-normal, and metabolic screening is unremarkable. The CRBN kindred
    provides the fullest published documentation of this negative. Because
    adequacy of assessment scales with cohort size, the designation should be
    recorded as provisional (see `mechanistic_hypotheses`).
  phenotype_term:
    preferred_term: Absence of dysmorphic, malformative, and extra-neural abnormality
  diagnostic: true
  evidence:
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: REFUTE
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Detailed physical and neurologic examinations in all study participants
      did not show microcephaly, organomegaly, phakomata, spasticity, weakness,
      neuropathy, visual deficits, feeding problems, or psychiatric disorders.
    explanation: >-
      Documents the absence of extra-neural and additional neurological
      features. REFUTE is used deliberately: the evidence refutes the presence
      of dysmorphic/extra-neural abnormality, which is the definitional
      requirement for this class.
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: REFUTE
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The phenotype does not include congenital anomalies or dysmorphic
      features.
    explanation: >-
      The explicit non-syndromic statement for an MRT-series entity.
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: REFUTE
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Our data suggest that in contrast to other genetic defects of
      glycosylation, inactivation of TUSC3 causes nonsyndromic MR
    explanation: >-
      A second gene in which the absence of the expected multisystem syndromic
      features is the reported finding.
  notes: >-
    Curated as a phenotype entry because the absence is definitional and
    diagnostically load-bearing, not incidental; the evidence items are REFUTE
    because they document the absence of the finding rather than its presence.
    Deliberately left without an ontology `term:` binding. HP:0000271
    Abnormality of the face was considered and rejected: the constellation
    asserted here also covers absent organomegaly, spasticity, weakness,
    neuropathy, visual deficits, feeding problems, and metabolic derangement,
    so a facial term would anchor only a fraction of the claim while implying
    the rest was scoped to the face. HPO has no single class for "no syndromic
    gestalt on expert examination", so this is a needs-term (NTR) candidate;
    the individual absent features are enumerated in the description and
    evidence snippets instead.

genetic:
- name: PRSS12
  gene_term:
    preferred_term: PRSS12
    term:
      id: hgnc:9477
      label: PRSS12
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT1
  notes: >-
    Encodes neurotrypsin, a neuronal serine protease enriched at presynaptic
    terminals. The first gene identified for this disease class (2002) and the
    anchor of the OMIM phenotypic series.
  evidence:
  - reference: PMID:12459588
    reference_title: "Truncating neurotrypsin mutation in autosomal recessive nonsyndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A 4-base pair deletion in the neuronal serine protease neurotrypsin gene
      was associated with autosomal recessive nonsyndromic mental retardation
      (MR).
    explanation: Primary gene-disease evidence for PRSS12 in AR-NSID.
  - reference: PMID:12459588
    reference_title: "Truncating neurotrypsin mutation in autosomal recessive nonsyndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In situ hybridization experiments on human fetal brains showed that
      neurotrypsin was highly expressed in brain structures involved in learning
      and memory.
    explanation: >-
      Human fetal-brain expression pattern linking the gene to the learning and
      memory substrate.

- name: CRBN
  gene_term:
    preferred_term: CRBN
    term:
      id: hgnc:30185
      label: CRBN
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT2
  notes: >-
    Encodes cereblon, an ATP-dependent Lon protease family member. Homozygous
    R419X in a large North American sectarian kindred; associated with the mild
    severity band and the fullest documented non-syndromic negative examination
    in the class.
  evidence:
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A nonsense mutation causing a premature stop codon in a novel gene
      (cereblon; CRBN) was identified that encodes for an ATP-dependent Lon
      protease.
    explanation: Primary gene-disease evidence for CRBN in AR-NSID.

- name: CC2D1A
  gene_term:
    preferred_term: CC2D1A
    term:
      id: hgnc:30237
      label: CC2D1A
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT3
  notes: >-
    Founder protein-truncating variant on 19p13.12; positive regulator of the
    IKK/NF-kB cascade and a restraint on PDE4D-dependent cAMP hydrolysis.
    Associated with severe non-syndromic intellectual disability.
  evidence:
  - reference: PMID:16033914
    reference_title: "The CC2D1A, a member of a new gene family with C2 domains, is involved in autosomal recessive non-syndromic mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A protein truncating mutation was identified in the gene CC2D1A in nine
      consanguineous families with severe autosomal recessive NSMR.
    explanation: Primary gene-disease evidence for CC2D1A across nine pedigrees.
  - reference: PMID:16033914
    reference_title: "The CC2D1A, a member of a new gene family with C2 domains, is involved in autosomal recessive non-syndromic mental retardation."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      Expression of CC2D1A mRNA was shown in the embryonic ventricular zone and
      developing cortical plate in staged mouse embryos, persisting into
      adulthood, with highest expression in the cerebral cortex and hippocampus.
    explanation: >-
      Developmental expression pattern consistent with the cortical-circuit
      mechanism. MODEL_ORGANISM because the expression series is in mouse
      embryos.

- name: NSUN2
  gene_term:
    preferred_term: NSUN2
    term:
      id: hgnc:25994
      label: NSUN2
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT5
  notes: >-
    tRNA methyltransferase; three independent loss-of-function alleles in
    Iranian and Kurdish consanguineous families at the MRT5 locus. Accompanying
    facial dysmorphism places this subtype on the syndromic boundary.
  evidence:
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      that cause a loss of the tRNA-methyltransferase-encoding NSUN2 main
      transcript in homozygotes
    explanation: Primary gene-disease evidence for NSUN2 in AR-NSID.

- name: GRIK2
  gene_term:
    preferred_term: GRIK2
    term:
      id: hgnc:4580
      label: GRIK2
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT6
  notes: >-
    Kainate-type ionotropic glutamate receptor subunit GLUK6. Complex homozygous
    rearrangement removing the ligand-binding domain, transmembrane domain, and
    pore loop, with electrophysiologically confirmed complete loss of function.
  evidence:
  - reference: PMID:17847003
    reference_title: "A defect in the ionotropic glutamate receptor 6 gene (GRIK2) is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      cosegregates with moderate-to-severe nonsyndromic autosomal recessive
      mental retardation in a large, consanguineous Iranian family
    explanation: Primary gene-disease evidence for GRIK2 in AR-NSID.

- name: TUSC3
  gene_term:
    preferred_term: TUSC3
    term:
      id: hgnc:30242
      label: TUSC3
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT7
  notes: >-
    Subunit of the ER oligosaccharyltransferase complex; homozygous deletion on
    chromosome 8. The first AR-NSID gene with mutations in more than one family.
  evidence:
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Haplotype analyses and copy-number studies led to the identification of a
      homozygous deletion partly removing TUSC3 (N33) in all patients.
    explanation: Primary gene-disease evidence for TUSC3 in AR-NSID.
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      TUSC3 is only the fifth gene implicated in NS-ARMR and the first for which
      mutations have been reported in more than one family.
    explanation: >-
      Documents both the historical position of TUSC3 and its multi-family
      replication.

- name: TRAPPC9
  gene_term:
    preferred_term: TRAPPC9
    term:
      id: hgnc:30832
      label: TRAPPC9
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT13
  frequency: >-
    One of the most frequently reported single-gene causes of AR-NSID, reported
    independently by two groups in 2009 and in many families since.
  notes: >-
    TRAPP-complex vesicle-trafficking subunit (NIBP) that also activates NF-kB
    via IKK-beta, converging with CC2D1A on the same signalling theme. Unique in
    this class for an established parent-of-origin expression bias in brain
    (see `mechanistic_hypotheses`).
  evidence:
  - reference: PMID:20004765
    reference_title: "Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In a second large NS-ARMR/ID family, previously linked to 8q24 in a study
      of Iranian families, we identified a 4 bp deletion within exon 14 of
      TRAPPC9, also segregating with the phenotype and truncating the protein.
    explanation: >-
      Two independent pedigrees with distinct truncating TRAPPC9 alleles within
      a single report.
  - reference: PMID:20004763
    reference_title: "A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      Simultaneous independent discovery by a second group in a different
      population - unusually strong replication for this class.

- name: MAN1B1
  gene_term:
    preferred_term: MAN1B1
    term:
      id: hgnc:6823
      label: MAN1B1
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT15
  frequency: >-
    One of the few genes in this class with an elevated mutation frequency
    across multiple populations rather than a single-family report.
  notes: >-
    Alpha-1,2-mannosidase at 9q34.3; a second glycosylation entry point
    alongside TUSC3. Missense alleles reduce catalytic rate roughly 1,300-fold
    or destabilise the protein.
  evidence:
  - reference: PMID:21763484
    reference_title: "Mutations in the alpha 1,2-mannosidase gene, MAN1B1, cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      MAN1B1 is one of the few NS-ARID genes with an elevated mutation frequency
      in patients with NS-ARID from different populations.
    explanation: Supports the cross-population recurrence claim for MAN1B1.
  - reference: PMID:21763484
    reference_title: "Mutations in the alpha 1,2-mannosidase gene, MAN1B1, cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      Both missense mutations are at amino acid residues that are conserved
      across the animal kingdom, and they either reduce k(cat) by ∼1300-fold or
      disrupt stable protein expression in mammalian cells.
    explanation: >-
      Functional demonstration that the missense alleles are severe
      loss-of-function, satisfying the recessive model.

- name: MED23
  gene_term:
    preferred_term: MED23
    term:
      id: hgnc:2372
      label: MED23
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT18
  notes: >-
    Mediator complex subunit; homozygous hypomorphic missense p.R617Q with a
    selective defect in JUN/FOS immediate-early-gene induction.
  evidence:
  - reference: PMID:21868677
    reference_title: "MED23 mutation links intellectual disability to dysregulation of immediate early gene expression."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Here, we report a missense mutation (p. R617Q) in MED23 that cosegregates
      with nonsyndromic autosomal recessive intellectual disability.
    explanation: Primary gene-disease evidence for MED23 in AR-NSID.

- name: MBOAT7
  gene_term:
    preferred_term: MBOAT7
    term:
      id: hgnc:15505
      label: MBOAT7
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  subtype: MRT57
  notes: >-
    Lysophosphatidylinositol acyltransferase (LPIAT1), which transfers
    arachidonic acid onto lysophosphatidylinositol. Six consanguineous families
    identified from a cohort of more than 5,000.
  evidence:
  - reference: PMID:27616480
    reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      From a cohort of >5,000 families affected by neurodevelopmental disorders,
      we identified six consanguineous families harboring homozygous
      inactivating variants in MBOAT7, encoding lysophosphatidylinositol
      acyltransferase (LPIAT1).
    explanation: >-
      Primary gene-disease evidence for MBOAT7 in AR-NSID, with cohort
      denominator.
  - reference: PMID:27616480
    reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: >-
      LPIAT1 is a membrane-bound phospholipid-remodeling enzyme that transfers
      arachidonic acid (AA) to lysophosphatidylinositol to produce AA-containing
      phosphatidylinositol.
    explanation: >-
      Defines the biochemical function placing this gene in the membrane-lipid
      theme.

- name: ADAT3
  gene_term:
    preferred_term: ADAT3
    term:
      id: hgnc:25151
      label: ADAT3
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  notes: >-
    tRNA-editing deaminase subunit. Listed here as a founder-effect exemplar
    rather than as a modelled MRT subtype: a single recurrent homozygous founder
    allele (p.V128M) accounts for a substantial share of recessive intellectual
    disability in the Arabian peninsula. It is also the clearest worked example
    of a gene migrating out of the non-syndromic category once a second cohort
    was deeply phenotyped, and is cited as such under `mechanistic_hypotheses`.
  evidence:
  - reference: PMID:26842963
    reference_title: "ADAT3-related intellectual disability: Further delineation of the phenotype."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A single homozygous founder mutation (c.382G>A:p.(V128M)) in the ADAT3
      gene, which encodes a protein that functions in tRNA editing, was
      identified in all affected individuals.
    explanation: >-
      Documents the recurrent founder allele - the population-genetic pattern
      that makes a few AR-ID genes locally common despite global rarity.
  - reference: PMID:26842963
    reference_title: "ADAT3-related intellectual disability: Further delineation of the phenotype."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In this report, we present additional 15 individuals from 11 families (10
      Saudis and 1 Emirati) who are homozygous for the same founder mutation.
    explanation: >-
      Replication of the founder allele. PARTIAL with respect to THIS entry
      because the expanded phenotype places ADAT3-related disease at or beyond
      the non-syndromic boundary.

- name: Homozygous copy-number variants
  gene_term:
    preferred_term: TUSC3
    term:
      id: hgnc:30242
      label: TUSC3
  relationship_type: CAUSATIVE
  variant_origin: GERMLINE
  notes: >-
    Not a distinct gene but a distinct LESION CLASS that cuts across genes:
    homozygous intragenic and whole-gene deletions are a recurrent cause in this
    disease class (TUSC3 is the archetype), and copy-number variants accounted
    for 14% of solved cases in a large consanguineous intellectual-disability
    cohort. This is the reason chromosomal microarray retains a place in the
    workup even where recessive point mutation is the expected mechanism.
  evidence:
  - reference: PMID:27431290
    reference_title: "Clinical genomics expands the morbid genome of intellectual disability and offers a high diagnostic yield."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      These included copy number variants in 14% (n=54, 15% are novel), and
      point mutations revealed by multi-gene panel and exome sequencing in the
      remaining 43% (1% were found to have Fragile-X).
    explanation: >-
      Quantifies the copy-number share of solved cases in a highly
      consanguineous intellectual-disability cohort.
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Haplotype analyses and copy-number studies led to the identification of a
      homozygous deletion partly removing TUSC3 (N33) in all patients.
    explanation: >-
      The archetypal homozygous-deletion lesion within this disease class.

diagnosis:
- name: Exome or Genome Sequencing with Autozygosity Mapping
  description: >-
    Exome (increasingly genome) sequencing is the definitive diagnostic modality
    and is recommended as a first-tier test for unexplained neurodevelopmental
    disorders. Because there are no prevalent ARID genes, pathways, or protein
    complexes, targeted gene selection is explicitly not justified - broad
    sequencing rather than a small panel is the correct first step. In
    consanguineous families the yield is substantially raised by combining
    sequencing with homozygosity-by-descent mapping from SNP genotyping:
    filtering candidate variants to those inside runs of homozygosity shared by
    all affected siblings converts an unmanageable variant list into a short
    one. Trio or family-based testing remains valuable, because de novo dominant
    lesions occur in consanguineous families too and a homozygosity-only
    filtering strategy will miss them.
  diagnosis_term:
    preferred_term: molecular genetic testing
    term:
      id: NCIT:C19770
      label: Molecular Analysis
  evidence:
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Thus, in a regular case, there is no reasoning for picking a few genes for
      a first diagnostic step, and a genetic diagnosis of ID in general, and
      ARID specifically, is better made using large panels or exome sequencing.
    explanation: >-
      The explicit recommendation against targeted gene selection in this
      disease class, and for broad sequencing instead.
  - reference: PMID:31182824
    reference_title: "Meta-analysis and multidisciplinary consensus statement: exome sequencing is a first-tier clinical diagnostic test for individuals with neurodevelopmental disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Our review demonstrates that ES consistently outperforms CMA for
      evaluation of unexplained NDDs. We propose a diagnostic algorithm placing
      ES at the beginning of the evaluation of unexplained NDDs.
    explanation: >-
      Consensus recommendation placing exome sequencing first-tier for
      unexplained neurodevelopmental disorders.
  - reference: PMID:31182824
    reference_title: "Meta-analysis and multidisciplinary consensus statement: exome sequencing is a first-tier clinical diagnostic test for individuals with neurodevelopmental disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Yield of ES was 36% overall, 31% for isolated NDD, and 53% for the NDD
      plus associated conditions.
    explanation: >-
      Quantifies yield, including the ISOLATED (non-syndromic) figure of 31%
      that is most relevant here and is notably LOWER than for syndromic cases.
      PARTIAL because the meta-analysis is not restricted to consanguineous or
      recessive cases.
  - reference: PMID:28397838
    reference_title: "Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Here, we combined microarray genotyping, homozygosity-by-descent (HBD)
      mapping, copy number variation (CNV) analysis, and whole exome sequencing
      (WES) to identify disease genes/mutations in 192 multiplex Pakistani and
      Iranian consanguineous families with non-syndromic ID.
    explanation: >-
      Describes the combined HBD-mapping plus exome strategy specific to this
      disease class.
  - reference: PMID:28397838
    reference_title: "Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The high yield of AR mutations for ID indicated that this approach has
      excellent clinical potential and should inform clinical diagnostics,
      including clinical whole exome and genome sequencing, for populations in
      which consanguinity is common.
    explanation: >-
      Explicit clinical-diagnostic recommendation for the consanguineous
      setting.
  - reference: PMID:27431290
    reference_title: "Clinical genomics expands the morbid genome of intellectual disability and offers a high diagnostic yield."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Standard clinical evaluation suggested a diagnosis in 16% of cases
      (54/337) but only 70% of these (38/54) were subsequently confirmed. On the
      other hand, the genomic approach revealed a likely diagnosis in 58%
      (n=196).
    explanation: >-
      Head-to-head demonstration that genomics-first massively outperforms
      standard clinical evaluation in a highly consanguineous cohort - the
      strongest single argument for sequencing-first in this population.
  - reference: PMID:27431290
    reference_title: "Clinical genomics expands the morbid genome of intellectual disability and offers a high diagnostic yield."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      When applied directly on all cases with negative molecular karyotyping,
      the diagnostic yield of exome sequencing was 60% (77/129).
    explanation: >-
      Quantifies exome yield after a negative microarray in the consanguineous
      setting.
  - reference: PMID:28097321
    reference_title: "Diagnostic Yield and Novel Candidate Genes by Exome Sequencing in 152 Consanguineous Families With Neurodevelopmental Disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Of these, a clear clinical genetic diagnosis was made in 56 families
      (36.8%) owing to 57 (likely) pathogenic variants in 50 genes already
      established in neurodevelopmental disorders (46 autosomal recessive, 2
      X-linked, and 2 de novo) or in 7 previously proposed recessive candidates.
    explanation: >-
      Diagnostic yield in consanguineous families, with the recessive share of
      solved cases quantified and de novo dominant cases still present.

- name: Chromosomal Microarray
  description: >-
    Chromosomal microarray remains a standard early test for unexplained
    developmental delay or intellectual disability and is not made redundant by
    sequencing in this class: several AR-NSID lesions are homozygous deletions
    (TUSC3 is the archetype) that a copy-number-naive exome pipeline can miss,
    copy-number variants accounted for 14% of solved cases in a large
    consanguineous cohort, and in consanguineous families the array
    simultaneously supplies the genome-wide homozygosity map used to prioritise
    sequence variants.
  diagnosis_term:
    preferred_term: chromosomal microarray
    term:
      id: NCIT:C19770
      label: Molecular Analysis
  evidence:
  - reference: PMID:20466091
    reference_title: "Consensus statement: chromosomal microarray is a first-tier clinical diagnostic test for individuals with developmental disabilities or congenital anomalies."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Available evidence strongly supports the use of CMA in place of G-banded
      karyotyping as the first-tier cytogenetic diagnostic test for patients
      with DD/ID, ASD, or MCA.
    explanation: >-
      Establishes chromosomal microarray as the first-tier cytogenetic test in
      this clinical population.
  - reference: PMID:27431290
    reference_title: "Clinical genomics expands the morbid genome of intellectual disability and offers a high diagnostic yield."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      These included copy number variants in 14% (n=54, 15% are novel), and
      point mutations revealed by multi-gene panel and exome sequencing in the
      remaining 43% (1% were found to have Fragile-X).
    explanation: >-
      Quantifies the copy-number contribution that justifies retaining
      microarray alongside sequencing.
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Haplotype analyses and copy-number studies led to the identification of a
      homozygous deletion partly removing TUSC3 (N33) in all patients.
    explanation: >-
      Worked example of an AR-NSID lesion that is a homozygous deletion,
      requiring copy-number analysis to detect.

- name: Exclusion of Recognised Syndromic, Metabolic, and Treatable Causes
  description: >-
    A non-syndromic designation is a diagnosis of exclusion and requires expert
    dysmorphology assessment plus exclusion of recognised aetiologies (fragile
    X, chromosomal disorders, congenital infection, prenatal exposures, acquired
    perinatal injury, and inborn errors of metabolism). This is not merely
    nosological housekeeping: exome sequencing in consanguineous
    intellectual-disability cohorts uncovers TREATABLE inborn errors of
    metabolism in a small but clinically decisive minority, so a case
    provisionally labelled non-syndromic must not bypass metabolic evaluation.
    Developmental regression, in particular, argues against this diagnosis and
    toward a degenerative or metabolic alternative.
  diagnosis_term:
    preferred_term: disease screening
    term:
      id: NCIT:C15419
      label: Disease Screening
  evidence:
  - reference: PMID:28097321
    reference_title: "Diagnostic Yield and Novel Candidate Genes by Exome Sequencing in 152 Consanguineous Families With Neurodevelopmental Disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In 5 of these families, potentially treatable disorders were diagnosed
      (mutations in PAH, CBS, MTHFR, CYP27A1, and HIBCH), and in 1 family, 2
      disease-causing homozygous variants in different genes were identified.
    explanation: >-
      Directly documents treatable metabolic disorders surfacing within a
      consanguineous intellectual-disability cohort - the clinical reason this
      exclusion step matters.
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Affected individuals had normal laboratory studies on the following: MRI
      of the brain, high-resolution (ISCN G-banding 550 resolution) cytogenetic
      studies, DNA testing for a CCG repeat expansion in fragile X, fasting
      plasma amino acids, urine amino acids and organic acids, pyruvate,
      lactate, and ammonia.
    explanation: >-
      Specifies the exclusion workup that operationally assigns a family to the
      non-syndromic category.
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All individuals had normal newborn screening examinations for
      hypothyroidism, galactosemia, maple syrup urine disease, phenylketonuria,
      and biotinidase.
    explanation: >-
      Documents the metabolic exclusions performed before assigning a
      non-syndromic recessive label.
  - reference: PMID:27431290
    reference_title: "Clinical genomics expands the morbid genome of intellectual disability and offers a high diagnostic yield."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Standard clinical evaluation suggested a diagnosis in 16% of cases
      (54/337) but only 70% of these (38/54) were subsequently confirmed.
    explanation: >-
      Important caveat on the exclusion step: clinical impression alone is
      unreliable, so exclusion should be evidence-based rather than
      gestalt-based. PARTIAL because it bears on the reliability of the step
      rather than on the step's content.

treatments:
- name: Genetic Counselling and Recurrence-Risk Assessment
  description: >-
    The highest-value intervention in this class, and the point at which it
    differs most sharply from the dominant class. Once biallelic causation is
    confirmed, sibling recurrence risk is 25% and carrier testing can be offered
    to at-risk relatives; in consanguineous kindreds where the allele is
    identical by descent, extended-family cascade testing and preconception or
    prenatal testing become possible. Counselling must be delivered with care
    for the cultural context in which consanguineous marriage occurs; the
    objective is informed reproductive choice. Two technical caveats: where
    multi-locus (oligogenic) causation is present, single-locus recurrence
    figures may be wrong; and for a gene with parent-of-origin expression bias
    the standard 25% figure may not apply.
  treatment_term:
    preferred_term: genetic counseling
    term:
      id: NCIT:C15240
      label: Genetic Counseling
  evidence:
  - reference: PMID:17718851
    reference_title: "Genetics of autosomal recessive non-syndromic mental retardation: recent advances."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The future objective will be the development of diagnostic kits for
      molecular diagnosis in mentally retarded individuals in order to offer
      at-risk families pre-natal diagnosis to detect affected offspring.
    explanation: >-
      States the counselling and prenatal-diagnosis rationale that molecular
      diagnosis in this class enables.
  - reference: PMID:24176302
    reference_title: "Genetics of recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      For autosomal recessive ID (ARID) the recurrence risk is high and, in
      populations with frequent parental consanguinity, ARID is the most common
      form of ID.
    explanation: >-
      Establishes the high recurrence risk that makes counselling the primary
      actionable output of diagnosis.
  - reference: PMID:24176302
    reference_title: "Genetics of recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      Frames diagnosis-enabled prevention as the field's stated objective for
      this disease class.
  - reference: PMID:18452889
    reference_title: "A defect in the TUSC3 gene is associated with autosomal recessive mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All obligate carriers of this family were heterozygous, but none of 192
      unrelated healthy individuals from the same population carried this
      deletion.
    explanation: >-
      Supports the counselling message that heterozygous carriers, including
      both parents, are unaffected.
  - reference: PMID:27457812
    reference_title: "Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes for recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In another eight families segregation of multiple pathogenic variants was
      observed, affecting 19 genes that were either known or are novel
      candidates for ID.
    explanation: >-
      Supports the multi-locus counselling caveat. PARTIAL because the report
      documents multi-locus segregation but does not quantify its effect on
      recurrence risk.

- name: Early Intervention, Rehabilitation, and Special Education
  description: >-
    No disease-modifying therapy exists for any AR-NSID subtype. Management is
    the standard, non-gene-specific developmental package: early intervention
    services, individualised special education, and habilitative therapies,
    directed at maximising adaptive function rather than altering the underlying
    lesion. The functional-outcome data available for this class (for example
    reading limited to first-grade level in the CRBN kindred) underline how much
    is at stake in delivering this well.
  treatment_term:
    preferred_term: early intervention and special education
    term:
      id: NCIT:C15315
      label: Rehabilitation
  therapeutic_modality: BEHAVIORAL
  target_phenotypes:
  - preferred_term: Global developmental delay
    term:
      id: HP:0001263
      label: Global developmental delay
  - preferred_term: Intellectual disability
    term:
      id: HP:0001249
      label: Intellectual disability
  evidence:
  - reference: PMID:26503795
    reference_title: "Genetic studies in intellectual disability and related disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This approach has provided novel insights into the biological pathways
      underlying ID, improved the diagnostic process and offered new targets for
      therapy.
    explanation: >-
      Supports the framing that therapy for intellectual disability remains
      prospective - genetics has provided TARGETS, not treatments. PARTIAL
      because the review does not evaluate educational or rehabilitative
      interventions.
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In five affected adult individuals, writing skills were limited to signing
      their own names.
    explanation: >-
      Documents the adult functional attainment that educational and
      habilitative intervention aims to improve.

- name: Speech and Language Therapy
  description: >-
    Directed at the delayed speech and language development that accompanies
    global developmental delay in this class. Symptomatic, with no
    subtype-specific modification.
  treatment_term:
    preferred_term: speech therapy
    term:
      id: NCIT:C159273
      label: Speech Language Therapy
  therapeutic_modality: BEHAVIORAL
  target_phenotypes:
  - preferred_term: Global developmental delay
    term:
      id: HP:0001263
      label: Global developmental delay
  evidence:
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Developmental milestones were mildly delayed since early childhood."
    explanation: >-
      Establishes the developmental delay that speech and language therapy
      addresses. PARTIAL because the source documents the indication, not the
      efficacy of the intervention.

- name: Antiseizure Pharmacotherapy for Comorbid Epilepsy
  description: >-
    Indicated only for the subset of subtypes in which seizures occur, most
    clearly MBOAT7/MRT57. Choice of agent follows standard epilepsy practice; no
    AR-NSID subtype has an established gene-directed antiseizure strategy, so
    the agent is recorded at drug-class rather than individual-drug granularity.
  treatment_term:
    preferred_term: anticonvulsant therapy
    term:
      id: NCIT:C64172
      label: Anticonvulsant Therapy
    therapeutic_agent:
    - preferred_term: anticonvulsant agent
      term:
        id: NCIT:C264
        label: Anticonvulsant Agent
  therapeutic_modality: SMALL_MOLECULE
  target_phenotypes:
  - preferred_term: Seizure
    term:
      id: HP:0001250
      label: Seizure
  evidence:
  - reference: PMID:27616480
    reference_title: "Mutations in MBOAT7, Encoding Lysophosphatidylinositol Acyltransferase I, Lead to Intellectual Disability Accompanied by Epilepsy and Autistic Features."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Subjects presented with ID frequently accompanied by epilepsy and autistic features."
    explanation: >-
      Establishes the epilepsy indication in this subtype. PARTIAL because the
      report documents the phenotype, not antiseizure treatment response.

- name: PDE4 Inhibition in CC2D1A-Related Disease (preclinical)
  description: >-
    The single mechanism-targeted therapeutic lead in this disease class, and
    PRECLINICAL ONLY - no human data exist and it must not be presented as a
    therapy. In Cc2d1a-deficient mice, loss of the CC2D1A restraint on PDE4D
    causes PDE4D hyperactivity, hippocampal cAMP depletion, and reduced CREB
    signalling; a PDE4 inhibitor restores spatial memory. Two hard caveats: the
    rescue is MALE-SPECIFIC (the underlying molecular deficit is absent in
    female mice, and the inhibitor has no effect in them), and the rescued
    readout is rodent spatial memory rather than the human cognitive construct.
  treatment_term:
    preferred_term: Pharmacotherapy
    term:
      id: NCIT:C15986
      label: Pharmacotherapy
    therapeutic_agent:
    - preferred_term: rolipram
      term:
        id: CHEBI:104872
        label: rolipram
  therapeutic_modality: SMALL_MOLECULE
  target_mechanisms:
  - target: Impaired Synaptic Transmission and Plasticity
    treatment_effect: INHIBITS
    description: >-
      PDE4 inhibition restores hippocampal cAMP/CREB signalling that is depleted
      by PDE4D hyperactivity in CC2D1A deficiency, targeting the synaptic
      plasticity node rather than the upstream genetic lesion.
    evidence:
    - reference: PMID:30732858
      reference_title: "Male-Specific cAMP Signaling in the Hippocampus Controls Spatial Memory Deficits in a Mouse Model of Autism and Intellectual Disability."
      supports: SUPPORT
      evidence_source: MODEL_ORGANISM
      snippet: >-
        Restoring PDE4D activity using an inhibitor rescues cognitive deficits
        in male mice but has no effect on female mice.
      explanation: >-
        Demonstrates target engagement at the plasticity node with cognitive
        rescue. PARTIAL and MODEL_ORGANISM: mouse only, and male-specific.
  evidence:
  - reference: PMID:30732858
    reference_title: "Male-Specific cAMP Signaling in the Hippocampus Controls Spatial Memory Deficits in a Mouse Model of Autism and Intellectual Disability."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      We found that in Cc2d1a-deficient male mice PDE4D is hyperactive, leading
      to a reduction in cAMP response element binding protein signaling, but
      this molecular deficit is not present in female mice.
    explanation: >-
      Establishes the drug target and the sex restriction. PARTIAL and
      MODEL_ORGANISM because the entire result is in mice.
  - reference: PMID:30732858
    reference_title: "Male-Specific cAMP Signaling in the Hippocampus Controls Spatial Memory Deficits in a Mouse Model of Autism and Intellectual Disability."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      Cc2d1a-deficient male mice show a deficit in spatial memory, which is not
      present in Cc2d1a-deficient female mice.
    explanation: >-
      The rescued phenotype is itself sex-restricted in the model, compounding
      the translational uncertainty.
  notes: >-
    Curated deliberately as an explicitly preclinical entry rather than omitted,
    because it is the only druggable-mechanism lead in the class and because the
    male-specific effect is a translational caveat that should be recorded
    rather than lost. See the `arnsid_model_translational_validity` discussion.

- name: Disease-Specific Metabolic Therapy After Reclassification
  description: >-
    Initiate the specific metabolic therapy - dietary restriction, cofactor or
    vitamin supplementation, or substrate-lowering pharmacotherapy as dictated
    by the gene - in any child whose provisional AR-NSID label is overturned by
    sequencing in favour of a treatable inborn error of metabolism. In a large
    consanguineous cohort this applied to 5 of 152 families (PAH, CBS, MTHFR,
    CYP27A1, HIBCH). The therapy is disease-modifying for those children, which
    is what distinguishes this from every other entry in this section: AR-NSID
    proper has no disease-modifying treatment, so the single largest
    therapeutic gain available in this clinical population comes from correctly
    identifying the minority who do not actually have it. The complementary
    diagnostic step is curated under `diagnosis` as "Exclusion of Recognised
    Syndromic, Metabolic, and Treatable Causes"; this entry records the
    therapeutic action that follows.
  treatment_term:
    preferred_term: dietary intervention
    term:
      id: NCIT:C15447
      label: Dietary Intervention
  therapeutic_modality: BEHAVIORAL
  evidence:
  - reference: PMID:28097321
    reference_title: "Diagnostic Yield and Novel Candidate Genes by Exome Sequencing in 152 Consanguineous Families With Neurodevelopmental Disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In 5 of these families, potentially treatable disorders were diagnosed
      (mutations in PAH, CBS, MTHFR, CYP27A1, and HIBCH), and in 1 family, 2
      disease-causing homozygous variants in different genes were identified.
    explanation: >-
      Direct evidence that treatable metabolic disorders are found within
      cohorts ascertained as recessive intellectual disability, and therefore
      that this therapeutic action has a real indication in this population.
  notes: >-
    The treatment term is necessarily generic because the therapy is
    gene-specific and heterogeneous (low-phenylalanine diet for PAH, betaine
    for CBS, chenodeoxycholic acid for CYP27A1, and others); Dietary
    Intervention is the closest single NCIT clinical action covering the
    commonest members. Each specific therapy properly belongs to the individual
    metabolic-disease entry rather than here. This entry is retained in
    `treatments` rather than `diagnosis` because the asserted action is
    therapeutic; the diagnostic counterpart is curated separately under
    `diagnosis`.

differential_diagnoses:
- name: Autosomal Dominant Non-Syndromic Intellectual Disability
  description: >-
    The sibling class and the principal differential in a simplex case. Both
    present as isolated intellectual disability without a syndromic gestalt, and
    the two cannot be distinguished clinically - only by the molecular result
    and the pedigree. Distinguishing them matters enormously for counselling:
    recurrence risk is 25% for recessive versus low (mosaicism-limited) for a de
    novo dominant lesion.
  disease_term:
    preferred_term: autosomal dominant non-syndromic intellectual disability
    term:
      id: MONDO:0015802
      label: autosomal dominant non-syndromic intellectual disability
  distinguishing_features:
  - Dominant cases are typically simplex with a heterozygous de novo variant in a constraint-intolerant gene and no consanguinity.
  - Recessive cases are more often multiplex with affected siblings, homozygous variants inside a run of homozygosity, and frequently consanguineous parents.
  - Only trio or family-based molecular testing resolves the two; de novo dominant lesions also occur within consanguineous families, so a consanguineous pedigree does not settle the question.
  evidence:
  - reference: PMID:29302074
    reference_title: "Genetics of intellectual disability in consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Yet, compared to autosomal dominant de novo mutations, which are the
      predominant cause of ID in Western countries, the identification of AR-ID
      genes has lagged behind.
    explanation: >-
      States the population-dependent split between the recessive and de novo
      dominant classes that drives the differential.
  - reference: PMID:28097321
    reference_title: "Diagnostic Yield and Novel Candidate Genes by Exome Sequencing in 152 Consanguineous Families With Neurodevelopmental Disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      50 genes already established in neurodevelopmental disorders (46 autosomal
      recessive, 2 X-linked, and 2 de novo)
    explanation: >-
      Shows de novo dominant diagnoses arising inside a consanguineous cohort -
      the practical reason a recessive assumption must not be allowed to filter
      them out.

- name: Syndromic Intellectual Disability
  description: >-
    Any recognised syndrome in which intellectual disability is accompanied by a
    characteristic pattern of dysmorphic, malformative, metabolic, or
    neuroimaging features. The boundary is genuinely unstable in this class:
    several MRT-series genes (NSUN2, MAN1B1, ADAT3) have been reclassified as
    syndromic once larger series were phenotyped, and the class review notes
    that many ARID genes are pleiotropic with broad phenotypic spectra. A
    non-syndromic label assigned on a small cohort should be treated as
    provisional.
  distinguishing_features:
  - Presence of a consistent extra-neural gestalt on expert dysmorphology examination.
  - A recognisable neuroimaging malformation rather than subtle volume or connectivity change.
  - A diagnostic metabolic or biochemical abnormality.
  - Absence of all of these on adequate assessment defines the non-syndromic category, but adequacy of assessment scales with cohort size and follow-up duration.
  evidence:
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In addition, in the last few months, evidence has been growing that many
      ARID genes are pleiotropic and that the resulting phenotypes may have a
      broad spectrum.
    explanation: >-
      Class-level statement that the syndromic/non-syndromic boundary is
      unreliable for recessive intellectual-disability genes.
  - reference: PMID:26842963
    reference_title: "ADAT3-related intellectual disability: Further delineation of the phenotype."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      ADAT3-related intellectual disability is an important recognizable cause
      of intellectual disability in Arabia.
    explanation: >-
      Worked example of a recessive intellectual-disability gene becoming a
      recognisable (syndromic) entity after further phenotyping.
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      suggesting that mutations in this gene might even induce a syndromic form
      of ID
    explanation: >-
      Shows the same boundary problem raised prospectively by the discovering
      authors.

- name: Treatable Inborn Errors of Metabolism
  description: >-
    Metabolic disorders such as phenylketonuria, homocystinuria, MTHFR
    deficiency, cerebrotendinous xanthomatosis, and HIBCH deficiency are
    themselves autosomal recessive and are enriched in exactly the same
    consanguineous populations. When mild or late-treated they can present as
    apparently isolated intellectual disability. This is the one part of the
    differential where the distinction changes management immediately.
  distinguishing_features:
  - Diagnostic biochemical abnormality on plasma amino acids, urine organic acids, homocysteine, acylcarnitines, or sterol profiling.
  - Often a history of episodic decompensation or a progressive rather than static course.
  - A specific dietary or pharmacological treatment exists, so the distinction is management-changing.
  - Newborn screening excludes several but is not universally available in the populations where AR-NSID is concentrated.
  evidence:
  - reference: PMID:28097321
    reference_title: "Diagnostic Yield and Novel Candidate Genes by Exome Sequencing in 152 Consanguineous Families With Neurodevelopmental Disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In 5 of these families, potentially treatable disorders were diagnosed
      (mutations in PAH, CBS, MTHFR, CYP27A1, and HIBCH), and in 1 family, 2
      disease-causing homozygous variants in different genes were identified.
    explanation: >-
      Quantifies how often treatable metabolic disease is found in cohorts
      ascertained as recessive intellectual disability.
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All individuals had normal newborn screening examinations for
      hypothyroidism, galactosemia, maple syrup urine disease, phenylketonuria,
      and biotinidase.
    explanation: >-
      Illustrates the metabolic exclusions performed before assigning a
      non-syndromic recessive label.

- name: X-Linked Intellectual Disability
  description: >-
    The historically better-characterised class of inherited non-syndromic
    intellectual disability, with more than 90 genes known well before the
    autosomal recessive class was opened up. In a family with affected males
    only, X-linkage and autosomal recessive inheritance can be
    indistinguishable on the pedigree, and fragile X remains a mandatory
    exclusion.
  distinguishing_features:
  - Male-only affectation with transmission through unaffected carrier females and absence of male-to-male transmission favours X-linkage.
  - Affected siblings of both sexes with consanguineous parents favours autosomal recessive inheritance.
  - Large consanguineous cohort studies still recover X-linked genes alongside autosomal recessive ones, so an assumption of autosomal recessive inheritance should not restrict variant filtering.
  - FMR1 repeat testing is a standard first-tier exclusion regardless of pedigree structure.
  evidence:
  - reference: PMID:21937992
    reference_title: "Deep sequencing reveals 50 novel genes for recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      More than 90 different gene defects have been identified for
      X-chromosome-linked intellectual disability alone, but research into the
      more frequent autosomal forms of intellectual disability is still in its
      infancy.
    explanation: >-
      Establishes the relative maturity of the X-linked class and the size of
      its gene set.
  - reference: PMID:29302074
    reference_title: "Genetics of intellectual disability in consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In 219 of these, we found likely causative variants, involving 77 known
      and 77 novel AR-ID (candidate) genes, 21 X-linked genes, as well as 9
      genes previously implicated in diseases other than ID.
    explanation: >-
      Shows X-linked genes being recovered within a consanguineous cohort
      ascertained for recessive disease - the practical reason not to pre-filter
      on inheritance mode.

- name: Progressive or Degenerative Neurological Disease
  description: >-
    Neurodegenerative and neurometabolic disorders can present in early
    childhood with apparent developmental delay before the progressive nature is
    evident. Because AR-NSID is a static developmental encephalopathy - a
    mis-built rather than a damaged circuit, with no degenerative, ischaemic, or
    inflammatory arm - loss of previously acquired skills is the single
    cleanest discriminator against this diagnosis.
  distinguishing_features:
  - Developmental regression (loss of acquired skills) effectively excludes AR-NSID.
  - Progressive neurological signs, evolving neuroimaging abnormality, or a deteriorating clinical course argue for a degenerative alternative.
  - In AR-NSID the gap versus peers widens with age because the developmental trajectory is shallower, not because skills are lost.
  evidence:
  - reference: PMID:15557513
    reference_title: "A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Detailed physical and neurologic examinations in all study participants
      did not show microcephaly, organomegaly, phakomata, spasticity, weakness,
      neuropathy, visual deficits, feeding problems, or psychiatric disorders.
    explanation: >-
      Documents the absence of the progressive neurological signs that would
      point to a degenerative alternative, in the best-characterised AR-NSID
      kindred.

discussions:
- discussion_id: arnsid_no_class_specific_prevalence
  prompt: >-
    What fraction of all intellectual disability is attributable to autosomal
    recessive NON-SYNDROMIC gene defects, in consanguineous and in outbred
    populations?
  kind: KNOWLEDGE_GAP
  status: OPEN
  attaches_to:
  - prevalence#
  rationale: >-
    Every published figure is either the all-cause intellectual disability
    prevalence (about 1%), a whole-class share (about 10% of affected children
    in an outbred population, syndromic and non-syndromic combined), or a solve
    rate within a clinic-ascertained multiplex consanguineous cohort (37-58%).
    None is a population estimate for the non-syndromic recessive class. Solve
    rates are inflated by multiplex ascertainment - families were selected for
    having two or more affected offspring, which enriches for recessive
    inheritance by construction - and are further confounded by cohorts that mix
    syndromic and non-syndromic cases; in one large consanguineous series 92% of
    families had additional features beyond intellectual disability. Without a
    population-based denominator the burden of this class, and therefore the
    case for carrier-screening programmes, cannot be quantified.
  proposed_experiments:
  - experiment_id: exp_arnsid_population_based_sequencing
    name: Unselected population-based sequencing of intellectual disability in a high-consanguinity setting
    description: >-
      Ascertain intellectual disability cases from a defined population rather
      than from genetics clinics, apply exome or genome sequencing to all, and
      have a clinical geneticist classify each case as syndromic or
      non-syndromic blinded to the genotype. This yields the first genuine
      population numerator and denominator for the non-syndromic recessive
      class, and simultaneously measures the size of the multiplex
      ascertainment inflation affecting existing estimates.
  - experiment_id: exp_arnsid_multiplex_bias_reanalysis
    name: Reanalysis of existing consanguineous cohorts stratified by sibship structure
    description: >-
      Re-analyse published large consanguineous cohorts stratifying solve rate
      by number of affected siblings and by syndromic status, to estimate and
      correct the ascertainment inflation in the headline yields and derive a
      bias-adjusted class share.
  evidence:
  - reference: PMID:28097321
    reference_title: "Diagnostic Yield and Novel Candidate Genes by Exome Sequencing in 152 Consanguineous Families With Neurodevelopmental Disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      additional features were present in 140 of the families (92.1%)
    explanation: >-
      Documents the syndromic-contamination confounder directly: 92% of families
      in this consanguineous cohort had features beyond intellectual disability,
      i.e. were not strictly non-syndromic, so the headline yield cannot be read
      as an AR-NSID-specific estimate.
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The prevalence of ARID in a cohort of affected children of an outbred
      population is estimated to be about 10%, with an upward tendency in still
      unclarified cases.
    explanation: >-
      The best available figure, and it illustrates the gap: it is a share of
      affected children rather than a population rate, and it covers ARID as a
      whole rather than the non-syndromic subset.

- discussion_id: arnsid_candidate_gene_validation_backlog
  prompt: >-
    How many of the hundreds of AR-ID "candidate" genes reported from
    consanguineous cohorts are genuinely causal, given that most rest on a
    single family with no functional follow-up?
  kind: KNOWLEDGE_GAP
  status: OPEN
  attaches_to:
  - pathophysiology#Biallelic Loss-of-Function Variant in an Autosomal Neurodevelopmental Gene
  rationale: >-
    The discovery engine for this class - one consanguineous family, one
    autozygous interval, one homozygous variant - produces gene-disease
    assertions faster than they can be replicated or functionally validated. In
    a single-family autozygous interval many rare homozygous variants
    co-segregate perfectly with disease by construction, so segregation carries
    much less evidential weight than in an outbred pedigree. The founding papers
    describe their own findings as "probably disease-causing variants" in
    "candidate" genes, and the largest cohort reports novel candidates at parity
    with confirmed genes. The consequence is that gene-disease validity for much
    of this class is unassessed, which propagates directly into clinical variant
    interpretation.
  proposed_experiments:
  - experiment_id: exp_arnsid_clingen_validity_curation
    name: Systematic gene-disease validity curation of the MRT series
    description: >-
      Apply ClinGen-style gene-disease validity classification to every MRT
      series entry and to candidate genes reported from consanguineous cohorts,
      producing an auditable Definitive/Moderate/Limited/Disputed assignment per
      gene and a quantified estimate of how much of the reported gene set is
      currently below reportable evidence thresholds.
  - experiment_id: exp_arnsid_matchmaking_second_families
    name: Cross-cohort matchmaking for independent second families
    description: >-
      Systematically submit single-family AR-NSID candidate genes to
      matchmaking platforms to seek independent second families with concordant
      genotype and phenotype before clinical reporting, and measure the
      proportion of candidates that do and do not replicate.
  evidence:
  - reference: PMID:21937992
    reference_title: "Deep sequencing reveals 50 novel genes for recessive cognitive disorders."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This study, the largest published so far, has revealed additional
      mutations in 23 genes previously implicated in intellectual disability or
      related neurological disorders, as well as single, probably
      disease-causing variants in 50 novel candidate genes.
    explanation: >-
      The discovering authors themselves label the findings "single, probably
      disease-causing variants" in "candidate" genes.
  - reference: PMID:29302074
    reference_title: "Genetics of intellectual disability in consanguineous families."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      In 219 of these, we found likely causative variants, involving 77 known
      and 77 novel AR-ID (candidate) genes, 21 X-linked genes, as well as 9
      genes previously implicated in diseases other than ID.
    explanation: >-
      Quantifies the backlog: novel candidates match confirmed genes
      one-for-one in the largest cohort published.
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      For an exhaustive deciphering of the genetics of ARID, we suggest research
      at the level of single genes rather than large meta-analyses.
    explanation: >-
      The field's own recommendation, which is effectively a call for the
      per-gene validation work this gap describes.

- discussion_id: arnsid_theme_to_phenotype_mapping
  prompt: >-
    Why do lesions in themes as different as glutamate receptor signalling,
    tRNA methylation, N-glycosylation, and phospholipid remodelling all converge
    on the same clinical phenotype, and does the affected theme predict severity
    or accessory features?
  kind: KNOWLEDGE_GAP
  status: OPEN
  attaches_to:
  - pathophysiology#Loss of a Gene-Specific Molecular Function Required by Developing Neurons
  rationale: >-
    Thematic recurrence is documented statistically - the gene set is enriched
    for co-expression and protein-protein interaction - but it is explicitly NOT
    pathway convergence, since the class review finds no prevalent pathways or
    protein complexes. It is therefore unresolved whether the themes act through
    a shared final common pathway (all ultimately degrading synaptic plasticity)
    or are genuinely parallel routes to a phenotype coarse enough to absorb
    them. The question is testable because the class already shows
    theme-correlated variation: the synaptic member GRIK2 and the
    transcriptional member MED23 produce clean cognitive phenotypes, whereas the
    lipid member MBOAT7 carries epilepsy and autistic features and the
    trafficking member TRAPPC9 carries microcephaly, white-matter change, and
    obesity. Whether that pattern is real or reflects reporting noise across
    tiny cohorts is unknown.
  proposed_experiments:
  - experiment_id: exp_arnsid_multigene_deep_phenotyping
    name: Multi-gene AR-NSID deep-phenotyping cohort
    description: >-
      Assemble a cohort with sufficient numbers per gene across several
      mechanistic themes and apply uniform deep phenotyping (standardised
      cognitive and adaptive testing, EEG, longitudinal head circumference,
      dysmorphology, MRI volumetry), then test whether theme membership predicts
      severity band, epilepsy risk, microcephaly, and syndromic reclassification.
  - experiment_id: exp_arnsid_ipsc_theme_convergence
    name: Comparative profiling of patient-derived neurons across mechanistic themes
    description: >-
      Generate patient iPSC-derived cortical neurons or organoids for
      representatives of each theme and compare transcriptomic, morphometric,
      and electrophysiological readouts, to test whether the themes converge on
      a shared downstream cellular signature or remain distinguishable.
  evidence:
  - reference: PMID:30459488
    reference_title: "Genetics of autosomal recessive intellectual disability."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      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.
    explanation: >-
      States the negative finding that makes the convergence question genuinely
      open rather than already answered by a shared pathway.
  - reference: PMID:26748517
    reference_title: "Systematic Phenomics Analysis Deconvolutes Genes Mutated in Intellectual Disability into Biologically Coherent Modules."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: >-
      Systematic identification of highly enriched functional themes and
      phenotypes revealed typical phenotype combinations characterizing
      process-defined groups of ID disorders, such as chromatin-related
      disorders and deficiencies in DNA repair.
    explanation: >-
      Demonstrates that theme membership CAN predict phenotype combinations for
      some processes, making the question tractable for the recessive themes
      catalogued here.

- discussion_id: arnsid_model_translational_validity
  prompt: >-
    Do the rodent models used for AR-NSID genes report on human intellectual
    disability, given that the rescued readouts are sex-restricted rodent memory
    assays and that a key dosage mechanism (parent-of-origin bias) is
    established only in mouse brain?
  kind: HUMAN_MODEL_MISMATCH
  status: OPEN
  attaches_to:
  - pathophysiology#Impaired Synaptic Transmission and Plasticity
  - pathophysiology#Impaired Cortical Circuit Formation and Connectivity
  rationale: >-
    Evidence exists in the models; what is open is whether it transfers. Three
    specific mismatches are documented rather than hypothetical. First,
    intellectual disability is a construct (IQ plus adaptive functioning) with
    no animal equivalent, so every model readout is a proxy - Drosophila
    short-term memory for NSUN2, mouse spatial and social memory for CC2D1A and
    TRAPPC9. Second, the one pharmacological rescue in the class is
    MALE-SPECIFIC: the PDE4D signalling deficit is absent in Cc2d1a-deficient
    female mice and the inhibitor has no effect in them, so a therapy developed
    on this model could fail in half of patients for reasons the model cannot
    reveal. Third, the parent-of-origin dosage mechanism for Trappc9 is
    established in mouse brain, and its human correlate rests on a single child
    with increased food intake - a metabolic rather than cognitive readout at
    n=1. This is a HUMAN_MODEL_MISMATCH rather than a KNOWLEDGE_GAP because in
    each case the model data are solid and it is their translational validity
    that is uncertain.
  proposed_experiments:
  - experiment_id: exp_arnsid_sex_stratified_human_correlate
    name: Sex-stratified analysis of CC2D1A-related disease in humans
    description: >-
      Assemble the published and unpublished CC2D1A cohort and test whether
      cognitive severity, autistic features, and behavioural phenotype differ by
      sex as the mouse model predicts, before any PDE4-inhibitor trial is
      contemplated. A negative result would indicate the male-specific mouse
      mechanism does not transfer.
  - experiment_id: exp_arnsid_human_brain_trappc9_allelic_bias
    name: Allele-specific TRAPPC9 expression in human brain and phenotyping of monoallelic carriers
    description: >-
      Measure allele-specific TRAPPC9 expression in human brain tissue to test
      whether the ~70% maternal bias seen in mouse holds in humans, and
      systematically phenotype heterozygous carriers of maternally inherited
      TRAPPC9 loss-of-function alleles for subclinical cognitive or metabolic
      effects.
  - experiment_id: exp_arnsid_human_ipsc_cortical_readouts
    name: Human iPSC cortical models as a bridge readout
    description: >-
      Use patient-derived iPSC cortical neurons and organoids, which retain
      human-specific cortical progenitor biology absent from rodent models, to
      test whether the cellular deficits identified in mouse (neurite outgrowth,
      cAMP/CREB signalling, progenitor depletion) are reproduced in human cells
      of the correct genotype.
  evidence:
  - reference: PMID:30732858
    reference_title: "Male-Specific cAMP Signaling in the Hippocampus Controls Spatial Memory Deficits in a Mouse Model of Autism and Intellectual Disability."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      Restoring PDE4D activity using an inhibitor rescues cognitive deficits in
      male mice but has no effect on female mice.
    explanation: >-
      The sex-restricted rescue that makes translational validity the open
      question for the class's only druggable lead.
  - reference: PMID:30732858
    reference_title: "Male-Specific cAMP Signaling in the Hippocampus Controls Spatial Memory Deficits in a Mouse Model of Autism and Intellectual Disability."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      We propose that male-specific signaling mechanisms are involved in
      establishing sex bias in neurodevelopmental disorders.
    explanation: >-
      The authors themselves frame the finding as a general claim about
      sex-specific mechanism, which is precisely what needs human confirmation.
  - reference: PMID:32877400
    reference_title: "Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      A role for Trappc9 in energy balance was further supported by increased ad
      libitum food intake in a child with TRAPPC9 deficiency.
    explanation: >-
      The single human datum supporting the mouse imprinting model, and it
      concerns metabolism rather than cognition at n=1 - the mismatch this gap
      describes.
  - reference: PMID:22541559
    reference_title: "Mutations in NSUN2 cause autosomal-recessive intellectual disability."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: >-
      When the Drosophila melanogaster NSUN2 ortholog was deleted, severe
      short-term-memory (STM) deficits were observed; STM could be rescued by
      re-expression of the wild-type protein in the nervous system.
    explanation: >-
      Illustrates the proxy-readout problem at its most extreme: Drosophila
      short-term memory standing in for human intellectual disability.

notes: >-
  Scope and modelling decisions. (1) This entry models the CLASS, following the
  same pattern as its sibling Autosomal Dominant Non-Syndromic Intellectual
  Disability. The MRT-series subtypes under `has_subtypes` are exemplars chosen
  to span the mechanistic theme set and the severity range, each anchored on a
  primary discovery paper with a verified snippet; they are not an exhaustive
  enumeration of the series, which exceeds 80 numbered entries and is still
  growing. The deep-research report suggested this class might alternatively be
  modelled as a `kb/groupings/` Grouping (a union defined by phenotype plus
  inheritance mode plus a negation); that is a reasonable follow-up but is not
  done here, both because the sibling AD class is already modelled as a Disease
  and because a Grouping requires its members to exist as separate Disease
  entries, which the MRT series does not yet. (2) The pathophysiology graph is
  deliberately kept at the ARCHITECTURE level with worked exemplars, rather than
  attempting per-gene chains for tens of genes. Critically, it does NOT assert
  pathway convergence: the class review states there are no prevalent ARID
  genes, pathways, or protein complexes, so the middle node is curated as
  thematic recurrence with that negative finding cited on the node itself. (3)
  No GeneReviews chapter exists for the class as a whole (PubMed searches for
  "autosomal recessive intellectual disability GeneReviews[TI]" and
  "nonsyndromic intellectual disability GeneReviews[TI]" both return zero
  results), so the GeneReviews phenotype baseline step could not be applied; the
  phenotype set is built instead from the primary MRT discovery literature and
  the large consanguineous cohort studies. (4) The HPO annotations propagated to
  MONDO:0019502 from its descendants (spasticity, cerebral visual impairment,
  and an ear-morphology cluster) were deliberately NOT transcribed: they are
  inherited from syndromic descendants of the grouping node and are precisely
  the features whose presence disqualifies a case from the non-syndromic label.
  (5) ADAT3 appears under `genetic` but deliberately NOT under `has_subtypes`:
  it is included as a founder-effect exemplar and as the worked case of
  migration out of the non-syndromic category, and modelling it as an MRT
  subtype here would assert a non-syndromic status that its own follow-up
  literature contradicts. (6) TECR (MRT14) and ST3GAL3 (MRT12) are genuine
  members of the series per MONDO but are omitted because no primary report
  could be retrieved and cached, and this repository does not permit uncited
  assertions. (7) `prevalence` deliberately carries no class-specific numeric
  rate; see the `arnsid_no_class_specific_prevalence` discussion. (8) MONDO
  asserts `is_a MONDO:0017706` (disorder of carbohydrate transmembrane transport
  and absorption) for MONDO:0019502. That parentage appears to be an upstream
  ontology defect, presumably leaked from a glycosylation-related descendant
  (TUSC3, MAN1B1, ST3GAL3); it is recorded here rather than reflected in
  `parents`, and is worth filing upstream with MONDO. (9) Deep research for this
  entry used the `claude_code` provider rather than `falcon`: no Edison /
  FutureHouse API key was available in the curation environment and
  `deep-research-client` reported `claude_code` as the only available provider.
  All DR-sourced PMIDs were independently fetched with `just fetch-reference`
  and every snippet verified against the cached abstract. (10) Two phenotypes
  deliberately carry a free-text `preferred_term` with NO ontology `term:`
  binding, each with its reasoning in the phenotype's own `notes`: "Impaired
  Literacy and Academic Attainment" (HP:0001328 Specific learning disability is
  a false match, since its definition explicitly excludes impairment related to
  a global deficiency of intelligence) and "Absence of Dysmorphism and
  Extra-Neural Involvement" (HP:0000271 Abnormality of the face would anchor
  only the facial fraction of a much wider negative constellation). Both are
  needs-term (NTR) candidates for HPO. (11) Pathograph wiring: the mechanism
  chain is connected to the phenotype list and to the causal gene list, so the
  entry states in machine-readable form which mechanism produces which clinical
  feature and which gene acts at which node. Genes are attached to the node
  their own discovery literature implicates (synaptic node: GRIK2, PRSS12,
  MED23, MBOAT7, CC2D1A; cortical-circuit node: TRAPPC9, TUSC3, MAN1B1;
  molecular-function node: NSUN2, ADAT3, CRBN), reproducing assignments already
  asserted in the node descriptions rather than adding new claims. Every causal
  edge carries its own evidence item. TWO phenotypes are deliberately left
  causally unconnected: "Obesity", because its route runs through hypothalamic
  energy balance rather than the neurodevelopmental-cognition architecture this
  graph models, and modelling it would require a hypothalamic node the class-level
  graph intentionally does not carry; and "Absence of Dysmorphism and Extra-Neural
  Involvement", because it is a definitional negative and a causal edge into an
  absence would be a category error. Both omissions are principled, not gaps to
  be closed by a later connectivity sweep.

📚

References & Deep Research

References

8

PMID:30459488

Genetics of autosomal recessive intellectual disability.

No top-level findings curated for this source.

PMID:24176302

Genetics of recessive cognitive disorders.

No top-level findings curated for this source.

PMID:21937992

Deep sequencing reveals 50 novel genes for recessive cognitive disorders.

No top-level findings curated for this source.

PMID:29302074

Genetics of intellectual disability in consanguineous families.

No top-level findings curated for this source.

PMID:28397838

Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families.

No top-level findings curated for this source.

PMID:27457812

Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes for recessive intellectual disability.

No top-level findings curated for this source.

PMID:17718851

Genetics of autosomal recessive non-syndromic mental retardation: recent advances.

No top-level findings curated for this source.

PMID:26503795

Genetic studies in intellectual disability and related disorders.

No top-level findings curated for this source.

Deep Research

1

Claude Code ▸

Comprehensive Research Report

claude-haiku-4-5-20251001, claude-opus-5[1m] 20 citations 2026-08-01T01:49:10.868008

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:

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)
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)

Resource	Identifier	Notes
MONDO	`MONDO:0019502`	Verified locally via OAK. Def: "Autosomal recessive form of non-syndromic intellectual disability." `is_a` MONDO:0000509 (non-syndromic ID) and MONDO:0100597 (intellectual disability, autosomal recessive); logical definition `MONDO:0000509 and RO:0000053 some HP:0000007`
OMIM Phenotypic Series	`OMIMPS:249500` (PS249500)	Anchor entry MIM 249500 = MRT1 (PRSS12)
Orphanet	`ORPHA:88616`	ORDO subset `ordo_subtype_of_a_disorder`
DOID	`DOID:0060308`
MedGen / UMLS	MedGen UID 1826073 / `UMLS:C5680181`	Verified via MedGen
GARD	`GARD:0018643`
ICD-10	F70–F79 (no AR-NSID-specific code)	Etiology is not coded in ICD-10; use F70 mild / F71 moderate / F72 severe / F73 profound
ICD-11	6A00 Disorders of intellectual development (6A00.0–6A00.4)	Plus an etiology code from LD2F (monogenic) where applicable
MeSH	D008607 Intellectual Disability; D000015 Abnormalities, Multiple; qualifier "Genetic Diseases, Inborn" (D030342)	No AR-NSID-specific MeSH descriptor
HPO inheritance	`HP:0000007` Autosomal recessive inheritance (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]:

Gene	Families	% of 1,131	Note
VPS13B	10	2.3%	Cohen syndrome — syndromic on full ascertainment
MAN1B1	9	2.0%	`hgnc:6823`
ADAT3	8	~0.7%	`hgnc:25151`
AP4M1	8	~0.7%	AP-4 deficiency — spastic paraplegia, syndromic

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

Phenotype	HPO term (verified via OAK)	Frequency	Notes
Intellectual disability	`HP:0001249` Intellectual disability	Obligate (100%)	Definitional
Global developmental delay	`HP:0001263` Global developmental delay	Very frequent	The pre-school presentation
Delayed speech and language development	`HP:0000750` Delayed speech and language development	Frequent	Often the presenting complaint
Motor delay	`HP:0001270` Delayed gross motor development	Variable	Often spared in classic NS forms

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].

Severity	HPO term (verified)	Comment
Mild ID	`HP:0001256`	Under-ascertained in the AR‑NSID literature
Moderate ID	`HP:0002342`	Common
Severe ID	`HP:0010864`	Common; enriched in the classic multiplex-family series
Profound ID	`HP:0002187`	Present in the Monarch descendant annotations

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:

Feature	HPO (verified)	Association
Microcephaly (usually postnatal)	`HP:0000252`	TRAPPC9 (PMID:20004763 — "associated with variable postnatal microcephaly"; PMID:20004765 — "Microcephaly is present in some but not all affected individuals") [both verbatim]
Seizure	`HP:0001250`	Subset of MRT genes
Autism / autistic behaviour	`HP:0000717`	CC2D1A strongly (see §6.2)
ADHD / short attention span	`HP:0007018` / `HP:0000736`	Frequent behavioural comorbidity
Absent speech	`HP:0001344`	Severe forms
Obesity	`HP:0001513`	TRAPPC9 — >50% of cases (see §4.5, §6.3)
Facial dysmorphism	`HP:0001999`	NSUN2 — "Affected individuals displayed moderate to severe ID and facial dysmorphism" (PMID:22541559) [fragment] — i.e. NSUN2 is arguably mis-classified as non-syndromic

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

Dimension	Value	dismech slot
Onset	Congenital lesion; clinically apparent in infancy–early childhood	`onset_category: INFANTILE` / `CHILDHOOD`
Progression	Static (non-progressive) encephalopathy	`clinical_course: STABLE`
Regression	Absent — `HP:0002376` developmental regression is a red flag against AR‑NSID and toward a metabolic/degenerative diagnosis	Curate as a differential-diagnosis discriminator
Duration	Lifelong	`CHRONIC`

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)

Gene	HGNC	MRT	Protein / function	Key citation
PRSS12	`hgnc:9477`	MRT1	Neurotrypsin — presynaptic serine protease; cleaves agrin	PMID:12459588
CRBN	`hgnc:30185`	MRT2	Cereblon; CUL4-DDB1 E3 ligase substrate receptor	Higgins 2004
CC2D1A	`hgnc:30237`	MRT3	Freud-1/Aki1; NF‑κB and cAMP–PKA–PDE4D regulator	PMID:16033914
GRIK2	`hgnc:4580`	MRT6	Kainate receptor GluK2	Motazacker 2007
TUSC3	`hgnc:30242`	MRT7	OST complex subunit; N-glycosylation	PMID:18452889; PMID:21513506
TRAPPC9	`hgnc:30832`	MRT13	NIBP; TRAPPII subunit; NIK/IKKβ-binding	PMID:20004763; PMID:20004765
NSUN2	`hgnc:25994`	MRT5	tRNA m⁵C methyltransferase	PMID:22541559; PMID:22541562
TECR	`hgnc:4551`	MRT14	trans-2-enoyl-CoA reductase; VLCFA elongation	Çalışkan 2011
MAN1B1	`hgnc:6823`	MRT15	ER α-1,2-mannosidase; glycoprotein quality control	PMID:21763484
ST3GAL3	`hgnc:10866`	MRT12	Sialyltransferase	Hu 2011
MED23	`hgnc:2372`	MRT18	Mediator complex subunit 23	Hashimoto 2011
ELP2	`hgnc:18248`	MRT58	Elongator complex subunit 2	Cohen 2015
IMPA1	`hgnc:6050`	MRT59	Inositol monophosphatase 1 (lithium target)	Figueiredo 2016
METTL23	`hgnc:26988`	MRT44	Methyltransferase-like 23	Bernkopf/Reiff 2014
LINS1	`hgnc:30922`	MRT27	Wnt-signalling regulator	Akawi 2013
ZNF526	`hgnc:29415`	—	Zinc-finger transcription factor	Najmabadi 2011 cohort
CDK5R1	`hgnc:1775`	—	p35, CDK5 activator (17q11.2 candidate)	PMC9946902
NDST1	`hgnc:7680`	MRT46	Heparan sulfate N-deacetylase/N-sulfotransferase
TNIK	`hgnc:30765`	MRT54	TRAF2/NCK-interacting kinase
PGAP1	`hgnc:25712`	MRT42	GPI-anchor remodelling
WASHC4 (KIAA1033/SWIP)	`hgnc:29174`	MRT43	WASH complex — endosomal actin	Ropers 2011
TRAPPC6B	`hgnc:23066`	—	TRAPP complex subunit (new in Harripaul 2018)	PMID:28397838
MAPK8 (JNK1)	`hgnc:6881`	—	Stress-activated MAP kinase	PMID:28397838
MPDZ	`hgnc:7208`	—	Multi-PDZ domain scaffold	PMID:28397838
TBC1D23	`hgnc:25622`	—	Golgi–endosome tethering	PMID:28397838
ADAT3	`hgnc:25151`	MRT36	tRNA adenosine deaminase	Recurrent in Arab populations
CRADD	`hgnc:2340`	MRT34	PIDDosome; lissencephaly (syndromic)
VPS13B	`hgnc:2183`	—	Cohen syndrome (syndromic on full workup)
AP4M1	`hgnc:574`	—	AP-4 deficiency / SPG50 (syndromic)
ADK	`hgnc:257`	—	Adenosine kinase deficiency (metabolic)

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:

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.
Severity modulation. Iodine deficiency, malnutrition, and educational deprivation act additively on adaptive outcome.
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 term	Label
`GO:0007399`	nervous system development
`GO:0022008`	neurogenesis
`GO:0030182`	neuron differentiation
`GO:0048666`	neuron development
`GO:0021987`	cerebral cortex development
`GO:0021895`	cerebral cortex neuron differentiation
`GO:0016358`	dendrite development
`GO:0030030`	cell projection organization
`GO:0050808`	synapse organization
`GO:0007268`	chemical synaptic transmission
`GO:0050804`	modulation of chemical synaptic transmission
`GO:0007611` / `GO:0007612`	learning or memory / learning
`GO:0016192`	vesicle-mediated transport
`GO:0006487`	protein N-linked glycosylation
`GO:0007249`	canonical NF-kappaB signal transduction
`GO:0001510` / `GO:0006400`	RNA methylation / tRNA modification
`GO:0006417`	regulation of translation

⚠️ 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

Theme	Genes	GO anchor
Membrane/vesicle trafficking	TRAPPC9, TRAPPC6B, WASHC4, TBC1D23, PGAP1, AP4M1	`GO:0016192`
Glycosylation & glycoprotein QC	TUSC3, MAN1B1, ST3GAL3, PGAP1, NDST1	`GO:0006487`
RNA/tRNA modification & translation	NSUN2, ADAT3, ELP2, METTL23	`GO:0001510`, `GO:0006400`, `GO:0006417`
Transcriptional regulation	MED23, ZNF526, TET1	—
Synaptic function & proteolysis	PRSS12, GRIK2, MPDZ	`GO:0050808`, `GO:0007268`
Intracellular signalling	CC2D1A, TRAPPC9, MAPK8, TNIK, LINS1 (Wnt)	`GO:0007249`
Metabolism / lipid	TECR, ADK, IMPA1	—
Cell-cycle / cytoskeleton / neurogenesis	CDK5R1, ABI2, SLAIN1, CRADD	`GO:0022008`

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

Level	Structure	Ontology term (verified)	Involvement
System	Nervous system	`UBERON:0001016` (verify)	Primary and, by definition, sole
Organ	Brain	`UBERON:0000955`	Primary
	Cerebral cortex	`UBERON:0000956`	Primary — TRAPPC9 and CC2D1A both show highest expression here
	Hippocampal formation / Ammon's horn	`UBERON:0002421` / `UBERON:0001954`	Learning-and-memory substrate; disproportionately reduced in Trappc9 KO
	Cerebral white matter	`UBERON:0002316`	Hypoplasia in TRAPPC9; reduced DTI-derived integrity in the KO mouse
	Cerebellum	`UBERON:0002037`	Mild atrophy in isolated families only
Tissue	Nervous tissue; grey and white matter
Cell	Neuron	`CL:0000540`	Principal target
	Glutamatergic / GABAergic neuron	`CL:0000679` / `CL:0000617`	E/I balance
	Neural stem cell	`CL:0000047`	Sox2⁺ NSPC depletion (Trappc9)
	Oligodendrocyte	`CL:0000128`	White matter arm
	Astrocyte	`CL:0000127`	Supporting
Subcellular	Synapse / presynapse	`GO:0045202` / `GO:0098793` (verify)	PRSS12, GRIK2, MPDZ
	Axon / dendrite	`GO:0030424` / `GO:0030425` (verified)	TRAPPC9
	Endoplasmic reticulum	`GO:0005783` (verify)	TUSC3, MAN1B1, TECR
	Golgi / recycling endosome	`GO:0005794` / `GO:0055037` (verify)	TRAPPC9, TBC1D23, WASHC4
	Nucleolus	`GO:0005730` (verify)	NSUN2 mislocalization (PMID:22541562)

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

Quantity	Value	Source
ID overall prevalence	1–3% of all populations	PMC9946902; PMID:28397838 ("Approximately 1% of the global population is affected by intellectual disability") [verbatim]; PMID:20004765 ("It occurs with a prevalence of approximately 2%") [verbatim]
ARID as fraction of diagnosable ID, outbred populations	~10%	PMID:30459488 — "we estimate that ARID makes up about 10% of all diagnosable ID cases in an outbred population" [fragment]
ARID in consanguineous populations	Leading genetic cause	PMID:29302074 [verbatim]
Total ID risk, consanguineous vs outbred	2–3× higher	PMID:30459488 [fragment]
Consanguineous share of world population	~10.4% (second cousins or closer)	Bittles & Black
Recessive share of point mutations in a highly consanguineous ID cohort	81%	PMID:27431290 — "The identified point mutations were mostly recessive (n=117, 81%)" [verbatim]

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):

Setting	Modality	Yield	Source
337 ID subjects (highly consanguineous)	Genomics-first (CMA + panel + ES)	58% — vs 16% suggested / 11% confirmed by standard clinical evaluation	PMID:27431290
Same cohort, ES applied to all CMA-negative cases	ES	60% (77/129)	PMID:27431290
192 multiplex Pakistani/Iranian consanguineous families	Microarray + HBD + CNV + WES	51% of families, 72 genes	PMID:28397838
121 large consanguineous Pakistani families	WES	60/121 (~50%) with a single-gene answer (30 known + 30 novel genes)	PMID:27457812
404 consanguineous families (mostly Iranian)	WES/WGS	219/404 (54%) likely causative; 77 known + 77 novel AR‑ID genes, 21 X-linked, 9 previously non-ID genes	PMID:29302074
Mixed ID (non-consanguineous), meta-analysis	ES	36% overall NDD; 31% isolated (non-syndromic) NDD; 53% syndromic NDD	cited in PMID:34930158
Genome sequencing, cumulative	GS	62% (de novo SNV 39%, de novo CNV 21%, recessive 2% — outbred cohort)	Gilissen, cited in PMID:34930158
Iran, 2025	Proband-only ES	40.4%, → 45.4% with CNV, → 50.5% with parental testing	Ghalamkari 2025, AJMG A

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

Category	Discriminating feature
De novo autosomal dominant ID	The dominant cause in outbred/simplex settings; distinguished only by trio sequencing
X-linked ID	Male-predominant pedigree; maternal transmission; FMR1 first
Syndromic ARID	Dysmorphism, malformation, growth or organ involvement — often only apparent on re-examination or with age
Chromosomal / CNV disorders	CMA-detectable
Treatable inborn errors of metabolism	PKU, homocystinuria, creatine deficiency syndromes, biotinidase deficiency, CDG. Highest-priority exclusion — these change management
Congenital hypothyroidism	Newborn screening; treatable
Fetal alcohol spectrum disorder	Exposure history; facial features
Congenital infection (CMV, Zika, rubella, toxoplasma)	Serology/PCR; imaging calcifications; hearing loss
Perinatal hypoxic-ischaemic encephalopathy	Birth history; MRI pattern
Cerebral palsy with ID	Motor signs predominate
Autism spectrum disorder without ID	Formal cognitive testing separates them
Progressive/degenerative disease	Regression (`HP:0002376`) excludes AR‑NSID — the key discriminator
Severe psychosocial deprivation	History; partial catch-up with intervention

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)

Intervention	NCIT term (verified via OAK)	`therapeutic_modality`
Early intervention / developmental therapy	`NCIT:C15315` Rehabilitation	`BEHAVIORAL`
Speech and language therapy	`NCIT:C159273` Speech Language Therapy	`BEHAVIORAL`
Occupational therapy	`NCIT:C121351` Occupational Therapy	`BEHAVIORAL`
Physical therapy	`NCIT:C15302` Physical Therapy	`BEHAVIORAL`
Special education / individualized education programme	`NCIT:C181743` Behavioral Counseling (nearest available)	`BEHAVIORAL`
Applied behaviour analysis / behavioural intervention	`NCIT:C181743` Behavioral Counseling	`BEHAVIORAL`
Genetic counselling	`NCIT:C15240` Genetic Counseling	`OTHER`
Supportive / multidisciplinary care	`NCIT:C15747` Supportive Care	`OTHER`

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

Gene	Models	Phenotype
NSUN2	Drosophila ortholog deletion; mouse KO	Severe short-term-memory deficit, rescuable by WT re-expression (PMID:22541559)
PRSS12	Mouse KO; Drosophila tequila	Impaired agrin cleavage; LTM formation defect in fly
TUSC3	Mouse KO; yeast OST-complex biology	Hypoglycosylation
GRIK2	Grik2/GluR6 KO mouse	Altered hippocampal synaptic transmission and LTP

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:

ID is not modellable. Rodent learning/memory assays are proxies for a construct (IQ + adaptive functioning) with no animal equivalent.
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.
Sex-dimorphic rescue. Cc2d1a PDE4-inhibitor rescue works in males only; the human correlate is unknown.
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.
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:

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.
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.
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.
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.
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.
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.
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):

PMID	Citation
21937992	Najmabadi H, et al. Deep sequencing reveals 50 novel genes for recessive cognitive disorders. Nature. 2011;478(7367):57-63
28397838	Harripaul R, et al. Mapping autosomal recessive intellectual disability: combined microarray and exome sequencing identifies 26 novel candidate genes in 192 consanguineous families. Mol Psychiatry. 2018;23(4):973-984
27431290	Anwar/Alkuraya et al. Clinical genomics expands the morbid genome of intellectual disability and offers a high diagnostic yield. Mol Psychiatry. 2017;22(4):615-624
24176302	Musante L, Ropers HH. Genetics of recessive cognitive disorders. Trends Genet. 2014;30(1):32-9
27457812	Riazuddin S, et al. Exome sequencing of Pakistani consanguineous families identifies 30 novel candidate genes for recessive intellectual disability. Mol Psychiatry. 2017;22(11):1604-1614
29302074	Hu H, et al. Genetics of intellectual disability in consanguineous families. Mol Psychiatry. 2019;24(7):1027-1039
12459588	Molinari F, et al. Truncating neurotrypsin mutation in autosomal recessive nonsyndromic mental retardation. Science. 2002;298(5599):1779-81
20004763	Mochida GH, et al. A truncating mutation of TRAPPC9 is associated with autosomal-recessive intellectual disability and postnatal microcephaly. Am J Hum Genet. 2009;85(6):897-902
20004765	Mir A, et al. Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-beta-binding protein, in nonsyndromic autosomal-recessive mental retardation. Am J Hum Genet. 2009;85(6):909-15
32877400	Wang H, et al. Trappc9 deficiency causes parent-of-origin dependent microcephaly and obesity. PLoS Genet. 2020
26912939	Davies G, et al. Examining non-syndromic autosomal recessive intellectual disability (NS-ARID) genes for an enriched association with intelligence differences. Intelligence. 2016
34930158	Chiurazzi-group review. Intellectual disability genomics: current state, pitfalls and future challenges. BMC Genomics. 2021

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

Ask OpenScientist

OpenScientist Report

Inheritance

Subtypes

Mechanistic Hypotheses

Discussions and Knowledge Gaps

Pathophysiology

Pathograph

Phenotypes

Genetic Associations

Medical Actions

Diagnosis

Prevalence

Differential Diagnoses

Source YAML

References & Deep Research

References

Deep Research

Comprehensive Research Report

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

1. Disease Information

1.1 Overview

1.2 Key identifiers (verified)

1.3 Synonyms (from the MONDO record, verified)

1.4 Source type

2. Etiology

2.1 Primary causal factor

2.2 Genetic risk factors

2.3 Environmental risk factors

2.4 Protective factors

2.5 Gene–environment interactions

3. Phenotypes

3.1 The defining phenotype

3.2 Severity distribution

3.3 The Monarch-propagated annotation set

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

3.5 Neuroimaging

3.6 Temporal characteristics per phenotype

3.7 Quality of life

4. Genetic / Molecular Information

4.1 Scale of the gene set

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

4.3 Variant classification, type, and frequency

4.4 Support for pathogenicity beyond segregation

4.5 Epigenetics and imprinting

4.6 Chromosomal abnormalities

5. Environmental Information

6. Mechanism / Pathophysiology

6.1 The class-level causal chain

6.2 Worked mechanism 1 — PRSS12 / neurotrypsin: synaptic proteolysis

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

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

6.5 Recurrent mechanistic themes across the AR‑NSID gene set

6.6 Immune, metabolic, and tissue-damage mechanisms

6.7 Molecular profiling

7. Anatomical Structures Affected

8. Temporal Development

9. Inheritance and Population

9.1 Epidemiology

9.2 Inheritance mechanics

9.3 Population demographics

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

10. Diagnostics

10.1 Genetic testing — the diagnostic backbone

10.2 Laboratory and biochemical testing

10.3 Imaging and electrophysiology

10.4 Clinical diagnostic criteria

10.5 Differential diagnosis

10.6 Screening

11. Outcome / Prognosis

12. Treatment

12.1 Supportive and rehabilitative care (the mainstay)

12.2 Pharmacotherapy — symptomatic only

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

12.4 Advanced therapeutics

12.5 Clinical trials

12.6 Pharmacogenomics

13. Prevention

14. Other Species / Natural Disease

15. Model Organisms

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

15.6 Shared limitations (candidates for `kind: HUMAN_MODEL_MISMATCH`)