LYFE Sciences · Project HERA
Variant Interpretation · Classification Report
Generated: 2026-07-19
Case ID: NM_003073.4_c.1091_1093del_20260719_044940
Framework: ACMG/AMP 2015
Variant classification summary

SMARCB1  · NP_003064.2:p.(Lys364del)  · NM_003073.4
GRCh37: chr22:24175856 GAGA>G  ·  GRCh38: chr22:23833669 GAGA>G
Gene: SMARCB1 Transcript: NM_003073.4
Final call
All criteria require review: For research and educational purposes only.
Variant details
Gene
SMARCB1
Transcript
NM_003073.4
Protein
NP_003064.2:p.(Lys364del)
gnomAD AF
6.195065258817436e-07 (v4.1)
ClinVar
Pathogenic
OncoKB
Likely Oncogenic
Classification rationale
Interpretation summary
Generated evidence synthesis
1
NM_003073.4:c.1091_1093del (p.Lys364del) is an in-frame deletion in exon 8 of SMARCB1, removing a single lysine residue from the C-terminal domain alpha-helix.
2
The variant has been identified in 9 independent individuals with Coffin-Siris syndrome, a rare autosomal dominant neurodevelopmental disorder characterized by intellectual disability, coarse facial features, and hypoplastic fifth digit nails.
3
One confirmed de novo occurrence with both maternity and paternity confirmed has been reported in a CSS patient.
4
The variant is essentially absent from large population databases (gnomAD v4.1: 1/1,614,188 alleles; gnomAD v2.1: absent), consistent with a rare pathogenic variant.
5
Comprehensive functional studies by Valencia et al. (2019) directly demonstrated that the K364del variant disrupts SMARCB1 C-terminal domain binding to the nucleosome acidic patch, significantly attenuates mSWI/SNF complex nucleosome remodeling and ATPase activity, and impairs genome-wide DNA accessibility.
6
In iPSC-derived neuronal differentiation models, heterozygous K364del resulted in significantly diminished neurite outgrowth and reduced neuron counts, with phenotypic rescue upon restoration of wild-type SMARCB1, establishing a direct causal link between this variant and neurodevelopmental deficits.
7
Residue K364 is located within the SMARCB1 C-terminal alpha-helical domain (aa 357-377), a well-characterized functional domain that directly binds the nucleosome acidic patch and is essential for mSWI/SNF-mediated chromatin remodeling. This domain is a hotspot for CSS-associated mutations with no benign variation.
8
The variant is classified as Pathogenic in ClinVar (Variation ID: 30201) by 8 clinical laboratories; however, the aggregate review status is 'criteria provided, single submitter' (1-star), which does not meet the 3-star threshold for PP5 application.
9
Applying generic ACMG/AMP 2015 criteria: 2 strong (PS3, PS4), 4 moderate (PS2, PM1, PM2, PM4) pathogenic criteria are met. No benign criteria are met. This combination exceeds the threshold for Pathogenic classification (requires 2 strong OR 1 strong + ≥3 moderate).
Final determination:
Criteria assessment
ACMG/AMP criteria review
Criteria shown when status is available
All criteria require review: For research and educational purposes only.
Criterion Status Rationale Evidence used
PVS1 N/A NM_003073.4:c.1091_1093del is an in-frame deletion (p.Lys364del) that does not fall into the default PVS1 null-variant buckets (nonsense, frameshift, or canonical ±1,2 splice consensus). The variant does not trigger nonsense-mediated decay. Published literature indicates SMARCB1 mutations in Coffin-Siris syndrome exert dominant-negative or gain-of-function effects, excluding haploinsufficiency as the disease mechanism (Tsurusaki et al. 2012). PVS1 is not applicable for this non-null variant with a dominant-negative mechanism.
pvs1_generic_framework PMID:22426308 PMID:31759698
PS1 N/A PS1 applies to nucleotide substitutions where a different nucleotide change at the same position has been established as pathogenic. This is an in-frame deletion, not a nucleotide substitution.
PS2 Met A de novo occurrence of NM_003073.4:c.1091_1093del (p.Lys364del) with confirmed maternity and paternity was identified in one Coffin-Siris syndrome patient (subject 4) by Tsurusaki et al. 2012. Two additional CSS patients (subjects 21, 22) also harbor this variant but parental confirmation was not possible. One confirmed de novo in a disorder consistent with the variant's associated phenotype supports PS2 at moderate strength.
PMID:22426308
PS3 Met Valencia et al. 2019 (PMID:31759698) directly tested the K364del variant in comprehensive functional assays. mSWI/SNF complexes containing K364del SMARCB1 exhibited significantly attenuated nucleosome remodeling activity and reduced ATPase activity on nucleosome substrates compared to wild-type. The K364del mutation completely disrupted SMARCB1 CTD binding to the nucleosome acidic patch. In SMARCB1-deficient MRT cell rescue experiments, K364del mutant complexes showed normal genome-wide targeting but substantially diminished DNA accessibility (ATAC-seq) and reduced nucleosome eviction (MNase-seq). Heterozygous K364del iPSCs demonstrated decreased DNA accessibility at pluripotency loci and impaired neuronal differentiation with significantly diminished neurite outgrowth, rescued by wild-type SMARCB1 restoration. The exact variant was directly tested with unequivocal deleterious functional effects across biochemical, genomic, and cellular differentiation assays.
PMID:31759698
PS4 Met The p.Lys364del variant is the most recurrent SMARCB1 mutation in Coffin-Siris syndrome, identified in 9 independent CSS cases across published cohorts (Tsurusaki et al. 2012: 3 cases; Miyake et al. 2014: additional cases; Valencia et al. 2019: 9 total). The variant is essentially absent from population databases (gnomAD v4.1: 1/1,614,188 alleles, AF = 6.2e-07; gnomAD v2.1: absent). The extreme enrichment in affected individuals versus population controls meets PS4 at strong strength.
PMID:22426308 PMID:25081545 PMID:31759698 gnomad_v2 gnomad_v4
PS5 N/A PS5 is intended for situations where a reputable source reports the variant as pathogenic but the evaluating laboratory lacks access to the primary evidence. The evidence for this variant is directly available and has been independently evaluated; PS5 is not applicable.
PM1 Met The p.Lys364del variant is located in the SMARCB1 C-terminal domain (CTD; amino acids 357-377), a critical functional domain characterized by Valencia et al. 2019 as the nucleosome acidic patch binding domain. The CTD alpha-helix contains a dense cluster of basic residues required for mSWI/SNF complex-mediated nucleosome remodeling. This is a well-defined functional domain where missense and in-frame deletion mutations are enriched in both CSS and cancer, with no benign variation observed. PM1 is met at moderate strength based on the variant's location in this critical functional domain.
PMID:31759698
PM2 Met The variant is essentially absent from large population databases. gnomAD v2.1: 0 alleles. gnomAD v4.1: 1 allele in 1,614,188 (AF = 6.2e-07, 0.000062%), well below the PM2 threshold of <0.1%. This extreme rarity in the general population supports PM2 at moderate strength.
gnomad_v2 gnomad_v4
PM3 N/A PM3 applies to recessive disorders. Coffin-Siris syndrome is an autosomal dominant disorder; PM3 is not applicable.
PM4 Met NM_003073.4:c.1091_1093del is an in-frame deletion of a single amino acid (p.Lys364del) in a non-repeat region of SMARCB1. The deletion removes a critical residue within the CTD nucleosome acidic patch binding domain, resulting in a protein length change with demonstrated functional consequences. PM4 is met at moderate strength.
PMID:31759698
PM5 N/A PM5 requires a novel missense change at an amino acid residue where a different pathogenic missense change has been previously identified. This variant is an in-frame deletion, not a missense substitution. Additionally, PM5 candidate harvesting was unable to identify same-residue comparator missense variants with confirmed pathogenic classifications. PM5 is not applicable.
PM6 N/A PM6 applies to de novo observations without confirmed maternity and paternity. One de novo occurrence with confirmed parentage has already been applied as PS2 (moderate). The additional cases without parental confirmation (subjects 21, 22 in Tsurusaki et al. 2012) cannot be separately counted without double-counting the de novo evidence. PM6 is not applied to avoid evidence double-counting.
PP1 N/A PP1 requires cosegregation data in multiple affected family members. No family segregation data is available in the provided evidence. Coffin-Siris syndrome cases with this variant are sporadic with de novo occurrence.
PP2 N/A PP2 applies specifically to missense variants in genes with a low rate of benign missense variation. This is an in-frame deletion, not a missense variant.
PP3 Not met In silico pathogenicity prediction tools (REVEL, BayesDel) are not applicable for in-frame deletions and were not computed for this variant. SpliceAI predicts no significant splicing impact (max delta score = 0.00). No HCI prior score is available for SMARCB1. There is no computational evidence to support PP3.
spliceai
PP4 Not met While patients with this variant present with classical Coffin-Siris syndrome features (severe intellectual disability, coarse facies, hypoplastic fifth digit nails), CSS is genetically heterogeneous — caused by mutations in at least six different BAF complex genes (SMARCB1, SMARCA4, SMARCA2, SMARCE1, ARID1A, ARID1B). The phenotype is not highly specific for a single genetic etiology, and PP4 is therefore not met.
PMID:22426308 PMID:23637025
PP5 Not met While ClinVar classifies this variant as Pathogenic (Variation ID: 30201) with submissions from 8 clinical laboratories, the aggregate review status is 'criteria provided, single submitter' (1-star). PP5 requires ClinVar 3-star expert panel review to be applied at supporting strength. The current review status does not meet the 3-star threshold for PP5.
clinvar
BA1 Not met The variant has an allele frequency of 6.2e-07 (0.000062%) in gnomAD v4.1, far below the BA1 threshold of >5%. BA1 is not met.
gnomad_v4
BS1 Not met The variant has an allele frequency of 6.2e-07 (0.000062%) in gnomAD v4.1, far below the BS1 threshold of >0.3% for a dominant disorder. BS1 is not met.
gnomad_v4
BS2 Not met A single allele is observed in gnomAD v4.1 (1/1,614,188). CSS is a severe pediatric-onset dominant disorder with full penetrance expected early in life. The single observation may represent a mosaic carrier, technical artifact, or an individual with subclinical features. A single observation in a population database that screens out severe pediatric disease does not constitute sufficient evidence for BS2.
gnomad_v4
BS3 Not met Well-established functional studies (Valencia et al. 2019) demonstrate that K364del has a damaging effect on SMARCB1 function — disrupting nucleosome acidic patch binding, reducing chromatin remodeling activity, and impairing neuronal differentiation. BS3 requires functional studies showing no damaging effect; the available evidence demonstrates the opposite.
PMID:31759698
BS4 N/A BS4 requires lack of segregation in affected family members. No family segregation data is available; all reported cases are sporadic de novo occurrences.
BP1 N/A BP1 applies specifically to missense variants in genes where primarily truncating variants cause disease. This is an in-frame deletion, not a missense variant. Additionally, SMARCB1 CSS-associated mutations are non-truncating, so BP1 is not applicable.
BP2 N/A BP2 requires observation of the variant in trans with a pathogenic variant for a fully penetrant dominant disorder. CSS is dominant and all reported cases are heterozygous; no trans configuration data exists or is expected.
BP3 Not met BP3 applies to in-frame deletions in repetitive regions without known function. The K364del deletion occurs in the SMARCB1 C-terminal alpha-helical domain, which is a well-characterized functional domain (nucleosome acidic patch binding region). This is not a repetitive region without function. BP3 is not met.
PMID:31759698
BP4 Not met Computational predictors (REVEL, BayesDel) are not applicable for in-frame deletions. SpliceAI predicts no splicing impact (max delta = 0.00), which is neutral but does not constitute positive evidence for a benign effect on protein function. There is no computational evidence suggesting the variant is benign; BP4 is not met.
spliceai
BP5 N/A BP5 requires the variant to be found in a case with an alternative molecular basis for disease. No such data is available for this variant.
BP6 N/A BP6 requires a reputable source to report the variant as benign. ClinVar classifies this variant as Pathogenic (Variation ID: 30201). BP6 is not applicable.
BP7 N/A BP7 applies to synonymous variants with no predicted splicing impact. This is an in-frame deletion, not a synonymous variant.
Disclaimer: The content and results provided by LYFE Sciences are for research and educational purposes only and must not be used as a substitute for professional medical judgment, diagnosis, or treatment. Always consult a qualified healthcare professional before making any clinical decisions.