ARID1A c.1217_1229dup, p.(Pro411ThrfsTer216)

Reasoning

No ARID1A-specific ClinGen VCEP PVS1 specification is available for this case (cspec.found=false, final_classification_framework.found=false, framework_mode=generic_acmg), so the generic ClinGen SVI PVS1 decision tree (PMC6185798) governs. NM_006015.5:c.1217_1229dup is a 13-bp in-frame-shifting duplication producing NP_006006.3:p.(Pro411ThrfsTer216), a frameshift starting at residue 411 with a new stop codon at codon 626 of the ~2285-residue ARID1A protein. Using the transcript exon/CDS map (exon2 c.1138-1350 harbors the duplication start; CDS total length c.1-6858; final exon begins at c.5125), the premature termination codon falls at approximately c.1876 (~codon 626), which lies well upstream (>50 nt) of the last exon-exon junction and is not located in the final exon or within the terminal 50 bp of the penultimate exon. The transcript is therefore predicted to undergo nonsense-mediated decay (NMD) rather than produce a stable truncated/elongated protein product. Germline LOF is an established ARID1A disease mechanism: pvs1_gene_context confirms germline disease-context literature (BAFopathies/Coffin-Siris spectrum and related developmental phenotypes attributable to ARID1A haploinsufficiency) supporting loss-of-function as the operative mechanism, satisfying the gene-level PVS1 gate. The variant affects a canonical, constitutively used coding exon (exon 2 of 20) with no evidence of biologically irrelevant/alternatively-spliced status or of population-level enrichment for LoF variants at this position, so no exon-importance-based attenuation is warranted. There is no reason to invoke a splicing-mechanism alternative pathway (SpliceAI predicts no significant splice impact, max delta 0.108, consistent with a straightforward frameshift/NMD mechanism rather than exon skipping). Per the ClinGen SVI PVS1 decision tree, a predicted-NMD frameshift/truncating variant in a gene with an established, non-exon-skip-modifiable LOF disease mechanism, affecting a biologically relevant constitutive exon, meets PVS1 at full (very strong) strength.

Rule

PMC6185798 (ClinGen SVI PVS1 recommendations) generic decision tree: null variant (frameshift) predicted to trigger NMD, in a gene with established loss-of-function disease mechanism, affecting a biologically relevant constitutive exon -> PVS1 at very strong strength.

Evidence considered

• pvs1_variant_assessment classifies the variant as consequence_class=frameshift, variant_bucket=frameshift, with suggested_default_strength=PVS1 and apply_generic_pvs1_framework=true under the PMC6185798 framework.
• pvs1_gene_context reports pvs1_gene_gate=eligible and lof_mechanism_supported=true, citing germline disease-focused literature (including ARID1A gene variants and fetal hydrocephalus / BAFopathy spectrum, PMID:40962111) as evidence that ARID1A loss-of-function is an established germline disease mechanism.
• Mutalyzer/VariantValidator exon and CDS coordinate maps (prefetch.json) place the c.1217_1229 duplication in exon 2 (c.1138-1350) and the resulting premature stop (~codon 626, ~c.1876) far upstream of the final exon (beginning c.5125) and outside the terminal 50 bp of the penultimate exon, consistent with NMD-predicted transcript decay rather than protein truncation/elongation escape.
• SpliceAI (spliceai) shows no significant splice-altering signal (max delta score 0.108), so the frameshift/NMD mechanism is not confounded by an alternative exon-skipping rescue pathway.
• No VCEP-specific PVS1 exon-importance caveats or population LoF-enrichment flags were found for this exon in the available materials.

Still missing

• No official ARID1A-specific ClinGen VCEP PVS1 specification exists; assessment relies on the generic SVI framework and literature-derived gene-level LOF mechanism support rather than a locus-specific expert panel document.
• Exact intron/exon boundary and NMD-junction distance were derived from Mutalyzer selector coordinates rather than an experimentally validated NMD assay for this specific variant.

Sources

Reasoning

PM2 is met at supporting strength. The exact 13-bp duplication NM_006015.5:c.1217_1229dup, corresponding to NC_000001.10:g.27056221_27056233dup and the GRCh38 locus queried in the bundle, is absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0. This absence is compatible with a rare disease-associated variant, but the bundle lacks allele-number, coverage, and ancestry-stratified metrics needed to quantify residual detection power and penetrance-related expectations.

Rule

Generic ACMG/AMP PM2 supporting evidence: the variant is absent or extremely rare in population databases appropriate for the disorder. No ARID1A-specific population rule was retrieved; absence is used only at supporting strength because database sample size, coverage, ancestry representation, disease penetrance, and prevalence thresholds are not quantified.

Evidence considered

• gnomAD v2.1 reports the exact variant as absent.
• gnomAD v4.1 reports the exact variant as absent.
• gnomAD-Canada v1.0 reports the exact variant as absent.
• The case bundle identifies the variant as a frameshift duplication in ARID1A, but no population allele counts or quantitative maximum credible allele-frequency threshold are provided.

Still missing

• No allele counts, total allele numbers, coverage metrics, or ancestry-specific results are supplied for the absent calls.
• No disease prevalence, penetrance, or gene-specific maximum credible allele-frequency calculation is available in the bundle.
• The absence call should be confirmed against current population-database records and transcript/genome normalization before final clinical use.

Sources

Reasoning

No ARID1A-specific VCEP CSPEC PP3/BP4 lookup was retrieved for this case (cspec.found=false; final_classification_framework.found=false), so the generic ACMG/AMP in-silico calibration in generic_acmg_classification_rules.md governs. This variant, NM_006015.5:c.1217_1229dup (p.(Pro411ThrfsTer216)), is a 13-bp coding duplication causing a frameshift, not a missense substitution, so REVEL is not applicable per the framework's mechanism-matching rule; the SpliceAI predictor path applies instead per the generic non-missense branch. SpliceAI max delta score is 0.108 (DS_DG=0.108 gain-of-donor being the largest component), which falls inside the defined gray zone [0.1, 0.2] where neither PP3 nor BP4 is met under this lab's SVI-aligned thresholds.

Rule

generic_acmg_pp3_bp4_spliceai

Evidence considered

• SpliceAI Lookup scores for NM_006015.5:c.1217_1229dup: DS_AG=0.01, DS_AL=0.002, DS_DG=0.108, DS_DL=0.003; max delta score = 0.108, predicting no significant splice impact.
• Generic ACMG/AMP in-silico calibration (this lab's internal SVI-aligned thresholds, reflected in VCEP CSPECs such as ATM's per Walker et al. 2023): SpliceAI max delta >0.2 -> PP3, <0.1 -> BP4, [0.1,0.2] -> neither met. 0.108 falls in the neither-met gray zone.

Still missing

• SpliceAI max delta (0.108) sits in the indeterminate zone (0.1-0.2), precluding a PP3 call under the generic calibration.
• No gene-specific VCEP CSPEC PP3/BP4 lookup spreadsheet was available for ARID1A to override the generic threshold.

Sources

Reasoning

BA1 is not met. The exact variant NM_006015.5:c.1217_1229dup is reported absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0; therefore there is no population allele frequency at or above a stand-alone benign threshold.

Rule

Generic ACMG/AMP fallback applies because no ARID1A-specific CSPEC/VCEP framework was retrieved. BA1 requires a population frequency high enough to establish stand-alone benign evidence; an absent variant cannot satisfy that requirement.

Evidence considered

• gnomAD v2.1 direct variant query completed with search_status=absent and the evidence sentence 'Absent from gnomAD v2.1.'
• gnomAD v4.1 direct variant query completed with search_status=absent and the evidence sentence 'Absent from gnomAD v4.1.'
• gnomAD-Canada v1.0 query completed with search_status=absent and the evidence sentence 'Absent from gnomAD-Canada v1.0.'

Still missing

• The bundle does not provide allele-number, coverage, or ancestry-stratified counts, so a quantitative threshold comparison cannot be independently reproduced from the bundle alone.

Sources

Reasoning

BS1 is not met. The exact variant is absent from the three population datasets queried, so no allele frequency above the expected disease-prevalence threshold is available.

Rule

Generic ACMG/AMP fallback: BS1 requires an allele frequency greater than expected for the disorder; absence from population databases provides no BS1 frequency evidence.

Evidence considered

• The gnomAD v2.1 exact variant query returned absent.
• The gnomAD v4.1 exact variant query returned absent.
• The gnomAD-Canada v1.0 exact variant query returned absent.

Still missing

• The bundle does not provide allele-number, coverage, or ancestry-stratified counts, so a quantitative BS1 threshold comparison cannot be independently reproduced from the bundle alone.

Sources

Reasoning

Same SpliceAI-path analysis as PP3: no ARID1A VCEP CSPEC PP3/BP4 lookup was available, so the generic non-missense SpliceAI calibration applies to this frameshift duplication. Max delta score = 0.108 falls just inside the gray zone [0.1, 0.2], above the BP4 threshold of <0.1, so BP4 is not met under this framework. REVEL/BayesDel are not applicable since this is not a missense substitution (BayesDel would in any case be unusable per pipeline policy absent a verified published threshold/PMID).

Rule

generic_acmg_pp3_bp4_spliceai

Evidence considered

• SpliceAI Lookup scores for NM_006015.5:c.1217_1229dup: max delta score = 0.108 (evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.11)').
• Generic ACMG/AMP in-silico calibration: SpliceAI max delta <0.1 -> BP4 (supporting). 0.108 does not clear this threshold, so BP4 is not applied.

Still missing

• SpliceAI max delta (0.108) is just above the BP4 cutoff of <0.1, leaving it in the indeterminate zone rather than clearly benign-supporting.
• No gene-specific VCEP CSPEC PP3/BP4 lookup spreadsheet was available for ARID1A.

Sources

Reasoning

No case-level evidence is provided for a confirmed de novo occurrence in the proband. Parental genotypes, maternity/paternity confirmation, and phenotype information are unavailable, so PS2 cannot be assessed.

Rule

Generic ACMG/AMP PS2: de novo occurrence in a patient with the disease and no family history, with confirmed maternity and paternity; stronger use requires appropriate phenotype consistency and parental testing quality.

Evidence considered

• The case-evidence bundle contains no proband phenotype description, parental genotypes, trio testing result, or maternity/paternity confirmation for NM_006015.5:c.1217_1229dup.
• The available literature packet contains functional/somatic ARID1A papers and does not provide a documented de novo observation for this variant.

Still missing

• Documented parental testing showing the variant is absent in both biological parents
• Confirmation of maternity and paternity or equivalent relationship verification
• Proband phenotype consistent with the relevant ARID1A-associated disorder

Reasoning

No ClinGen VCEP or gene-specific framework is available for ARID1A (cspec.found=false; framework_mode=generic_acmg), so generic ACMG/AMP PS3 rules apply, which require a well-established functional assay demonstrating a damaging effect specific to this variant. Literature retrieved (PMID:21900401, PMID:22009941, PMID:24899687, PMID:25625625) consists of gene-level ARID1A tumor-suppressor functional studies (SWI/SNF complex formation, mouse knockout ovarian tumorigenesis models, pan-cancer somatic mutation surveys, ARID1A-PIK3CA cooperativity in OCCC mouse models). None of these publications report a functional assay (e.g., complementation, rescue, reporter, or biochemical activity assay) performed on the specific variant NM_006015.5:c.1217_1229dup / p.(Pro411ThrfsTer216); a full-text search for the coding change, protein change, or residue P411 across the retrieved full-text papers returned no matches. No variant-specific functional assay evidence is therefore available to apply PS3.

Evidence considered

• case_summary.json cspec.found=false and vcep_materials.json final_classification_framework.found=false confirm no ARID1A-specific VCEP/CSPEC framework is available for this case; framework_mode is generic_acmg.
• literature_pass.json high_priority_papers (PMID:21900401, PMID:22009941, PMID:24899687, PMID:25625625) are gene-level ARID1A functional/tumor-suppressor studies, not variant-specific functional assays for c.1217_1229dup.
• Targeted search of full-text files under publications/ for the variant coding change, protein change, and residue number (P411) returned zero matches, confirming none of the retrieved papers characterize this specific variant.

Still missing

• No variant-specific functional/complementation assay data for p.(Pro411ThrfsTer216) identified in the retrieved literature packet.
• No ClinGen ARID1A VCEP specification available to define an assay-based PS3 rule or strength calibration for this gene.

Reasoning

No case-control cohort, statistically significant disease enrichment, or eligible affected-case series for the exact variant was provided. The available literature entries concern ARID1A gene-level cancer biology and do not establish PS4 for this variant.

Rule

Generic ACMG/AMP PS4: use statistically significant enrichment in affected individuals or affected cases compared with appropriate controls; no applicable ARID1A-specific VCEP framework was retrieved.

Evidence considered

• The consolidated case-evidence bundle contains no variant-specific case-control counts, statistical enrichment analysis, or eligible affected-case series for NM_006015.5:c.1217_1229dup.
• The available literature packet is gene-level functional/somatic cancer literature and does not report this exact germline variant in an evidence set establishing PS4.

Still missing

• Variant-specific case-control enrichment data or a sufficiently documented affected-case series with an appropriate comparison group.

Reasoning

No affected-proband observations, a confirmed pathogenic variant in trans, phase data, or inheritance information were provided for this ARID1A variant. The available evidence bundle contains no allelic or segregation observations, so PM3 cannot be assessed.

Rule

Generic ACMG/AMP PM3: pathogenic or likely pathogenic variants in trans with the variant under assessment support the criterion; strength depends on the number and quality of informative observations. No qualifying observation is available here.

Evidence considered

• The governing framework search found no ARID1A-specific VCEP/CSPEC specification; generic ACMG/AMP fallback therefore applies.
• The consolidated case-evidence bundle contains no affected-proband, phase, in-trans comparator, or inheritance evidence for PM3.

Still missing

• Affected-proband case observations are absent.
• No confirmed pathogenic or likely pathogenic variant in trans is documented.
• No phase information (cis/trans) is documented.
• No relevant inheritance or parental testing information is documented.

Reasoning

No case-level evidence is provided for a presumed de novo occurrence. Without parental testing or documented circumstances supporting de novo status, PM6 cannot be assessed.

Rule

Generic ACMG/AMP PM6: assumed de novo, without confirmation of paternity and maternity; it is weaker than PS2 and requires documented case-level de novo context.

Evidence considered

• The case-evidence bundle does not report parental testing, a de novo call, family history, or phenotype details for this variant.
• No reviewed publication in the bundle documents this exact variant as presumed de novo.

Still missing

• A documented presumed-de-novo observation in the proband
• Available parental testing or an explicit explanation of why parental testing was unavailable
• Phenotype and family-history information supporting a de novo interpretation

Reasoning

No affected relatives, informative meioses, or segregation data are available. Therefore, cosegregation of the variant with disease cannot be assessed.

Rule

Generic ACMG/AMP PP1: cosegregation with disease in multiple affected family members in a gene definitively known to cause the disease; strength may be increased when multiple informative meioses support segregation.

Evidence considered

• The case-evidence bundle contains no pedigree, family-history narrative, affected-relative testing, or segregation analysis for NM_006015.5:c.1217_1229dup.
• The literature packet does not contain a family-based report of this exact variant that could provide segregation evidence.

Still missing

• Pedigree with clinically characterized relatives
• Genotypes and phenotypes for affected relatives
• Number of informative meioses and segregation results
• Assessment of phenotype specificity and possible phenocopies

Reasoning

The case bundle does not provide the patient's phenotype, a defined ARID1A-associated disease diagnosis, or evidence that the phenotype is highly specific for a disorder caused by ARID1A variation. PP4 therefore cannot be assessed.

Rule

Generic ACMG/AMP PP4: phenotype or family history is highly specific for a disease with a single-gene etiology; generic cancer gene-level descriptions are insufficient.

Evidence considered

• case_summary.json identifies the variant and gene but provides no patient phenotype or disease-specific clinical findings.
• No applicable ARID1A-specific VCEP specification was retrieved; generic ACMG/AMP is the fallback framework.

Still missing

• A documented patient phenotype and clinical diagnosis mapped to a specific ARID1A-associated disorder, including assessment of phenotypic specificity.

Reasoning

PP5 requires an exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification under the requested rule. The case bundle reports that this exact variant is absent from ClinVar, with no expert-panel assertion; OncoKB and ordinary literature context cannot substitute for the required ClinVar expert-panel evidence.

Rule

PP5 is applied only for an exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification; ordinary submissions, aggregate labels, non-expert assertions, and external database labels do not qualify.

Evidence considered

• The ClinVar search in the case bundle returned no exact-variant result: variation ID, classification, review status, submitter count, and expert panels are all absent or empty.
• The available OncoKB record is not a ClinVar expert-panel assertion and was not used to trigger PP5.

Still missing

• An exact-variant ClinVar expert-panel classification of Pathogenic or Likely pathogenic.

Sources

Reasoning

BS2 cannot be assessed from the available evidence. No source reports healthy individuals homozygous for this variant, but the population query outputs contain only absence calls and do not provide homozygote counts, genotype-quality details, or a sufficiently characterized healthy cohort.

Rule

Generic ACMG/AMP BS2 requires observation of the variant in a healthy adult or healthy individuals in a recessive disorder context, typically including homozygous or hemizygous observations as applicable; absence alone is insufficient to assess BS2.

Evidence considered

• All three population sources report the exact variant as absent, but none supplies homozygote counts or a direct healthy-homozygote observation.

Still missing

• No healthy homozygote data or explicit homozygote-count field is available.
• No genotype-quality, coverage, or ancestry-specific information is available to distinguish true absence from a population-database detection limitation.

Sources

Reasoning

No ClinGen VCEP or gene-specific framework is available for ARID1A, so generic ACMG/AMP BS3 rules apply, which require a well-established functional assay showing no damaging effect specific to this variant. No such variant-specific functional data (benign-calibrated assay result) for NM_006015.5:c.1217_1229dup was identified in the retrieved literature packet or case bundle; the retrieved papers are gene-level ARID1A tumor-suppressor characterizations, not variant-specific assays, and none mention this variant.

Evidence considered

• Same literature and framework review as PS3: no variant-specific functional assay evidence (damaging or benign) was found for this variant in the retrieved sources.

Still missing

• No variant-specific functional assay data of any kind (damaging or benign) identified for p.(Pro411ThrfsTer216).

Reasoning

No unaffected relatives have been tested or described, and no informative non-segregation analysis is available. BS4 cannot be assessed.

Rule

Generic ACMG/AMP BS4: lack of segregation in an affected family or observation of the variant in unaffected relatives when the phenotype is expected to be highly penetrant; interpretation requires informative relatives and appropriate phenotype assessment.

Evidence considered

• The case-evidence bundle provides no unaffected-relative genotypes, pedigree, age-dependent phenotype assessment, or non-segregation analysis for this variant.
• No reviewed publication in the bundle reports non-segregation of this exact variant with the relevant phenotype.

Still missing

• Genotypes of clinically unaffected relatives at appropriate ages
• Reliable phenotype characterization and penetrance considerations
• Informative non-segregation results demonstrating absence of the variant in unaffected affected-risk relatives

Reasoning

No confirmed observation of this variant in cis with a pathogenic variant, or in trans with a benign variant, was provided. The evidence bundle contains no phase-resolved allelic observations, so BP2 cannot be assessed.

Rule

Generic ACMG/AMP BP2: observation of the variant in cis with a pathogenic variant, or in trans with a benign variant, may support the criterion when phase and classifications are reliable. No qualifying observation is available here.

Evidence considered

• The governing framework search found no ARID1A-specific VCEP/CSPEC specification; generic ACMG/AMP fallback therefore applies.
• The consolidated case-evidence bundle contains no cis/trans phase data or other allelic evidence relevant to BP2.

Still missing

• No phase-resolved observation of this variant in cis with a pathogenic variant is documented.
• No phase-resolved observation of this variant in trans with a benign variant is documented.
• No family-based or specimen-based allelic testing information is documented.

Reasoning

No patient phenotype, tumor diagnosis, or independent pathogenic molecular finding establishing an alternative molecular etiology was provided. The available gene-level somatic cancer literature does not establish BP5 for this case.

Rule

Generic ACMG/AMP BP5: a strong alternative molecular basis for the patient's phenotype should be present; unrelated gene-level cancer biology is insufficient.

Evidence considered

• The case-evidence bundle contains no alternative pathogenic variant, established alternative molecular diagnosis, or clinical context from which BP5 could be evaluated.
• Absence of the exact variant from COSMIC is not evidence for an alternative molecular etiology and was not used for BP5.

Still missing

• A documented phenotype or disease context and a credible alternative molecular etiology that explains the findings better than this ARID1A variant.

Reasoning

BP6 requires an exact-variant ClinVar expert-panel Benign or Likely benign classification under the requested rule. The case bundle reports that this exact variant is absent from ClinVar, with no expert-panel assertion.

Rule

BP6 is applied only for an exact-variant ClinVar expert-panel Benign or Likely benign classification; ordinary submissions, aggregate labels, non-expert assertions, and external database labels do not qualify.

Evidence considered

• The ClinVar search in the case bundle returned no exact-variant result: variation ID, classification, review status, submitter count, and expert panels are all absent or empty.
• No ordinary laboratory submission, aggregate ClinVar label, non-expert assertion, or external database label was used to trigger BP6.

Still missing

• An exact-variant ClinVar expert-panel classification of Benign or Likely benign.

Sources