TSC2 c.3796_3797del, p.(Leu1266AlafsTer55)

Reasoning

NM_000548.4:c.3796_3797del in TSC2 is a 2-bp deletion causing a frameshift, NP_000539.2:p.(Leu1266AlafsTer55), i.e., a null variant under the ClinGen SVI PVS1 framework (PMC6185798). No TSC2 CSPEC/VCEP exists in this case (cspec found=false; vcep_materials empty), so the generic ACMG/AMP 2015 framework (PMID:25741868) applies with the SVI PVS1 specification, and pvs1_gene_context marks the gene gate 'eligible' (TSC2 LOF is an established germline disease mechanism: PMID:10205261 shows TSC1/TSC2 act as tumor suppressors with somatic loss or intragenic mutation of the wild-type allele in hamartomas and truncating mutations causing TSC; PMID:17304050 confirms TSC2 protein-truncating mutations are a major disease class, 80% of de novo cases). Transcript relevance: NM_000548.4 is TSC2 transcript variant 1, the principal transcript; ClinVar records this exact allele on NM_000548.5(TSC2):c.3796_3797del (p.Leu1266fs), so the variant is on the canonical transcript (NM_000548.4 vs .5 is only a RefSeq version difference). NMD: the frameshift introduces a premature termination codon at predicted amino acid 1320 (~c.3960-3962), which lies in exon 33 of 42, ~1300 nt upstream of the final exon-exon junction (c.5259/5260) — well beyond the 50-55 nt NMD threshold and not in the last exon or last 50 bp of the penultimate exon, so NMD is predicted with no escape features. Protein length change: 1808 aa wild type -> 1320 aa predicted mutant (loss of 488 aa, ~27% of the protein), deleting the C-terminal region including the RapGAP domain encoded by exons 34-38 (PMID:10205261), a critical functional domain — therefore no downgrade for a non-critical distal truncation. Repeat-region context: 2-bp deletion within a short repetitive run (normalization shift=2) but not in a polymorphic homopolymer; the allele is absent from gnomAD v2.1 and v4.1, so it is not a tolerated population LoF allele. SpliceAI max delta 0.01 indicates no competing splice effect. PVS1 is therefore met at very strong strength.

Rule

Generic ACMG/AMP 2015 (PMID:25741868) with ClinGen SVI PVS1 specification (PMC6185798, registry key pvs1_generic_framework): PVS1 (very strong) applies to a null variant (frameshift) in a gene with an established LOF disease mechanism when NMD is predicted and the variant is not in the last exon or last 50 bp of the penultimate exon.

Evidence considered

• pvs1_variant_assessment: variant classified as frameshift (variant_bucket=frameshift) on NM_000548.4, protein NP_000539.2:p.(Leu1266AlafsTer55), suggested_default_strength=PVS1 under PMC6185798; Mutalyzer protein prediction: position_last_original=1808, position_last_predicted=1320 (loss of 488 aa).
• pvs1_gene_context: TSC2 germline LOF mechanism supported, pvs1_gene_gate='eligible', with caution that no official CSPEC/VCEP exists and generic PVS1 fallback should follow PMC6185798.
• pvs1_generic_framework: ClinGen SVI PVS1 recommendations (PMC6185798) — null variant + established LOF mechanism + NMD predicted, not in last exon/last 50 bp of penultimate exon -> PVS1 very strong.
• PMID:10205261: TSC1/TSC2 established as tumor-suppressor genes; truncating/intragenic mutations of the wild-type allele in hamartomas; 'Exons 34-38 of TSC2 encode a GTPase-activating protein (GAP)-related domain' — supports LOF mechanism gene gate and criticality of the C-terminal region removed by this truncation.
• PMID:17304050: 325-individual TSC cohort; TSC2 mutations account for 80% of de novo cases and protein-truncating (PT) TSC2 mutations are a recognized disease-causing class — supports LOF mechanism gene gate.
• clinvar (variation 3724267): allele annotated on canonical transcript NM_000548.5(TSC2):c.3796_3797del (p.Leu1266fs) with germline origin — used only for transcript-relevance confirmation; the submitted Pathogenic/Likely pathogenic classifications were NOT used as evidence for PVS1.
• gnomad_v2 / gnomad_v4: variant absent from gnomAD v2.1 and v4.1 — no population LoF tolerance at this site, consistent with a null variant in a constrained gene.
• spliceai: max delta score 0.01 — no significant competing splice impact; frameshift remains the primary consequence.

Still missing

• No experimental RNA-based confirmation of NMD for this specific variant; NMD prediction is bioinformatic per SVI framework default.
• No TSC2-specific VCEP/CSPEC PVS1 specification available; generic SVI PVS1 fallback (PMC6185798) applied.
• The exact variant (c.3796_3797del / p.Leu1266AlafsTer55) is not reported in any full-text publication in this case bundle; PVS1 rests on consequence-based analysis.

Sources

Reasoning

PM2 (supporting) is met: the variant is absent from controls/population databases. Direct lookups returned no record in gnomAD v2.1 exomes (~125k alleles), gnomAD v4.1 exomes (the largest exome dataset, >700k individuals), and gnomAD-Canada v1.0 genomes (HostSeq), i.e. observed allele frequency 0, well below the local gnomAD non-VCEP PM2 threshold (<0.1%). TSC2 is an established early-onset autosomal dominant disease gene with loss-of-function as the germline mechanism (pvs1_gene_context: lof_mechanism_supported=true), and this variant is a frameshift (p.(Leu1266AlafsTer55)) in a highly LoF-constrained gene, so absence from large population datasets is expected for a pathogenic allele and is not a technical artifact. The Labcorp ClinVar submission independently corroborates 'this variant is not present in population databases (gnomad no frequency)' but was used only as corroboration (record flagged record_not_exact_match), never as the primary basis. PM2 applied at supporting strength per the generic ACMG/AMP 2015 framework.

Rule

PM2: absent in controls (or at extremely low frequency if recessive) in ESP/1000 Genomes/ExAC (ACMG/AMP 2015, PMID:25741868); local gnomAD non-VCEP framework threshold PM2<0.1%. Variant absent (AF=0) in all queried datasets.

Evidence considered

• PMID:25741868 — PM2 rule (absent in controls, or at extremely low frequency if recessive, in ESP/1000 Genomes/ExAC) governs; the variant is entirely absent, satisfying the rule without reliance on a numeric AF cutoff
• gnomad_v2 — variant absent from gnomAD v2.1 exomes (search_status=absent)
• gnomad_v4 — variant absent from gnomAD v4.1 exomes (search_status=absent)
• gnomad_canada — variant absent from gnomAD-Canada v1.0 genomes (search_status=absent)
• clinvar — Labcorp SCV005840460 corroborates 'this variant is not present in population databases (gnomad no frequency)'; corroboration only, not the primary basis
• PMID:17304050 — disease-context: early-onset autosomal dominant TSC with high de novo fraction supports interpreting absence from population databases as expected for a pathogenic TSC2 allele

Still missing

• Exact gnomAD allele counts and per-ancestry frequency breakdowns were not captured in the evidence bundle (source_data empty); absence rests on direct variant-page lookups returning no record (search_status=absent) plus screenshots (gnomad_v2.png, gnomad_v4.png).
• gnomAD v4.1 genome dataset was not queried separately; genome-based coverage is from gnomAD-Canada v1.0 (HostSeq) only.

Sources

Reasoning

No ClinGen CSPEC/VCEP specification exists for TSC2, so the generic ACMG/AMP 2015 framework (Richards et al., PMID:25741868) governs: PS3 requires well-established functional studies demonstrating a deleterious effect. No variant-specific well-established functional study of c.3796_3797del (p.(Leu1266AlafsTer55)) is present in the evidence bundle. A full-text search of every retrieved publication (PMID:10205261, 14561707, 14729330, 24529379, 25724664, 25741868) found zero mentions of c.3796_3797del, L1266, or Leu1266. OncoKB curates this exact variant as 'Likely Oncogenic' with biological effect 'Likely Loss-of-function' (variant-specific page), but this is a curated annotation inferred from the frameshift/truncating consequence in a somatic-cancer context, not an experimental assay with positive/negative controls; it does not satisfy the PS3 'well-established functional study' bar on its own. The OncoKB-cited functional papers establish gene-level TSC2 biology only (TSC2 is a GAP toward Rheb and a negative regulator of mTORC1; Tsc1/Tsc2 loss activates mTOR; TSC2 loss of function is common in HCC) and do not test this variant. PMID:10205261 (flagged as a PS3 candidate in the literature pass) provides tumor-suppressor/second-hit LOH context at gene level but does not mention this variant and contains no variant-specific functional assay. PS3 therefore not met at any strength; no functional evidence is contributed to the classification.

Rule

PS3 (Richards et al. 2015, PMID:25741868; generic ACMG/AMP fallback per generic_acmg_combination_rules): Well-established functional studies show a deleterious effect.

Evidence considered

• PMID:25741868 - ACMG/AMP 2015 PS3 rule text: 'Well-established functional studies show a deleterious effect'; assay validity judged by biological relevance, validation, reproducibility, and controls.
• generic_acmg_combination_rules - governing generic ACMG/AMP fallback rule file (no VCEP/CSPEC specification available for TSC2).
• oncokb - variant-specific curated annotation for TSC2 L1266Afs*55: Likely Oncogenic, biological effect 'Likely Loss-of-function'; curation/inference without assay controls; evaluated and found insufficient to meet PS3.
• PMID:10205261 - TSC2 tumor-suppressor function via somatic loss/LOH of the wild-type allele in hamartomas (150 TSC families); gene-level context only; variant not mentioned.
• PMID:14561707 - Tsc1/Tsc2 loss activates mTOR and disrupts PI3K-Akt signaling in MEFs and Tsc2+/- mouse tumors; gene-level context only; variant not mentioned.
• PMID:14729330 - TSC2 is a GAP toward Rheb (HEK293 in vitro GAP assays, GTP/GDP ratio readouts); gene-level context only; variant not mentioned.
• PMID:24529379 - TSC1-TSC2-TBC1D7 complex negatively regulates mTORC1 at the lysosome; gene-level context only; variant not mentioned.
• PMID:25724664 - TSC2 loss of function common in HCC (4/8 cell lines, 8/28 xenografts) with elevated mTOR signaling; gene-level context only; variant not mentioned.

Still missing

• No variant-specific functional assay for c.3796_3797del / p.(Leu1266AlafsTer55) (e.g., TSC2 GAP activity toward Rheb, mTORC1/S6K pathway readouts, or mRNA/protein expression or NMD assay) with appropriate validated controls was identified in any retrieved source.
• OncoKB 'Likely Loss-of-function' annotation is a curated inference, not a validated assay; a published, assay-controlled functional study of this exact variant would be required to meet PS3 (or PS3_moderate/PS3_supporting under SVI strength calibration).

Sources

Reasoning

PP3 requires multiple lines of computational evidence supporting a deleterious effect on the gene or gene product (ACMG/AMP 2015, PMID:25741868). For NM_000548.4:c.3796_3797del (TSC2, p.(Leu1266AlafsTer55), a 2-bp frameshift deletion in exon 31), the splice-impact path was assessed with SpliceAI: max delta score 0.01 (DS_AG 0.01, DS_AL 0.01, DS_DG 0.0, DS_DL 0.0; pangolin_SG 0.0, pangolin_SL -0.06) for a variant lying ~21 bp from the exon 31 donor splice site - far below the calibrated splice-impact threshold (delta >= 0.2, ClinGen SVI Splicing Subgroup, Walker LC et al., Am J Hum Genet 2023;110(7):1046-1067, PMID 37352859), so no splice-altering effect is predicted. The missense in-silico path (REVEL/BayesDel/aGVGD) is not applicable: the variant is a deletion, not an SNV, and the evidence pipeline explicitly skipped these predictors ('not snv or no coords'). No other calibrated computational evidence (e.g., conservation, CADD) is present in the evidence bundle. No TSC2 VCEP/CSPEC or gene-specific PP3/BP4 lookup spreadsheet exists to pre-assign a code (cspec found=false; vcep_materials.json empty; framework_mode=generic_acmg). The frameshift's protein-truncating consequence is a real annotated effect captured by PVS1 (assessed by the PVS1 group), not an in-silico prediction, and PP3 must not be stacked onto that same consequence. PP3 is therefore not met.

Rule

PP3: Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.) - ACMG/AMP 2015 (PMID:25741868); splice-impact line calibrated per ClinGen SVI Splicing Subgroup (PMID 37352859, SpliceAI delta >= 0.2).

Evidence considered

• spliceai — SpliceAI Lookup for NM_000548.4:c.3796_3797del returned max delta score 0.01 (DS_AG 0.01, DS_AL 0.01, DS_DG 0.0, DS_DL 0.0; pangolin_SG 0.0, pangolin_SL -0.06); no splice impact predicted
• spliceai — SpliceAI delta interpretation: delta >= 0.2 indicates predicted splice impact; the observed 0.01 is far below this threshold (ClinGen SVI Splicing Subgroup, Walker LC et al., Am J Hum Genet 2023;110(7):1046-1067, PMID 37352859)
• PMID:25741868 — PP3 rule: multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.), with the caveat that algorithms sharing input data are not independent and PP3 can be used only once per variant
• Case evidence bundle (prefetch.json/evidence.json) — REVEL and BayesDel explicitly skipped: 'not snv or no coords'; missense in-silico path not applicable to a 2-bp deletion frameshift

Still missing

• No TSC2 ClinGen VCEP/CSPEC PP3/BP4 calibration or gene-specific in-silico lookup spreadsheet exists; generic ACMG/AMP 2015 PP3/BP4 rules (PMID:25741868) with the ClinGen SVI Splicing Subgroup SpliceAI calibration (PMID 37352859) were applied instead.
• No conservation/evolutionary predictor data (e.g., phyloP/GERP/CADD) were captured in the evidence bundle; for a frameshift variant this path would not alter the PP3 determination, but it would be needed if PP3 were ever re-adjudicated for a missense SNV in TSC2.

Sources

Reasoning

PP5 is restricted, per governing instructions, to an exact-variant ClinVar expert-panel (3-star) classification of Pathogenic/Likely pathogenic. The ClinVar record for this variant (VCV003724267) has 3 submissions, all from ordinary clinical laboratories with 'criteria provided, single submitter' review status; expert_panel_submissions = 0 and review stars = 2 (multiple submitters, no conflicts). No expert-panel classification exists, so PP5 must not be applied. The aggregate Pathogenic/Likely pathogenic label from laboratory submissions does not qualify.

Rule

ACMG/AMP 2015 PP5: 'Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.' (PMID:25741868); per pipeline policy, triggered only by an exact-variant ClinVar expert-panel Pathogenic/Likely pathogenic classification

Sources

Reasoning

BA1 requires an allele frequency high enough to be inconsistent with pathogenicity (ACMG/AMP 2015: >5% in ExAC/1000 Genomes/ESP; local gnomAD non-VCEP framework: >1%). The variant NM_000548.4:c.3796_3797del (TSC2, p.(Leu1266AlafsTer55)) is absent from all queried population datasets: gnomAD v2.1 exomes (16-2131778-CCT-C, dataset=gnomad_r2_1), gnomAD v4.1 exomes (16-2081777-CCT-C, dataset=gnomad_r4), and gnomAD-Canada v1.0 genomes (HostSeq). Observed allele frequency is 0, far below any BA1 threshold, so BA1 is not met.

Rule

BA1: allele frequency >5% in ExAC/1000 Genomes/ESP (ACMG/AMP 2015, PMID:25741868); local gnomAD non-VCEP framework threshold BA1>1%. Absence (AF=0) cannot meet either threshold.

Evidence considered

• PMID:25741868 — BA1 rule (allele frequency >5% in population databases such as ExAC/1000 Genomes/ESP) defines the criterion; not satisfied by AF=0
• gnomad_v2 — variant absent from gnomAD v2.1 exomes (search_status=absent)
• gnomad_v4 — variant absent from gnomAD v4.1 exomes (search_status=absent)
• gnomad_canada — variant absent from gnomAD-Canada v1.0 genomes (search_status=absent)

Sources

Reasoning

BS1 requires an allele frequency greater than expected for the disorder (local gnomAD non-VCEP framework for dominant disorders: >0.3%). TSC is an early-onset autosomal dominant disorder (PMID:17304050); a truly pathogenic TSC2 allele is expected to be absent or extremely rare. The variant is absent from gnomAD v2.1 exomes, gnomAD v4.1 exomes, and gnomAD-Canada v1.0 genomes (AF=0), which is not greater than expected for TSC; BS1 is not met.

Rule

BS1: allele frequency greater than expected for disorder (ACMG/AMP 2015, PMID:25741868); local gnomAD non-VCEP framework threshold BS1>0.3% for dominant disorders. AF=0 is below the expected frequency for a pathogenic TSC2 allele.

Evidence considered

• PMID:25741868 — BS1 rule (allele frequency greater than expected for disorder) defines the criterion; not satisfied by AF=0
• PMID:17304050 — TSC is an early-onset autosomal dominant disorder (incidence ~1/6000-1/25000, ~two-thirds de novo); expected allele frequency for a pathogenic TSC2 variant is far below the BS1 threshold
• gnomad_v2 — variant absent from gnomAD v2.1 exomes (search_status=absent)
• gnomad_v4 — variant absent from gnomAD v4.1 exomes (search_status=absent)
• gnomad_canada — variant absent from gnomAD-Canada v1.0 genomes (search_status=absent)

Sources

Reasoning

BS2 requires observation of the variant in a healthy adult (heterozygous for a dominant disorder with full penetrance expected at an early age). No healthy-adult, homozygous, or hemizygous observations exist: the variant is absent from gnomAD v2.1 exomes, gnomAD v4.1 exomes, and gnomAD-Canada v1.0 genomes, and no such observations are reported in the case evidence. TSC is an early-onset, high-penetrance dominant disorder (PMID:17304050), so a truly benign variant would be expected to be observed in controls; the complete absence of population observations is the opposite of the BS2 evidence pattern. BS2 is not met.

Rule

BS2: observed in a healthy adult individual for a dominant disorder with full penetrance expected at an early age (ACMG/AMP 2015, PMID:25741868). No healthy adult observations reported.

Evidence considered

• PMID:25741868 — BS2 rule (observed in a healthy adult, full penetrance expected at an early age) defines the criterion; no such observation exists
• PMID:17304050 — TSC is an early-onset autosomal dominant disorder with substantial de novo fraction, the penetrance context under which BS2 healthy-adult observations would be evaluated
• gnomad_v2 — variant absent from gnomAD v2.1 exomes (search_status=absent)
• gnomad_v4 — variant absent from gnomAD v4.1 exomes (search_status=absent)
• gnomad_canada — variant absent from gnomAD-Canada v1.0 genomes (search_status=absent)

Still missing

• No healthy-adult or homozygote observations are possible to evaluate because the variant is absent from every queried population dataset; if a population observation is ever reported, BS2 should be re-adjudicated.

Sources

Reasoning

Generic ACMG/AMP 2015 (Richards et al., PMID:25741868) governs: BS3 requires well-established functional studies showing no damaging effect on protein function or splicing. No functional study demonstrating normal function of c.3796_3797del (p.(Leu1266AlafsTer55)) exists in the evidence bundle. SpliceAI predicts no significant splice impact (max delta = 0.01; all delta scores <= 0.01), but this is an in silico prediction that addresses splicing only and does not establish normal protein function for a frameshift variant whose primary consequence is a premature stop (p.Leu1266AlafsTer55, NMD-eligible in principle); it cannot substitute for a well-established functional study. No retrieved publication mentions this variant, so no paper-based functional evidence of a benign effect exists. BS3 therefore not met; the variant contributes no benign functional evidence.

Rule

BS3 (Richards et al. 2015, PMID:25741868; generic ACMG/AMP fallback per generic_acmg_combination_rules): Well-established functional studies show no damaging effect on protein function or splicing.

Evidence considered

• PMID:25741868 - ACMG/AMP 2015 BS3 rule text: 'Well-established functional studies show no damaging effect on protein function or splicing'.
• generic_acmg_combination_rules - governing generic ACMG/AMP fallback rule file (no VCEP/CSPEC specification available for TSC2).
• spliceai - SpliceAI max delta score 0.01 (DS_AG 0.01, DS_AL 0.01, DS_DG 0.0, DS_DL 0.0): no predicted splice impact; in silico only and does not demonstrate normal protein function; insufficient for BS3.

Still missing

• No well-established functional study showing normal function (no damaging effect on protein function or splicing) for c.3796_3797del / p.(Leu1266AlafsTer55).
• SpliceAI low delta scores (max 0.01) exclude a predicted splice effect but are in silico and irrelevant to the protein-truncating mechanism; an RNA/protein-level assay would be needed to support any BS3 consideration.

Sources

Reasoning

BP4 requires multiple lines of computational evidence suggesting no impact on the gene or gene product (ACMG/AMP 2015, PMID:25741868). The only computational line available for this variant is the splice-impact path: SpliceAI max delta score 0.01, predicting no splice-altering effect (threshold delta >= 0.2, PMID 37352859). BP4 is nevertheless not met because: (1) this is a single line of splice-only evidence, and the same prediction must not be double-counted across PP3/BP4/BP7; (2) a low splice score addresses splicing only and cannot establish 'no impact on the gene or gene product' for a variant whose annotated consequence is a frameshift, p.(Leu1266AlafsTer55), which truncates the TSC2 protein at residue 1266 after 55 novel amino acids - an unambiguous protein-level impact; (3) the missense in-silico path (REVEL/BayesDel/aGVGD) is not applicable to a non-SNV deletion and was skipped ('not snv or no coords'). Under the ACMG/AMP framework, BP4 is intended for variants such as synonymous changes where computational evidence supports no effect; applying it to a protein-truncating variant would contradict the established consequence. No TSC2 VCEP/CSPEC exists to pre-assign BP4. BP4 is therefore not met.

Rule

BP4: Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc.) - ACMG/AMP 2015 (PMID:25741868); caveat: algorithms sharing input cannot each count as independent lines. Single splice-only line plus known frameshift consequence preclude BP4.

Evidence considered

• spliceai — SpliceAI Lookup for NM_000548.4:c.3796_3797del returned max delta score 0.01 (DS_AG 0.01, DS_AL 0.01, DS_DG 0.0, DS_DL 0.0), a single splice-only line suggesting no splice impact
• spliceai — SpliceAI delta interpretation: delta >= 0.2 indicates predicted splice impact; the observed 0.01 is far below this threshold (ClinGen SVI Splicing Subgroup, Walker LC et al., Am J Hum Genet 2023;110(7):1046-1067, PMID 37352859)
• PMID:25741868 — BP4 rule: multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc.), with the caveat that algorithms sharing input data are not independent and BP4 can be used only once per variant
• Case evidence bundle (prefetch.json normalization) — variant consequence is frameshift NP_000539.2:p.(Leu1266AlafsTer55); the established protein-truncating effect is inconsistent with a 'no impact on gene product' claim; REVEL/BayesDel skipped ('not snv or no coords')

Still missing

• No TSC2 VCEP/CSPEC BP4 calibration or gene-specific in-silico lookup spreadsheet exists; generic ACMG/AMP 2015 rules applied.
• No RNA/splicing functional assay data (e.g., RT-PCR/minigene) are available for this variant; only computational splice prediction (SpliceAI) exists. If RNA evidence becomes available, splice-based PP3/BP4/BS3 adjudication should be revisited (BS3/PS3 functional evidence is owned by another group).

Sources

Reasoning

BP6 is restricted, per governing instructions, to an exact-variant ClinVar expert-panel (3-star) classification of Benign/Likely benign. The ClinVar record for this variant contains no benign or likely-benign submissions at all (2 Pathogenic, 1 Likely pathogenic, all ordinary clinical laboratories; expert_panel_submissions = 0). No qualifying expert-panel benign classification exists, so BP6 must not be applied.

Rule

ACMG/AMP 2015 BP6: 'Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.' (PMID:25741868); per pipeline policy, triggered only by an exact-variant ClinVar expert-panel Benign/Likely benign classification

Sources

Reasoning

PS2 requires a de novo occurrence of the variant in a proband with the disease, with both parents confirmed non-carriers and maternity/paternity confirmed. No de novo occurrence data for NM_000548.4:c.3796_3797del exists in this case: the ClinVar submission audit for VCV003724267 (3 single-submitter clinical testing records, germline origin) contains no criterion data (submission_audit: usable_for_criterion_leads = 0, criterion_leads = []), and the exact variant is not mentioned in any retrieved full-text publication (PMID:10205261, PMID:17304050). Population-level TSC de novo rates (approximately two-thirds of cases sporadic) are disease-level background and cannot establish de novo status of this specific variant. Evidence insufficient; parental testing results for the proband are the missing gap.

Rule

Generic ACMG/AMP 2015 (Richards et al., PMID 25741868): PS2 = de novo (maternity and paternity confirmed) in a patient with the disease and no family history.

Still missing

• No parental genotype data for the proband: both parents tested and non-carriers with confirmed maternity/paternity is not reported anywhere in the evidence bundle for c.3796_3797del.

Sources

Reasoning

PS4 requires the prevalence of the variant in affected individuals to be significantly increased versus controls (ACMG/AMP 2015), commonly operationalized for rare, highly penetrant dominant disorders as multiple unrelated probands carrying the exact variant with the same phenotype. No such case-level data exist for c.3796_3797del in this case bundle. Both candidate case-observation papers (PMID:10205261, 150 unrelated TSC families; PMID:17304050, 325 individuals with definite TSC) were reviewed in full text and do not report this exact variant (no c.3796_3797del, p.Leu1266Afs*55, or L1266 notation anywhere; their listed TSC2 deletions are at other positions, e.g. 4559del4bp, 5256del18bp, c.5238_5255del). ClinVar shows 2 Pathogenic + 1 Likely pathogenic clinical-laboratory submissions, but the submission audit marks all 3 as record_not_exact_match and 0 usable for criterion leads, 0 expert-panel submissions; per governing rules, aggregate ClinVar P/LP labels are final classifications and cannot be used to infer PS4 case enrichment. gnomAD absence (v2.1/v4.1/Canada) is population-frequency evidence relevant to PM2/BS1, not PS4. Case-control enrichment cannot be quantified for this exact variant.

Rule

ACMG/AMP 2015 PS4: 'The prevalence of the variant in affected individuals is significantly increased compared with the prevalence in controls.' (PMID:25741868)

Still missing

• No published or auditable case-level observations (affected proband counts) of the exact variant c.3796_3797del in TSC cohorts
• No case-control enrichment data (variant frequency in affecteds vs controls) available
• ClinVar submissions flagged record_not_exact_match (NM_000548.5 vs NM_000548.4) and provide no usable per-submission case counts

Sources

Reasoning

PM6 requires an assumed de novo occurrence without confirmation of maternity/paternity. Even an unconfirmed de novo status for c.3796_3797del is not reported: no variant-specific case report, ClinVar criterion data, or literature source documents that the proband's parents did not carry the variant. The reviewed cohort studies (PMID:10205261, PMID:17304050) report only aggregate de novo/sporadic proportions for TSC and do not mention this variant. Evidence insufficient; the gap is any parental testing observation (even without parentage confirmation) for this variant.

Rule

Generic ACMG/AMP 2015 (Richards et al., PMID 25741868): PM6 = assumed de novo, but without confirmation of paternity and maternity.

Still missing

• No assumed-de-novo observation (parents not tested for the variant, or tested without parentage confirmation) is available for c.3796_3797del in any reviewed source.

Sources

Reasoning

PP1 requires co-segregation of the variant with disease in multiple affected family members in a gene definitively known to cause the disease (TSC2 is definitively established as causing tuberous sclerosis complex). No segregation data exist for c.3796_3797del: no multiplex family carrying this exact variant is reported in the ClinVar submissions (no criterion data; records are single-submitter clinical testing) or in the retrieved literature (variant absent from PMID:10205261 and PMID:17304050 full texts; those papers provide TSC1/TSC2 cohort mutation frequencies, not family segregation for this variant). Evidence insufficient; the gap is any affected-family-member co-segregation count (meioses) for this variant.

Rule

Generic ACMG/AMP 2015 (Richards et al., PMID 25741868): PP1 = co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.

Still missing

• No co-segregation data: number of affected family members carrying c.3796_3797del (informative meioses) is not reported in any reviewed source.

Sources

Reasoning

PP4 requires the proband's phenotype or family history to be highly specific for a disease with a single genetic etiology. The case bundle contains no proband phenotype, clinical history, or family history data. The ClinVar record lists aggregate condition labels (Tuberous sclerosis 2, Lymphangiomyomatosis, Isolated focal cortical dysplasia type II, not provided) attached to laboratory submissions; these are not a documented phenotype for a specific carrier of this variant and, per the no-inference rule, cannot be used to establish phenotype specificity. Although the TSC phenotype is generally highly specific, PP4 cannot be applied without carrier phenotype documentation.

Rule

ACMG/AMP 2015 PP4: 'Phenotype or family history is highly specific for a disease with a single genetic etiology.' (PMID:25741868)

Still missing

• No proband phenotype/family history data present in the case bundle
• No documentation that the exact-variant carrier(s) meet TSC clinical diagnostic criteria

Sources

Reasoning

BS4 requires demonstrated lack of segregation, i.e., the variant present in an affected individual without the disease phenotype or absent in affected family members within a segregation study. No segregation (or non-segregation) analyses of c.3796_3797del exist in this case: no variant-specific family data in ClinVar (criterion_leads = []) and the exact variant is not reported in the retrieved literature (PMID:10205261, PMID:17304050). Absence of segregation data is not evidence of non-segregation, so BS4 cannot be applied. Evidence insufficient; the gap is a segregation study showing the variant in phenotypically unaffected carriers or absent in affected relatives.

Rule

Generic ACMG/AMP 2015 (Richards et al., PMID 25741868): BS4 = lack of segregation in affected family members.

Still missing

• No non-segregation observations: no family study reporting c.3796_3797del in an unaffected carrier or its absence in an affected family member.

Sources

Reasoning

BP2 (ACMG/AMP 2015) applies when the variant is 'observed in trans with a pathogenic variant for a fully dominant disease/gene; or observed in cis with a pathogenic variant in any inheritance pattern.' TSC2-related TSC is a fully dominant disorder, so the trans clause is applicable in principle. However, the case bundle contains no phased observations whatsoever: there is no proband genotype, no parental/family cis-trans testing, no segregation data, and none of the three ClinVar germline submissions (all single-submitter clinical-testing records; 0 usable for criterion leads) reports any phase or co-occurrence information. None of the available publications (PMID:10205261, PMID:17304050) mentions this exact variant or provides cis/trans data for it. With no evidence that c.3796_3797del was observed in trans with a pathogenic TSC2 variant (or in cis with a pathogenic variant), BP2 cannot be met or refuted on the available evidence.

Rule

Observed in trans with a pathogenic variant for a fully dominant disease/gene; or observed in cis with a pathogenic variant in any inheritance pattern (ACMG/AMP 2015, PMID 25741868).

Evidence considered

• ACMG/AMP 2015 (PMID 25741868) defines BP2 as observed in trans with a pathogenic variant for a fully dominant disease/gene, or observed in cis with a pathogenic variant in any inheritance pattern
• clinvar: 3 germline submissions (Sinai Health System, GeneDx, Labcorp) provide classifications only; no cis/trans or phase information reported; 0 submissions usable for criterion leads
• No proband/family genotype, phase, or segregation data found anywhere in the case evidence bundle
• PMID:17304050 confirms TSC is autosomal dominant (relevant to the BP2 trans clause context) but does not mention this variant

Still missing

• No phased (cis/trans) observation of c.3796_3797del with any other TSC2 variant is available
• No proband or family genotype/segregation data is present in the case bundle
• ClinVar submissions (3 germline clinical-testing records) contain no phase or co-occurrence information (0 usable criterion-lead submissions)

Sources

Reasoning

BP5 requires documentation that the variant was found in a case with an alternate molecular basis of disease. The case bundle contains no proband-level data (no phenotype, no prior genetic testing results, no evidence of an alternate etiology), so the criterion cannot be evaluated. Absence of proband workup data is a data gap, not evidence of an alternate molecular diagnosis.

Rule

ACMG/AMP 2015 BP5: 'Variant found in a case with an alternate molecular basis for disease.' (PMID:25741868)

Still missing

• No proband clinical workup or alternate-molecular-basis documentation available in the case bundle