TSC2 c.1292C>T, p.(Ala431Val)

Reasoning

PM2 is met at supporting strength. The variant is present but extremely rare in population databases: gnomAD v4.1 AF 0.000165598 (0.01656%; 267/1,612,342 alleles), gnomAD v2.1 AF 0.000380318 (0.03803%; 106/278,714 alleles), and gnomAD-Canada AF 0.000108578 (0.01086%; 2/18,420 alleles). No homozygotes are reported. These frequencies are below the generic rare-variant threshold of 0.1% and are compatible with PM2 supporting evidence for a rare Mendelian disease allele.

Rule

Generic PM2 threshold: absent or extremely rare in population controls; operational threshold applied here is AF <0.1% for supporting evidence.

Evidence considered

• gnomAD v4.1 reports 267/1,612,342 alleles, AF 0.000165598 (0.01656%), and 0 homozygotes; grpmax FAF is 0.00016638.
• gnomAD v2.1 reports 106/278,714 alleles, AF 0.000380318 (0.03803%), and 0 homozygotes; grpmax FAF is 0.0013996.
• gnomAD-Canada reports 2/18,420 alleles, AF 0.000108578 (0.01086%), and 0 homozygotes.
• The generic population operating threshold applied here is PM2 supporting when AF is below 0.1%; no applicable TSC2-specific VCEP specification was found.
• Richards et al. describes population frequency as evidence used in sequence-variant interpretation and is the governing generic ACMG/AMP publication for this case (PMID:25741868).

Sources

Reasoning

No VCEP-specific PP3/BP4 lookup spreadsheet is available for TSC2, so the generic ACMG in silico approach was applied. SpliceAI predicts no splicing impact for NM_000548.5:c.1292C>T: all four delta scores (acceptor gain 0.004, acceptor loss 0.005, donor gain 0.0, donor loss 0.0) are near zero, with a maximum delta score of 0.005, far below thresholds used to flag a splice-altering variant, indicating computational prediction of no effect on splicing. This single, clearly negative and well-calibrated splice-impact result is sufficient to apply BP4 at supporting strength for the splice-impact sub-path. The missense meta-predictor REVEL score (0.386) falls in an indeterminate zone between the calibrated BP4-supporting cutoff (<=0.290) and PP3-supporting cutoff (>=0.644) per ClinGen SVI/Pejaver et al. calibration (PMID 36413997), so it does not independently add to or detract from the BP4 call; the missense in silico line remains inconclusive rather than benign-supporting. BayesDel (0.284738) is not used because no verified published calibration threshold for BayesDel is available to this pipeline. This variant is a missense change (p.Ala431Val), not synonymous, so BP7 does not apply.

Rule

generic_acmg_2015_pp3_bp4

Evidence considered

• SpliceAI (spliceailookup.broadinstitute.org, NM_000548.5:c.1292C>T, hg19): DS_AG=0.004, DS_AL=0.005, DS_DG=0.0, DS_DL=0.0, max delta score=0.005 -- evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.01).' Consistent with computational prediction of no deleterious effect on splicing, supporting BP4.
• REVEL v1.3 local lookup score = 0.386, in the indeterminate zone between calibrated BP4 (<=0.290) and PP3 (>=0.644) thresholds (Pejaver et al. 2022, PMID 36413997); not used to add weight to BP4.
• BayesDel local lookup score = 0.284738 retrieved but not used for BP4 because no verified published calibration/threshold for BayesDel is available to this pipeline.
• Variant is a missense substitution (NP_000539.2:p.(Ala431Val)), not synonymous, so BP7 is not applicable.
• No gene-specific VCEP PP3/BP4 lookup table exists for TSC2 in this case (vcep_materials.json: cspec not found, files_found empty).

Still missing

• The missense meta-predictor line (REVEL) is indeterminate and does not independently corroborate BP4; the BP4 call rests solely on the SpliceAI splice-impact result.
• No named published SpliceAI delta-score threshold citation is available in this bundle for a numeric cutoff beyond the qualitative near-zero/no-impact interpretation already reported by the tool itself.

Sources

Reasoning

No VCEP-specific PP3/BP4 lookup spreadsheet is available for TSC2 (cspec not found; vcep_materials.json empty); generic ACMG in silico approach applies. Two independent computational lines were evaluated: (1) SpliceAI predicts no splicing impact (DS_AG=0.004, DS_AL=0.005, DS_DG=0.0, DS_DL=0.0; max delta=0.005), far below any threshold used to flag a deleterious splicing effect, so the splice-impact path does not support pathogenicity. (2) The missense meta-predictor REVEL score is 0.386, which falls in the indeterminate zone between the ClinGen SVI (Pejaver et al., PMID 36413997) calibrated BP4-supporting cutoff (REVEL <=0.290) and PP3-supporting cutoff (REVEL >=0.644); it therefore does not reach the PP3 threshold. BayesDel score (0.284738) is available but this pipeline has no verified published calibration/threshold for BayesDel, so it is treated as not_available for PP3 per policy and not used to trigger the criterion. No aGVGD result is present in evidence. Overall, no in silico line of evidence reaches a PP3-supporting (or higher) threshold for this variant, whether via the splicing or missense path.

Rule

generic_acmg_2015_pp3_bp4

Evidence considered

• SpliceAI (spliceailookup.broadinstitute.org, NM_000548.5:c.1292C>T): DS_AG=0.004, DS_AL=0.005, DS_DG=0.0, DS_DL=0.0, max delta score=0.005 -- no significant splice impact predicted.
• REVEL v1.3 local lookup score = 0.386 for 16-2062531-C-T, which lies between the ClinGen SVI-calibrated BP4-supporting cutoff (<=0.290) and PP3-supporting cutoff (>=0.644) (Pejaver et al. 2022, PMID 36413997), so it does not meet the PP3 threshold.
• BayesDel local lookup score = 0.284738 was retrieved but is not used for PP3 because no verified published calibration/threshold for BayesDel is available to this pipeline.
• No gene-specific VCEP PP3/BP4 lookup table exists for TSC2 in this case (vcep_materials.json: cspec not found, files_found empty).

Still missing

• REVEL score (0.386) falls in an indeterminate zone between calibrated BP4 and PP3 thresholds and cannot independently support PP3.
• BayesDel score is present but has no verified published calibration threshold accessible to this pipeline, so it cannot be used for PP3.
• No aGVGD score is available in the evidence bundle.

Sources

Reasoning

No exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic assertion is present.

Rule

Operational rule: apply PP5 only to an exact-variant ClinVar expert-panel Pathogenic/Likely pathogenic classification; ordinary laboratory submissions and aggregate labels are excluded.

Evidence considered

• ClinVar variation 207708 is an exact match for NM_000548.5:c.1292C>T (p.Ala431Val); the submission audit reports zero expert-panel submissions and the submitted classifications are benign or likely benign.

Sources

Reasoning

BA1 is not met. The highest reliable population frequency is below the generic benign-alone threshold of 1%: gnomAD v4.1 overall AF is 0.000165598 (0.01656%), the highest reported broad/subpopulation AF is 0.000563486 (0.05635%) in Finnish individuals, and gnomAD-Canada AF is 0.000108578 (0.01086%).

Rule

Generic BA1 threshold: population frequency >1% for benign-alone evidence; no applicable TSC2-specific VCEP specification was found.

Evidence considered

• gnomAD v4.1 reports 267/1,612,342 alleles, AF 0.000165598, with 0 homozygotes; the highest displayed population AF is Finnish 36/63,888 alleles, AF 0.000563486.
• gnomAD v2.1 reports overall AF 0.000380318 and grpmax FAF 0.0013996; no homozygotes were observed.
• gnomAD-Canada reports 2/18,420 alleles, AF 0.000108578, with 0 homozygotes.
• The generic ACMG/AMP framework treats BA1 as a benign-alone criterion, but these observed frequencies are far below 1%.

Sources

Reasoning

BS1 is not met. The large gnomAD v4.1 dataset shows an overall AF of 0.000165598 and maximum displayed population AF of 0.000563486, both below the generic >0.3% frequency threshold used for BS1. Although gnomAD v2 contains a small NFE_EST subgroup with AF 0.005816, the corresponding v2 grpmax FAF is 0.0013996 and the substantially larger v4 dataset does not reproduce a frequency above 0.3%, supporting a small-subgroup/threshold artifact rather than BS1 evidence.

Rule

Generic BS1 threshold: population frequency >0.3%; evaluate robust maximum population frequency and account for small-subgroup artifacts.

Evidence considered

• gnomAD v4.1 reports overall AF 0.000165598 and Finnish AF 0.000563486; both are below 0.003.
• gnomAD v2.1 reports grpmax FAF 0.0013996, below 0.003, despite a small NFE_EST subgroup AF of 0.005816 (28/4,814 alleles).
• gnomAD-Canada reports AF 0.000108578, also below 0.003.
• The generic population operating threshold for BS1 is AF >0.3%; no applicable TSC2-specific VCEP specification was found.

Sources

Reasoning

No exact-variant ClinVar expert-panel Benign or Likely benign assertion is present.

Rule

Operational rule: apply BP6 only to an exact-variant ClinVar expert-panel Benign/Likely benign classification; ordinary laboratory submissions and aggregate labels are excluded.

Evidence considered

• ClinVar variation 207708 is an exact match for NM_000548.5:c.1292C>T (p.Ala431Val); the submission audit reports zero expert-panel submissions. Its benign and likely benign labels are ordinary laboratory submissions and are excluded from BP6 under the requested rule.

Sources

Reasoning

No ClinGen VCEP specification exists for TSC2 (cspec.found=false; framework_mode=generic_acmg). PS1 requires identification of a different nucleotide change at the same codon that produces the identical amino acid change (p.Ala431Val) already established as pathogenic. No such comparator variant is present anywhere in the case bundle (pm5_candidates.json returned found=false with same-residue harvesting explicitly skipped, and no other evidence artifact lists alternate codon changes at residue 431).

Still missing

• No same-amino-acid-change (different codon) comparator variant with an established pathogenic classification was identified in the bundle to evaluate PS1.

Reasoning

No proband-level parental testing or confirmed de novo observation is documented for NM_000548.5:c.1292C>T (NP_000539.2:p.(A431V)). The variant is mentioned in a published autism-spectrum-disorder comparison table, but that source provides no parent genotypes, maternity/paternity confirmation, or de novo evidence. PS2 therefore cannot be assessed.

Rule

Generic ACMG/AMP PS2: de novo occurrence in a patient with a confirmed phenotype consistent with the disease, with confirmed maternity and paternity; apply only when parental testing and relevant clinical information are documented.

Evidence considered

• PMID:23514105 lists TSC2 p.A431V as an ASD variant in an inter-study comparison table, but reports no parent testing, de novo status, proband-level phenotype, or family details for this variant.
• The generic ACMG/AMP framework requires a documented de novo occurrence, with appropriate parental and maternity/paternity confirmation, for PS2; no such case evidence is present in this bundle.

Still missing

• Proband clinical phenotype and diagnostic consistency with TSC2-related disorder
• Validated parental genotypes for both biological parents
• Confirmation that the variant is absent in both parents using adequate-depth testing
• Confirmation of maternity and paternity and assessment for parental mosaicism

Sources

Reasoning

No TSC2 VCEP/gene-specific functional assay framework is available for this case (vcep_materials.json reports no cspec found; framework_mode is generic_acmg_fallback). No source in the bundle reports a validated functional assay (e.g., GAP-domain/Rheb-GTPase activity, mTOR pathway rescue, splicing assay) for NM_000548.5:c.1292C>T (p.Ala431Val). The retrieved literature (PMID:22558107, PMID:23514105, PMID:25862857 and others) consists of case-cohort resequencing/association studies and general ACMG/guideline papers; PMID:23514105 lists p.A431V only as an 'ASD variant' citation in a comparator table with no functional, biochemical, or splicing assay data. No assay evidence, controls, or biological-relevance data exist in the bundle to support PS3.

Still missing

• No functional/biochemical assay (e.g., TSC2 GAP-domain activity, Rheb-GTPase/mTOR pathway assay, splicing minigene) has been performed or reported for this variant in any source in this bundle.
• No TSC2-specific VCEP functional criteria specification is available to calibrate assay evidence even if one were found.

Sources

Reasoning

No exact-variant case-control enrichment estimate is available. The exact variant is listed as an ASD variant in an inter-study table, but the supplied evidence gives no carrier count, control count, or association statistic for p.A431V. A separate TSC2 gene-level autism enrichment study did not report this variant and cannot establish PS4 for it.

Rule

Generic ACMG/AMP PS4: prevalence of the exact variant in affected individuals must be significantly increased over controls; no TSC2-specific VCEP specification was retrieved.

Evidence considered

• PMID:23514105 Table 4 reports "TSC2 | A431V | - | - | ASD variant | -" and attributes the ASD-variant entry to Kelleher et al.; it does not provide p.A431V case/control counts, an odds ratio, or a P value.
• PMID:22558107 sequenced 290 idiopathic autism cases and 300 controls and reported a gene-level excess of rare TSC2 variants, but its extracted full text specifically does not mention c.1292C>T/p.A431V; gene-level enrichment is not evidence of enrichment of this allele.

Still missing

• An exact-variant case-control dataset with carrier counts and an appropriate control comparison is required.

Sources

Reasoning

TSC2 has no ClinGen VCEP specification, and no mutational hotspot, critical functional domain boundary, or missense-constraint (e.g., UniProt/Pfam feature, GAP-domain coordinates, gnomAD missense Z-score) data for residue 431 was retrieved in this case bundle. TSC2 is a large tumor-suppressor gene whose pathogenic variant spectrum is predominantly loss-of-function/truncating rather than clustered missense hotspots, but no source in the bundle characterizes residue 431 with respect to any specific functional domain, so no determination of domain/hotspot membership can be made.

Still missing

• No domain/hotspot annotation (e.g., UniProt feature table, Pfam/InterPro domain boundaries, or established TSC2 mutational hotspot map) covering residue 431 was retrieved.

Reasoning

PM5 requires a different missense change at the same residue (431) previously established as pathogenic. The dedicated pm5_candidates.json artifact explicitly reports found=false and eligible_for_classic_pm5_search=false, stating same-residue PM5 candidate harvesting was skipped ('Unable to confirm classic same-residue PM5 semantics safely; skipping automatic candidate harvesting'), and the recommendation level is not_applicable pending manual review. No alternate amino-acid substitution at codon 431 with an established pathogenic classification appears elsewhere in the bundle.

Evidence considered

• pm5_candidates.json: found=false, eligible_for_classic_pm5_search=false, recommendation.level=not_applicable, reason='Unable to confirm classic same-residue PM5 semantics safely; skipping automatic candidate harvesting.'

Still missing

• pm5_candidates.json harvesting was skipped for this variant; no verified alternate pathogenic missense change at residue 431 (Ala431) was located to evaluate PM5.

Reasoning

No unconfirmed de novo observation is documented for NM_000548.5:c.1292C>T (NP_000539.2:p.(A431V)). The only variant-level publication found lists p.A431V in a comparison table without parental genotypes or a statement that it arose de novo. PM6 therefore cannot be assessed.

Rule

Generic ACMG/AMP PM6: assumed de novo occurrence in a patient with a phenotype consistent with the disease when parental testing or confirmation is incomplete; do not apply without a documented case-level de novo suspicion.

Evidence considered

• PMID:23514105 explicitly provides no functional assay, segregation data, or additional case-level details for p.A431V; its table entry does not establish de novo origin.
• The generic ACMG/AMP framework uses PM6 for a suspected but unconfirmed de novo occurrence; the bundle contains no such observation for this variant.

Still missing

• A case report or cohort record identifying the proband carrying the variant
• Parental testing information, even if incomplete or not formally confirmed
• Clinical phenotype consistent with TSC2-related disorder

Sources

Reasoning

No informative affected or unaffected relatives, meioses, phase information, or segregation counts are reported for NM_000548.5:c.1292C>T (NP_000539.2:p.(A431V)). The variant-level publication only lists p.A431V in an inter-study comparison table and provides no family segregation data. PP1 cannot be assessed.

Rule

Generic ACMG/AMP PP1: cosegregation with disease in multiple affected family members; strength depends on the number and informativeness of meioses and the applicable disease model.

Evidence considered

• PMID:23514105 lists p.A431V as an ASD variant but states no additional case-level details or segregation data for it.
• No family pedigree, relative genotypes, phenotype assessments, or meioses are included in the consolidated case-evidence bundle.

Still missing

• Number and relationship of tested relatives
• Phenotype status of each relative
• Variant status in affected and unaffected relatives
• Informative meioses and any segregation likelihood or statistical calculation

Sources

Reasoning

PP2 requires that missense variants be a common mechanism of disease in the gene and that the gene have a low benign missense variation rate. No missense-constraint metric (e.g., gnomAD missense Z-score, ExAC/gnomAD constraint o/e ratio) for TSC2 was retrieved in this bundle, and gene_context.json/oncokb background describe TSC2's disease mechanism as tumor-suppressor loss-of-function, with somatic cancer variants noted to 'predominantly present as truncating loss-of-function mutations' rather than as a gene where missense change itself is the established common disease mechanism. Without a retrieved missense-constraint value or an established statement that missense is THE common mechanism for the germline TSC2 phenotype (tuberous sclerosis), PP2 cannot be confidently assessed either way.

Evidence considered

• gene_context.json: OncoKB background notes somatic TSC2 cancer mutations 'predominantly present as truncating loss-of-function mutations,' suggesting LOF, not missense, is the dominant reported mechanism, though this describes somatic cancer variants rather than germline TSC-causing variants specifically.

Still missing

• No gnomAD/ExAC missense constraint (Z-score or o/e ratio) for TSC2 was retrieved.
• No explicit statement in the bundle establishing missense variation as the predominant/common disease mechanism for TSC2 germline (tuberous sclerosis) phenotypes, as opposed to loss-of-function/truncating variants.

Reasoning

No phenotype for the proband or exact-variant carriers is supplied that is documented as highly specific for a TSC2-related disorder. The sole exact-variant literature entry labels p.A431V an ASD variant without clinical detail; ASD is not sufficient here to establish a highly specific phenotype.

Rule

Generic ACMG/AMP PP4: the patient's phenotype or family history must be highly specific for a disease with a single genetic etiology; no TSC2-specific VCEP specification was retrieved.

Evidence considered

• PMID:23514105 Table 4 contains only the entry "TSC2 | A431V | - | - | ASD variant | -" and reports no phenotype details for this variant.

Still missing

• Detailed proband phenotype and diagnostic evaluation, including whether clinical tuberous sclerosis complex criteria are met, are required.
• Exact-variant carrier phenotype data with sufficient specificity for a TSC2-related disorder are required.

Sources

Reasoning

BS2 cannot be established from the available population evidence. gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada each report zero homozygotes, but absence of homozygotes is not evidence that the variant is observed in healthy adults in the homozygous state. No sufficiently documented healthy-adult homozygote or hemizygote observation was identified.

Rule

Generic BS2 requires observation of the variant in a healthy adult in the homozygous or hemizygous state; zero homozygotes alone does not satisfy the criterion.

Evidence considered

• gnomAD v2.1 reports 0 homozygotes.
• gnomAD v4.1 reports 0 homozygotes.
• gnomAD-Canada reports 0 homozygotes.
• No source in the case bundle documents a healthy adult homozygote or hemizygote for this variant.

Still missing

• No documented healthy adult homozygote or hemizygote for NM_000548.5:c.1292C>T was identified.
• Population datasets provide homozygote counts but do not establish the clinical health status and adult ascertainment required for BS2.

Sources

Reasoning

No TSC2 VCEP/gene-specific functional assay framework is available for this case (vcep_materials.json reports no cspec found; framework_mode is generic_acmg_fallback). No source in the bundle reports a validated functional assay showing no damaging effect (e.g., normal GAP-domain activity, normal mTOR pathway signaling, normal splicing) for NM_000548.5:c.1292C>T (p.Ala431Val). The retrieved literature contains no functional/biochemical/splicing assay data for this variant, only case-cohort resequencing and general guideline/policy papers.

Still missing

• No functional/biochemical assay demonstrating a benign/normal effect has been performed or reported for this variant in any source in this bundle.
• No TSC2-specific VCEP functional criteria specification is available to calibrate assay evidence even if one were found.

Sources

Reasoning

No documented unaffected relatives carrying or lacking the variant, and no family-based non-segregation analysis, is available for NM_000548.5:c.1292C>T (NP_000539.2:p.(A431V)). The available variant-level publication does not report relative phenotypes or genotypes. BS4 cannot be assessed.

Rule

Generic ACMG/AMP BS4: lack of segregation in affected family members or presence of the variant in unaffected relatives when the phenotype is expected to be penetrant and the relatives are informative; apply only with adequate family and phenotype data.

Evidence considered

• PMID:23514105 provides no family-level phenotype or genotype information for p.A431V and therefore does not establish non-segregation.
• No informative unaffected-relative observations or other non-segregation evidence are present in the consolidated case-evidence bundle.

Still missing

• A well-characterized family with informative unaffected relatives
• Genotypes of unaffected relatives
• Reliable phenotype assessment and age-related penetrance considerations
• Evidence that the family structure and disease model make non-segregation informative

Sources

Reasoning

BP1 (missense variant in a gene for which primarily truncating/loss-of-function variants are known to cause disease) is the mechanistic converse of PP2 and requires the same missense-constraint/mechanism data that was not retrieved for TSC2 in this bundle. No ClinGen VCEP specification, missense constraint metric, or explicit mechanism statement adequate to determine whether missense variation is or is not an established disease mechanism for TSC2 germline phenotypes was found.

Still missing

• Same gap as PP2: no gnomAD/ExAC missense constraint metric or explicit TSC2 germline disease-mechanism statement (missense vs. truncating-predominant) was retrieved to support or refute BP1.

Reasoning

BP2 cannot be applied because the available evidence does not establish that NM_000548.5:c.1292C>T (p.Ala431Val) is in trans with a pathogenic variant in an affected individual with a fully penetrant dominant disorder, or in cis with a pathogenic variant. The exact variant is listed as an ASD variant in one paper, but no phase, second-variant pathogenicity, or affected-proband genotype details are provided.

Rule

Generic ACMG/AMP BP2: observed in trans with a pathogenic variant for a fully penetrant dominant disorder, or observed in cis with a pathogenic variant in any inheritance context.

Evidence considered

• PMID:23514105 lists TSC2 p.A431V in Table 4 as an ASD variant attributed to a prior study, but reports no phase information, no pathogenic second allele, and no additional affected-proband genotype details for p.A431V.
• The generic ACMG/AMP BP2 concept requires observation in trans with a pathogenic variant for a fully penetrant dominant disorder or in cis with a pathogenic variant; the bundle contains no such observation for this variant.

Still missing

• No family or proband-level genotype data establishing cis/trans phase.
• No documented pathogenic second allele and no affected-proband observation meeting the BP2 requirement.

Sources

Reasoning

No case-level evidence identifies an independent molecular cause that fully explains the affected individual's phenotype in a carrier of this variant.

Rule

Generic ACMG/AMP BP5: a variant is found in a case with an alternate molecular basis for disease; no TSC2-specific VCEP specification was retrieved.

Evidence considered

• The consolidated case evidence contains no proband-level alternative molecular diagnosis or co-occurrence evidence for c.1292C>T/p.A431V.

Still missing

• Case-level phenotype and comprehensive molecular testing results demonstrating an alternative explanatory diagnosis are required.