CDKN2A c.307_308insT, p.(Arg103LeufsTer17)

Reasoning

NM_000077.4:c.307_308insT is a frameshift insertion (NP_000068.1:p.(Arg103LeufsTer17)) in CDKN2A (p16INK4a). No official CSPEC/VCEP PVS1 specification is available for this gene in this case, so the generic ClinGen SVI PVS1 decision tree (PMC6185798) was applied. Gene-level review confirmed germline loss-of-function is an established disease mechanism for CDKN2A-related melanoma/FAMMM predisposition (pvs1_gene_context gate = eligible). The frameshift creates a premature termination codon 17 residues after codon 103 in a 156-residue protein. CDKN2A/p16INK4a is encoded by three exons (1alpha, 2, 3), with the bulk of the coding sequence, including codon 103 and the resulting new stop codon around codon 119, located within exon 2, well upstream (more than 50 nucleotides) of the terminal exon 2/exon 3 junction. Per the ClinGen SVI PVS1 frameshift/nonsense decision tree, a PTC located more than one exon away from (or >50 nt upstream of) the final exon-exon junction is predicted to trigger nonsense-mediated decay, consistent with loss of function. The transcript is the clinically relevant NM_000077.4 transcript, the truncation removes the majority (>60%) of the 156-aa protein including functionally critical ankyrin-repeat regions required for CDK4/CDK6 inhibition, and there is no evidence this exon is dispensable, alternatively spliced out, or a recognized hotspot for benign population-level LoF variation (variant and gene both absent from population LoF enrichment concerns per gnomAD absence). No competing in-frame rescue, alternative translation initiation, or exon-skipping evidence was identified. PVS1 is therefore met at very strong strength per the generic PMC6185798 framework, with PP3 not stacked for the same underlying prediction per framework guidance.

Rule

PVS1

Evidence considered

• pvs1_variant_assessment.json classifies the variant consequence as frameshift, transcript NM_000077.4:c.307_308insT, protein NP_000068.1:p.(Arg103LeufsTer17), and flags apply_generic_pvs1_framework=true with suggested_default_strength=PVS1, framework PMC6185798.
• pvs1_gene_context.json establishes pvs1_gene_gate=eligible: germline loss-of-function is a supported CDKN2A disease mechanism (lof_mechanism_supported=true), based on targeted germline literature review of CDKN2A-related melanoma predisposition; no official CSPEC/VCEP PVS1 framework exists for this gene so the generic fallback per PMC6185798 was directed.
• Applying the ClinGen SVI PVS1 decision tree (PMC6185798) to a frameshift variant: the resulting premature stop codon (p.Arg103LeufsTer17, terminating around codon 119 of the 156-aa p16INK4a protein) falls within exon 2, which is not the terminal exon and is well upstream of the final exon-exon junction, predicting NMD and truncation/loss of the C-terminal ankyrin-repeat region required for CDK4/CDK6 inhibitory function.
• case_summary.json compact_evidence shows the variant is absent from gnomAD v2.1, v4.1, and gnomAD-Canada, consistent with a rare, non-population-tolerated loss-of-function allele and no evidence contradicting pathogenic LoF interpretation at this exon.

Still missing

• No official CDKN2A-specific ClinGen VCEP PVS1 specification (e.g., exon-by-exon exception list) was available; assessment relies on generic PMC6185798 framework and general knowledge of CDKN2A exon/protein structure rather than a directly retrieved exon-boundary coordinate table for this case.
• No experimental NMD or minigene splicing data specific to this variant were available; NMD prediction is based on decision-tree position logic rather than direct transcript-level evidence.

Sources

Reasoning

No applicable CDKN2A VCEP/CSPEC population-frequency specification was retrieved; generic ACMG/AMP assessment applies. NM_000077.4:c.307_308insT is absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0. This satisfies generic ACMG/AMP PM2 (absent from population controls).

Rule

Generic ACMG/AMP PM2 (moderate): absent from controls in population databases.

Evidence considered

• The evidence bundle reports this exact normalized allele absent from gnomAD v2.1 (GRCh37 9-21971050-C-CA), gnomAD v4.1 (GRCh38 9-21971051-C-CA), and gnomAD-Canada v1.0.
• No CDKN2A VCEP/CSPEC population-frequency framework was available in the case materials; generic ACMG/AMP is the designated fallback.

Still missing

• The retrieved population records do not report allele-number or coverage metrics; confirmation from the live records could further document callable-sample depth but does not change the reported absence.

Sources

Reasoning

SpliceAI/Pangolin analysis of NM_000077.4:c.307_308insT returns a max delta score of 0.00, well below the ~0.1 benign-supporting threshold conventionally used with SpliceAI-family delta scores (per the score-interpretation framework in Jaganathan et al. 2019, PMID 30661751), so calibrated computational evidence does not predict a splice-altering effect. This variant is not a missense SNV, so the BayesDel/REVEL/aGVGD missense sub-path is not_available (prefetch explicitly skipped both due to 'not snv or no coords'); BayesDel is additionally excluded from PP3/BP4 use in this pipeline pending a verified published threshold. No CDKN2A VCEP/CSPEC lookup material overrides this. Because the variant's primary mechanism is a frameshift/PTC (already captured elsewhere as a PVS1 candidate), this BP4 call reflects only the absence of a predicted splicing consequence and carries low materiality against the established truncating mechanism; flagging for human review to confirm the splice-path BP4 assignment is appropriately weighted alongside the frameshift consequence rather than misread as broad benign support for the variant overall.

Rule

generic_acmg_bp4_spliceai

Evidence considered

• prefetch.json/evidence.json: SpliceAI max_delta_score = 0.00 (pangolin_SG 0.003, pangolin_SL -0.068) for NM_000077.4:c.307_308insT; evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.00).'
• Jaganathan et al. 2019 (PMID 30661751) established the SpliceAI delta-score framework; scores near 0.00, well under the ~0.1 low-confidence threshold, are conventionally used as benign-supporting (BP4) evidence against a splice-altering effect.
• prefetch.json: revel and bayesdel both skipped ('not snv or no coords'), confirming missense in-silico predictors are not_available for this frameshift insertion and cannot independently drive BP4.
• pvs1_variant_assessment.json: variant is classified as 'frameshift' consequence_class with the generic PVS1 framework as the primary evidence path, so this BP4 call is a narrow splice-path finding, not a statement about overall variant benignity.

Still missing

• No CDKN2A CSPEC/VCEP-specific PP3/BP4 lookup table was available to confirm or override this generic-threshold BP4 call.
• BayesDel score not computed and would not be usable for PP3/BP4 in this pipeline regardless (no verified published threshold available).

Sources

Reasoning

NM_000077.4:c.307_308insT is a frameshift insertion (p.(Arg103LeufsTer17)) located mid-exon in CDKN2A exon 2 (c.307/308, exon boundaries c.151-457), not an SNV, so no REVEL or BayesDel missense in-silico score is computable (both prefetch steps report 'skipped: not snv or no coords'); the missense sub-path is therefore not_available for this variant. The splice-impact sub-path was evaluated via SpliceAI/Pangolin: max delta score = 0.00 (pangolin_SG 0.003, pangolin_SL -0.068), far below any calibrated splice-altering threshold, so there is no computational evidence supporting a deleterious/splice-disrupting effect. No CDKN2A ClinGen VCEP was found (cspec.found=false) and no gene-specific PP3/BP4 lookup materials were retrieved (vcep_materials empty), so no VCEP override applies and generic calibrated thresholds govern.

Rule

generic_acmg_pp3_spliceai

Evidence considered

• prefetch.json/evidence.json: SpliceAI (reported alongside Pangolin scores) max_delta_score = 0.00 for NM_000077.4:c.307_308insT (pangolin_SG 0.003, pangolin_SL -0.068); evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.00).'
• Under the SpliceAI delta-score interpretation established in Jaganathan et al. 2019 (PMID 30661751), scores this far below the 0.2 (moderate) / 0.5 (strong) splice-altering thresholds provide no support for a splicing-disruptive effect, so PP3 is not met.
• prefetch.json: revel and bayesdel steps both report {"skipped": true, "reason": "not snv or no coords"} confirming the missense in-silico path is unavailable for this frameshift insertion.
• case_summary.json cspec.found=false and vcep_materials.summaries={} confirm no CDKN2A VCEP-specific PP3/BP4 lookup spreadsheet is present in this case to pre-assign a code.

Still missing

• No CDKN2A CSPEC/VCEP-specific PP3/BP4 lookup table was available to check for a pre-assigned gene-specific code.
• REVEL/BayesDel not computable because this is an indel, not an SNV.

Sources

Reasoning

The exact variant is absent from ClinVar; therefore there is no exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic assertion.

Rule

Apply PP5 only for an exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification.

Evidence considered

• ClinVar exact-match searches for NM_000077.4:c.307_308insT and genomic equivalents returned no result; the bundle states that this variant is absent from ClinVar.

Still missing

• An exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification would be required.

Sources

Reasoning

No applicable CDKN2A VCEP/CSPEC population-frequency specification was retrieved; generic ACMG/AMP assessment applies. The variant is absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0, so it does not meet the standalone high-frequency BA1 criterion.

Rule

Generic ACMG/AMP BA1: allele frequency is not high enough to satisfy the standalone benign criterion.

Evidence considered

• The evidence bundle reports this allele absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0.
• No CDKN2A VCEP/CSPEC population-frequency framework was available in the case materials; generic ACMG/AMP is the designated fallback.

Still missing

• The retrieved population records report absence but do not provide allele-number or coverage metrics; this does not affect the conclusion that BA1 is not met.

Sources

Reasoning

No applicable CDKN2A VCEP/CSPEC population-frequency specification was retrieved; generic ACMG/AMP assessment applies. The variant is absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0 and therefore is not more frequent than expected for a CDKN2A-related disorder.

Rule

Generic ACMG/AMP BS1: allele frequency is not greater than expected for the disorder.

Evidence considered

• The evidence bundle reports no observed allele in gnomAD v2.1, gnomAD v4.1, or gnomAD-Canada v1.0.
• No CDKN2A VCEP/CSPEC population-frequency framework was available in the case materials; generic ACMG/AMP is the designated fallback.

Still missing

• The retrieved population records do not include allele-number or ancestry-stratified coverage metrics; no observed allele frequency approaches a BS1 range.

Sources

Reasoning

No applicable CDKN2A VCEP/CSPEC specification was retrieved; generic ACMG/AMP assessment applies. The population datasets report the variant absent, with no heterozygous or homozygous carriers observed in healthy individuals. Thus, there is no BS2 evidence of the variant in healthy adults for a disorder expected to be fully penetrant at an early age.

Rule

Generic ACMG/AMP BS2: not met because no observation in healthy individuals is available.

Evidence considered

• The evidence bundle reports the variant absent from all three queried population datasets; consequently, it provides no healthy-carrier or homozygote observations.
• No CDKN2A VCEP/CSPEC framework was available in the case materials; generic ACMG/AMP is the designated fallback.

Still missing

• No phenotype-linked healthy-adult carrier cohort or homozygote observation was provided.

Sources

Reasoning

The exact variant is absent from ClinVar; therefore there is no exact-variant ClinVar expert-panel Benign or Likely benign assertion.

Rule

Apply BP6 only for an exact-variant ClinVar expert-panel Benign or Likely benign classification.

Evidence considered

• ClinVar exact-match searches for NM_000077.4:c.307_308insT and genomic equivalents returned no result; the bundle states that this variant is absent from ClinVar.

Still missing

• An exact-variant ClinVar expert-panel Benign or Likely benign classification would be required.

Sources

Reasoning

De novo occurrence cannot be assessed because the supplied case evidence contains no proband phenotype, parental genotypes, parentage confirmation, or testing methodology for this variant.

Rule

Generic ACMG/AMP PS2: de novo occurrence in an affected proband, with maternity and paternity confirmed.

Evidence considered

• No CDKN2A-specific ClinGen CSPEC, local gene-specific framework, or case-level de novo testing record was retrieved in the consolidated evidence bundle.
• The generic ACMG/AMP framework is the applicable fallback; PS2 requires a de novo variant in a patient with the disease and no family history, with maternity and paternity confirmed (Richards et al., 2015; PMID 25741868).

Still missing

• Proband phenotype and age of onset, assessed for consistency with a CDKN2A-associated germline condition.
• Laboratory report documenting the variant in the proband and its absence in both biological parents.
• Confirmation of biological parentage and assay sensitivity sufficient to evaluate possible parental mosaicism.

Reasoning

No CDKN2A ClinGen VCEP/CSPEC is available for this case (framework_mode: generic_acmg; cspec.found: false), so generic ACMG/AMP PS3 requirements apply, which require a validated functional assay demonstrating a damaging effect on protein function specifically for this variant. The only functional-theme literature retrieved (PMID:8603820, PMID:8668202) are 1996 studies of CDK4/CDK6-inhibitory activity and temperature-sensitive phenotypes of various CDKN2A missense and premature-termination mutants; their abstract snippets do not mention this specific variant (NM_000077.4:c.307_308insT / p.(Arg103LeufsTer17)), only full-text was not available for confirmation (fulltext_map.json is empty and only abstract screenshots, not extracted full text, were retrieved). Per strict citation-grounding rules, a paper can only be cited for PS3 if the exact variant is confirmed present in the abstract or full text, which could not be established here. No variant-specific functional assay evidence is present in the case bundle.

Still missing

• No full-text access was available to confirm whether PMID:8603820 or PMID:8668202 specifically tested the p.(Arg103LeufsTer17) frameshift/truncating allele rather than only missense or other premature-termination alleles at different positions.
• No dedicated functional assay report (e.g., CDK4/CDK6 binding, G1 arrest, senescence induction) for this exact frameshift variant was retrieved in this case bundle.

Reasoning

No variant-specific affected-case series, case-control comparison, or enrichment analysis is available for this variant.

Rule

Generic ACMG/AMP PS4 requires statistically increased prevalence in affected individuals compared with controls.

Evidence considered

• No CDKN2A-specific VCEP/CSPEC framework was retrieved; the case specifies the generic ACMG/AMP fallback.
• The literature triage contains no publication that mentions NM_000077.4:c.307_308insT or p.(Arg103LeufsTer17), and no PS4 case-control data are supplied.

Still missing

• Variant-specific counts in clinically ascertained affected individuals and an appropriate control dataset, with ascertainment and statistical analysis, are needed.

Sources

Reasoning

No affected-proband observation establishes this variant in trans with a pathogenic or likely pathogenic variant in a disorder with an applicable recessive inheritance model.

Rule

Generic ACMG/AMP PM3: in a recessive disorder, the variant is detected in trans with a pathogenic variant.

Evidence considered

• No CDKN2A VCEP/CSPEC or local gene-specific framework was retrieved; generic ACMG/AMP evaluation therefore applies.
• The consolidated case-evidence bundle contains no affected-proband observations, second germline variant, or phase information for NM_000077.4:c.307_308insT.

Still missing

• Affected-proband genotype data identifying a second pathogenic/likely pathogenic allele are unavailable.
• Phase between this variant and any second relevant allele is unavailable.
• No applicable recessive disease model or gene-specific PM3 specification was retrieved for this assessment.

Reasoning

De novo occurrence without confirmed parentage cannot be assessed because no parental testing or proband phenotype information is supplied.

Rule

Generic ACMG/AMP PM6: assumed de novo occurrence in an affected proband when parentage is not confirmed.

Evidence considered

• No CDKN2A-specific ClinGen CSPEC, local gene-specific framework, or case-level parental testing record was retrieved in the consolidated evidence bundle.
• The generic ACMG/AMP framework is the applicable fallback; PM6 is de novo occurrence without confirmation of paternity and maternity (Richards et al., 2015; PMID 25741868).

Still missing

• Proband phenotype and age of onset, assessed for consistency with a CDKN2A-associated germline condition.
• Parental genotype results showing whether the variant is absent from both parents.
• Whether biological parentage was confirmed; if it was not, documented parentage status is needed to distinguish PM6 from PS2.

Reasoning

Segregation cannot be assessed because no pedigree, affected or unaffected relative genotypes, phenotypes, ages, or informative meioses are supplied.

Rule

Generic ACMG/AMP PP1: cosegregation with disease in multiple affected relatives; the strength depends on the quality and quantity of informative family data.

Evidence considered

• No CDKN2A-specific ClinGen CSPEC, local gene-specific framework, or family segregation data was retrieved in the consolidated evidence bundle.
• The generic ACMG/AMP framework is the applicable fallback; PP1 requires cosegregation with disease in multiple affected family members in a gene definitively known to cause the disease (Richards et al., 2015; PMID 25741868).

Still missing

• Pedigree with affected status, relevant tumor/clinical diagnoses, and ages for tested relatives.
• Variant genotypes for informative affected and unaffected relatives, including testing method.
• Count of independent informative meioses and assessment of phenocopies, age-related penetrance, and incomplete family testing.

Reasoning

No proband phenotype, personal or family cancer history, age of onset, or phenotype-specificity assessment is provided.

Rule

Generic ACMG/AMP PP4 requires a patient phenotype or family history highly specific for a disease with a single genetic etiology.

Evidence considered

• No CDKN2A-specific VCEP/CSPEC framework was retrieved; the case specifies the generic ACMG/AMP fallback.
• The consolidated evidence bundle contains variant normalization and database searches but no individual-level phenotype data.

Still missing

• Clinical phenotype and family-history data are needed to determine whether the presentation is highly specific for a CDKN2A-related disorder.

Sources

Reasoning

No CDKN2A ClinGen VCEP/CSPEC is available for this case, so generic ACMG/AMP BS3 requirements apply, which require a validated functional assay showing no damaging effect on protein function for this specific variant. No such variant-specific benign/normal-function functional assay evidence was retrieved in this case bundle; the retrieved functional-theme literature (PMID:8603820, PMID:8668202) discusses CDKN2A mutations broadly and is not confirmed to include this exact frameshift allele.

Still missing

• No functional assay evidence of any kind (damaging or benign) was confirmed to test this exact variant, precluding a BS3 assessment.

Reasoning

Lack of segregation cannot be assessed because the bundle contains no genotype-phenotype data for affected relatives and no adequately evaluated unaffected carriers.

Rule

Generic ACMG/AMP BS4: lack of segregation with disease in informative affected relatives.

Evidence considered

• No CDKN2A-specific ClinGen CSPEC, local gene-specific framework, or family non-segregation data was retrieved in the consolidated evidence bundle.
• The generic ACMG/AMP framework is the applicable fallback; BS4 requires lack of segregation in affected members of a family (Richards et al., 2015; PMID 25741868).

Still missing

• Pedigree and genotype-phenotype data demonstrating discordance in informative relatives.
• For unaffected carriers, age, surveillance status, and disease penetrance context sufficient to determine whether non-segregation is informative.
• Assessment of phenocopies and alternative causes in affected non-carriers.

Reasoning

No observation documents this variant in cis with a pathogenic/likely pathogenic variant, or in trans with such a variant in a fully penetrant dominant disorder.

Rule

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

Evidence considered

• No CDKN2A VCEP/CSPEC or local gene-specific framework was retrieved; generic ACMG/AMP evaluation therefore applies.
• The consolidated case-evidence bundle contains no affected or unaffected individual-level co-occurrence data, second germline variant, or cis/trans phase result for NM_000077.4:c.307_308insT.

Still missing

• Phased genotype data from an informative individual are unavailable.
• The case bundle provides no co-occurring pathogenic/likely pathogenic variant and no phenotype or penetrance context needed to assess a trans observation under BP2.

Reasoning

No affected individual, phenotype attribution, or alternate molecular diagnosis is supplied.

Rule

Generic ACMG/AMP BP5 requires a variant found in a case with an alternate molecular basis for disease.

Evidence considered

• No CDKN2A-specific VCEP/CSPEC framework was retrieved; the case specifies the generic ACMG/AMP fallback.
• The consolidated evidence bundle contains no individual-level diagnostic findings or alternative causative variant.

Still missing

• A documented alternate molecular diagnosis in an affected individual and sufficient clinical correlation are needed.

Sources