CDKN2A c.221A>T, p.(Asp74Val)

Reasoning

The variant is absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0. This supports PM2 at supporting strength under the generic ACMG/AMP fallback; absence alone does not establish a stronger population-evidence level.

Rule

Generic ACMG/AMP fallback: absence from sufficiently large population reference datasets supports PM2 at supporting strength; no CDKN2A-specific CSPEC/VCEP population rule was available in the case materials.

Evidence considered

• Independent GRCh37 and GRCh38 gnomAD queries for the normalized variant returned absent.
• The gnomAD-Canada v1.0 query also returned absent.

Sources

Reasoning

CDKN2A has no ClinGen VCEP-specific CSPEC or gene-specific PP3/BP4 lookup available for this case (cspec found=false, final_classification_framework found=false), so generic ACMG/AMP calibrated computational evidence rules apply. NM_000077.4:c.221A>T is a missense variant (NP_000068.1:p.(Asp74Val)); the missense in-silico path governs, not the splice path. REVEL score = 0.903 (local REVEL v1.3 lookup), which meets the ClinGen SVI Pejaver et al. 2022 (PMID 36413997) calibrated REVEL threshold of >=0.773 for PP3_Moderate (though it does not reach the >=0.932 PP3_Strong threshold). SpliceAI predicts no splice impact (max delta = 0.00), so the splice-impact path contributes no additional pathogenic evidence and is not double-counted toward PP3. BayesDel score (0.514125) was retrieved but no verified published calibration threshold/publication for BayesDel is available to this pipeline, so it is treated as not_available and not used to support or oppose PP3.

Rule

generic_acmg_PP3_missense_REVEL_moderate

Evidence considered

• compact_evidence.revel.score = 0.903, found=true (local REVEL v1.3 predictor lookup, source_registry key 'revel').
• REVEL >=0.773 meets the ClinGen SVI-recommended PP3_Moderate threshold from Pejaver et al. 2022, 'Calibration of computational tools for missense variant pathogenicity classification and ClinGen recommendation for PP3/BP4 criteria' (PMID 36413997); score 0.903 does not reach the >=0.932 PP3_Strong threshold from the same publication.
• compact_evidence.spliceai.evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.00).' -- confirms the splice-impact sub-path is uninformative for this missense variant and is not being double-counted toward PP3.

Still missing

• BayesDel score is available (0.514125) but excluded from adjudication because no verified published calibration threshold/publication for this predictor is available to this pipeline.
• No CDKN2A-specific ClinGen VCEP PP3/BP4 lookup table exists for this case to override or refine the generic REVEL-based call.

Sources

Reasoning

No exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification is present. The available exact-variant ClinVar submissions are ordinary clinical-laboratory submissions and cannot trigger PP5.

Rule

Per task-specific rule, PP5 may be applied only for an exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification.

Evidence considered

• For exact ClinVar variation 820911, the submission audit reports zero expert-panel submissions.
• The listed submitters are Ambry Genetics and Labcorp Genetics; these are single-submitter clinical-laboratory assertions, not expert-panel assertions.

Still missing

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

Sources

Reasoning

The variant is absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0; it therefore does not meet BA1's stand-alone benign high-frequency requirement.

Rule

Generic ACMG/AMP fallback: BA1 requires a population allele frequency high enough to be stand-alone benign evidence; an absent variant cannot satisfy this criterion.

Evidence considered

• gnomAD v2.1 queried 9-21971137-T>A and returned absent.
• gnomAD v4.1 queried 9-21971138-T>A and returned absent.
• gnomAD-Canada v1.0 queried 9-21971138-T>A and returned absent.

Sources

Reasoning

No population allele frequency was observed in the queried population datasets, so the variant cannot exceed a disease-compatible BS1 frequency threshold.

Rule

Generic ACMG/AMP fallback: BS1 requires an allele frequency greater than expected for the disorder; absence cannot meet that requirement.

Evidence considered

• The variant was absent, rather than merely below a displayed frequency, in gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0.

Sources

Reasoning

The variant was absent from the queried population datasets; consequently, there are no observed adult healthy homozygotes or other population observations that could support BS2.

Rule

Generic ACMG/AMP fallback: BS2 requires observation in healthy adult individual(s) for a disorder expected to be fully penetrant early in life; no such observation is present.

Evidence considered

• No carriers were reported in gnomAD v2.1, gnomAD v4.1, or gnomAD-Canada v1.0, so homozygote evidence is unavailable.

Sources

Reasoning

BP1 applies to genes where disease is primarily/predominantly caused by truncating variants, such that missense variants are not expected to be pathogenic. OncoKB curated this specific missense variant (p.Asp74Val) as 'Likely Oncogenic' with a 'Likely Loss-of-function' biological effect, indicating that missense substitutions at this residue are recognized as capable of driving CDKN2A loss-of-function, i.e., missense change is not inherently benign-by-mechanism in this gene. The germline literature review underlying pvs1_gene_context also supports CDKN2A loss-of-function as the germline disease mechanism without restricting causative variants to truncating classes only. Together this argues against a BP1 pattern of 'primarily truncating variants cause disease' for CDKN2A.

Evidence considered

• OncoKB curates NM_000077.4 c.221A>T (p.D74V) as 'Likely Oncogenic', biological effect 'Likely Loss-of-function', indicating missense variants at this residue can plausibly cause CDKN2A functional loss.
• pvs1_gene_context.json (mechanism_rationale) states targeted germline literature review identified disease-focused publications supporting CDKN2A loss-of-function as a germline disease mechanism, without describing that mechanism as exclusive to truncating variant classes.

Still missing

• No explicit CDKN2A-specific statement quantifying the proportion of pathogenic variants that are truncating vs. missense was retrieved from this evidence bundle; the not_met call rests on functional/oncogenicity curation (OncoKB) rather than a direct gene-level truncating-vs-missense proportion statistic.

Sources

Reasoning

The calibrated missense in-silico evidence (REVEL = 0.903) falls in the pathogenic-supporting range (>=0.773, PMID 36413997) rather than the benign range, so BP4 cannot be applied for the missense path. SpliceAI's null splice-impact prediction (max delta = 0.00) does not independently support BP4 in the absence of a CDKN2A-specific VCEP splice-BP4 threshold/lookup; no such gene-specific material is present in this case (vcep_materials.files_found = [], cspec found=false), and no generically citable published SpliceAI calibration threshold with a named PMID was applied here. BayesDel is excluded from consideration for the same reason given under PP3 (no verified calibrated threshold publication available to this pipeline).

Rule

generic_acmg_BP4_missense_REVEL_not_met

Evidence considered

• compact_evidence.revel.score = 0.903 places the variant in the pathogenic-supporting/moderate range per Pejaver et al. 2022 (PMID 36413997), precluding a benign call under BP4 for the missense path.
• compact_evidence.spliceai.evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.00)' was reviewed but not applied toward BP4 because no CDKN2A VCEP-specific splice BP4 threshold/lookup is present in this case's materials.

Still missing

• No CDKN2A VCEP-specific splice or missense PP3/BP4 lookup table was found to check for a pre-assigned BP4 code.
• BayesDel score present but not usable without a named, verifiable calibration publication.

Sources

Reasoning

No exact-variant ClinVar expert-panel Benign or Likely benign classification is present. The available exact-variant ClinVar submissions are ordinary clinical-laboratory submissions and cannot trigger BP6.

Rule

Per task-specific rule, BP6 may be applied only for an exact-variant ClinVar expert-panel Benign or Likely benign classification.

Evidence considered

• For exact ClinVar variation 820911, the submission audit reports zero expert-panel submissions.
• Neither listed single-submitter clinical-laboratory assertion is Benign or Likely benign, and non-expert assertions are ineligible for BP6.

Still missing

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

Sources

Reasoning

PS1 requires an established pathogenic variant causing the exact same amino acid change (p.Asp74Val) via a different nucleotide substitution. No such comparator was identified in this case's evidence bundle. No CDKN2A VCEP/CSPEC was found (cspec.found=false; framework_mode=generic_acmg), so generic ACMG PS1 semantics apply, but the dedicated same-residue comparator search (pm5_candidates.json) was explicitly not run for this variant (eligible_for_classic_pm5_search=false, candidates=[]).

Evidence considered

• pm5_candidates.json reports found=false, pm5_mode='unknown', eligible_for_classic_pm5_search=false, with reason 'Unable to confirm classic same-residue PM5 semantics safely; skipping automatic candidate harvesting', and zero candidates collected (total_candidates_seen=0), leaving no basis to confirm or refute a same-amino-acid-change pathogenic comparator for PS1.

Still missing

• No same-amino-acid-change (p.Asp74Val, codon 74) pathogenic comparator variant was identified in the available evidence.
• Automatic PM5/PS1-relevant candidate harvesting was skipped by the pipeline (pm5_mode=unknown, eligible_for_classic_pm5_search=false); no independent literature/ClinVar search for other codon-74 variants was performed for this bundle.

Sources

Rule

Generic ACMG/AMP fallback applies because no CDKN2A CSPEC/VCEP framework was retrieved; PS2 requires a confirmed de novo occurrence in a patient with a phenotype consistent with the gene-associated disease.

Evidence considered

• The exact-variant ClinVar record is present in the case bundle, but its extracted submissions provide no parental testing, parentage confirmation, or de novo observation.

Still missing

• Obtain proband genotype and phenotype details.
• Obtain parental genotypes with parentage confirmation and determine whether either parent is affected.

Sources

Reasoning

No ClinGen CSPEC/VCEP or local custom gene-specific framework is available for CDKN2A in this case (vcep_materials.json: final_classification_framework.found = false, gene_vcep_dir = null), so generic ACMG/AMP applies. No well-established functional/biochemical assay data (e.g., p16INK4A CDK4/CDK6 kinase inhibition assay, cell-cycle arrest assay, or validated multiplexed functional assay) for NM_000077.4:c.221A>T (p.Asp74Val) was found anywhere in the case bundle (prefetch.json, evidence.json, literature_pass.json, vcep_materials.json). The only literature source tagged with theme 'functional' by the pipeline (PMID:25353071) concerns vemurafenib efficacy in BRAF(V600E)-positive papillary thyroid cancer and does not mention CDKN2A or this variant at all, so it is not grounded and cannot be used. OncoKB provides a curated 'Likely Loss-of-function' biological-effect label and 'Likely Oncogenic' classification, but this is a computational/curated aggregate assertion, not a validated functional assay result with documented assay type, replicates, or controls, so it cannot support PS3.

Evidence considered

• case_summary.json vcep_materials.final_classification_framework.found = false and gene_vcep_dir = null confirm no CSPEC/VCEP CDKN2A framework was retrieved for this case.
• evidence.json oncokb entry: biological_effect 'Likely Loss-of-function', classification 'Likely Oncogenic', citations_pmids = [] (no supporting PMID trail), so this curated call cannot be traced to a specific validated functional assay.
• literature_pass.json high_priority_papers: only PMID:25353071 is tagged theme 'functional', but its title/abstract ('Efficacy and tolerability of vemurafenib in patients with BRAF(V600E)-positive papillary thyroid cancer') concerns a different gene (BRAF) and does not mention CDKN2A or c.221A>T/p.Asp74Val, so it fails variant-grounding and cannot be cited for PS3.

Still missing

• No validated functional/biochemical assay (e.g., CDK4/CDK6 kinase inhibition, cell-cycle arrest, VAMP assay) publication identified for p.Asp74Val.
• No CSPEC/VCEP-specific PS3 assay framework available for CDKN2A to define required assay validation thresholds.
• OncoKB's Likely Loss-of-function annotation lacks a traceable primary functional-assay citation for this exact variant and cannot substitute for direct assay evidence.

Reasoning

No case-control, cohort, or other quantitative enrichment data for the exact CDKN2A p.(Asp74Val) variant were provided. The available ClinVar record contains laboratory assertions rather than prevalence data and cannot establish PS4.

Rule

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

Evidence considered

• No applicable CDKN2A VCEP/ClinGen specification was retrieved; generic ACMG/AMP evaluation is required.
• ClinVar variation 820911 has two laboratory submissions and no presented case-control or cohort enrichment data for p.(Asp74Val).
• The selectively screened available full texts did not mention CDKN2A, c.221A>T, p.Asp74Val, or p.D74V and therefore provide no exact-variant PS4 evidence.

Still missing

• Well-ascertained affected-case and ancestry-matched control counts for the exact variant, with a statistical enrichment analysis.

Sources

Reasoning

PM1 requires the variant to fall in a mutational hotspot or a well-established functional domain without benign variation. No CDKN2A-specific VCEP/CSPEC domain framework was found for this case. case_summary/evidence.json note that cancerhotspots.org flags codon D74 as a statistically significant hotspot and COSMIC lists 4 somatic occurrences (COSV58689077), but neither of those sources has a corresponding source_registry citation key in this bundle, so they cannot be formally cited as evidence_refs. OncoKB (a citable source) curated this variant as 'Likely Oncogenic' with a 'Likely Loss-of-function' biological effect, which is suggestive but is an oncogenicity/functional-impact call, not a structural domain or hotspot determination in itself. No curated protein-domain annotation (e.g., ankyrin-repeat boundaries or CDK4/6-binding interface residues) for position 74 is present in the bundle to independently establish domain criticality.

Evidence considered

• OncoKB (evidence.json/oncokb) curates NM_000077.4 c.221A>T (p.D74V) as 'Likely Oncogenic' with biological effect 'Likely Loss-of-function', consistent with but not itself proof of a hotspot/domain effect.
• case_summary.json compact_evidence.hotspots states 'This variant lies in a statistically significant hotspot' (cancerhotspots.org) and compact_evidence.cosmic reports 4 somatic occurrences (COSV58689077); these are described narratively here because neither 'hotspots' nor 'cosmic' has a source_registry key available for formal citation in this case.

Still missing

• cancerhotspots.org (statistically significant hotspot at D74) and COSMIC (COSV58689077, n=4 somatic occurrences) lack source_registry citation keys in this bundle and therefore cannot be cited as PM1 evidence_refs despite being described in case_summary/evidence.json.
• No curated structural/functional domain annotation for CDKN2A residue 74 is present in the evidence bundle.
• No germline-specific (as opposed to somatic pan-cancer) hotspot or domain constraint data was available to confirm applicability of PM1 in the germline familial-melanoma context.

Sources

Reasoning

No affected-proband observation with a second pathogenic/likely pathogenic CDKN2A variant, no phase determination, and no recessive disease context are available. PM3 cannot be assessed.

Rule

Generic ACMG/AMP PM3: apply only when a variant is observed in trans with a pathogenic variant in an affected individual for a recessive disorder.

Evidence considered

• No CDKN2A VCEP/CSPEC framework was retrieved; the case therefore uses the generic ACMG/AMP fallback.
• The case record identifies germline ClinVar submissions for CDKN2A p.Asp74Val but supplies no affected-proband genotype, companion-variant, segregation, or phase information.

Still missing

• Affected-proband genotype data identifying any second CDKN2A variant.
• Phase evidence establishing whether the target variant is in trans with a pathogenic/likely pathogenic allele.
• Disease-specific inheritance context supporting use of a recessive allelic criterion.

Sources

Reasoning

PM5 requires a different missense change at the same residue (codon 74) previously established as pathogenic. pm5_candidates.json explicitly reports the classic same-residue comparator search was not performed (eligible_for_classic_pm5_search=false) and returned zero candidates, so no basis exists to confirm or refute PM5.

Evidence considered

• pm5_candidates.json: pm5_mode='unknown', eligible_for_classic_pm5_search=false, reason 'Unable to confirm classic same-residue PM5 semantics safely; skipping automatic candidate harvesting'; candidate_summary shows total_candidates_seen=0, same_residue_candidates=0, expert_panel_plp_candidates=0.

Still missing

• pm5_candidates.json reports eligible_for_classic_pm5_search=false and same_residue_candidates=0; the classic PM5 same-residue alternate-pathogenic-missense search was not performed for this variant.
• No alternate pathogenic missense change at codon 74 (e.g., p.Asp74Gly/Asn/Tyr/etc.) was identified anywhere in the available evidence bundle to support or exclude PM5.

Sources

Reasoning

Not assessed: no presumed de novo occurrence without confirmed parentage is documented for this variant.

Rule

Generic ACMG/AMP fallback applies because no CDKN2A CSPEC/VCEP framework was retrieved; PM6 requires a de novo occurrence when parentage is not confirmed.

Evidence considered

• The exact-variant ClinVar record is present in the case bundle, but its extracted submissions provide no parental testing, parentage confirmation, or presumed de novo observation.

Still missing

• Obtain proband genotype and phenotype details.
• Obtain parental genotypes and state whether biological parentage was confirmed.

Sources

Reasoning

Not assessed: the bundle contains no informative affected or unaffected relatives, pedigree, genotypes, or count of segregating meioses.

Rule

Generic ACMG/AMP fallback applies because no CDKN2A CSPEC/VCEP framework was retrieved; PP1 requires cosegregation of the variant with disease in multiple affected family members.

Evidence considered

• The exact-variant ClinVar record is present in the case bundle, but its extracted submissions contain no family segregation data.

Still missing

• Provide a pedigree with phenotypes, ages, and disease status for tested relatives.
• Provide relative genotypes for this variant and the number of informative cosegregations/meioses.

Sources

Reasoning

PP2 requires a gene with a low rate of benign missense variation where missense is a common disease mechanism. No CDKN2A-specific VCEP/CSPEC PP2 threshold was found (cspec.found=false), and no gene-level missense constraint metric (e.g., missense Z-score/observed-over-expected ratio, or ClinGen gene-disease missense curation) is present anywhere in this evidence bundle to establish the required benign-missense-rate threshold.

Evidence considered

• case_summary.json cspec.found=false and framework_mode='generic_acmg' confirm no gene-specific PP2 threshold is available for CDKN2A in this case.
• No missense constraint/benign-variation-rate source was retrieved in prefetch.json, evidence.json, or vcep_materials.json for this gene.

Still missing

• No gene-level missense constraint statistic (gnomAD missense Z/o.e. or equivalent) is present in the evidence bundle for CDKN2A.
• No CDKN2A-specific VCEP/CSPEC PP2 rule or missense-mechanism statement was found; generic ACMG PP2 cannot be evaluated without this gene-level constraint data.

Reasoning

No individual-level phenotype, tumor phenotype, family history, or diagnostic exclusion data were provided for a carrier of the exact variant. Thus, phenotype specificity for a CDKN2A-associated condition cannot be assessed.

Rule

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

Evidence considered

• No applicable CDKN2A VCEP/ClinGen specification was retrieved; generic ACMG/AMP evaluation is required.
• The case bundle provides ClinVar condition labels but no carrier-level phenotype or diagnostic workup that can establish a highly specific phenotype.

Still missing

• Individual-level phenotype and family-history data for an exact-variant carrier, including evidence that the presentation is highly specific for a CDKN2A-associated disorder and that relevant alternative diagnoses have been excluded.

Reasoning

As with PS3, no ClinGen CSPEC/VCEP or local custom CDKN2A framework was found, so generic ACMG/AMP applies. No well-established functional assay showing normal/wild-type activity for p.Asp74Val was identified in the case bundle; the sole 'functional'-themed source (PMID:25353071) is unrelated to CDKN2A and does not mention this variant, so it cannot ground a benign functional call. Curated OncoKB output leans toward loss-of-function, which if anything argues against BS3, but is not a primary assay result and is not being used to call BS3 not_met with confidence beyond insufficiency.

Evidence considered

• case_summary.json vcep_materials.final_classification_framework.found = false and gene_vcep_dir = null confirm no CSPEC/VCEP CDKN2A framework was retrieved for this case.
• literature_pass.json high_priority_papers: only PMID:25353071 is tagged theme 'functional', but it concerns BRAF/vemurafenib in papillary thyroid cancer and does not mention CDKN2A or c.221A>T/p.Asp74Val, so it fails variant-grounding and cannot be used for BS3.

Still missing

• No functional/biochemical assay data demonstrating wild-type-like activity for p.Asp74Val was identified in this case bundle.
• No CSPEC/VCEP-specific BS3 assay framework available for CDKN2A.

Reasoning

Not assessed: the bundle contains no informative non-segregation observation, such as an affected relative lacking the variant or an unaffected relative carrying it in a context where penetrance and age are adequately addressed.

Rule

Generic ACMG/AMP fallback applies because no CDKN2A CSPEC/VCEP framework was retrieved; BS4 requires lack of segregation with disease in affected family members.

Evidence considered

• The exact-variant ClinVar record is present in the case bundle, but its extracted submissions contain no family non-segregation data.

Still missing

• Provide a pedigree with phenotypes, ages, and disease status for tested relatives.
• Provide relative genotypes for this variant and assess age-dependent penetrance and phenocopies before interpreting any non-segregation.

Sources

Reasoning

No observation establishes that CDKN2A p.Asp74Val is in cis with a pathogenic/likely pathogenic variant, or in trans with such a variant in a setting where this would support BP2. BP2 cannot be assessed.

Rule

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

Evidence considered

• No CDKN2A VCEP/CSPEC framework was retrieved; the case therefore uses the generic ACMG/AMP fallback.
• The available ClinVar case record contains classifications and condition labels but no co-occurring-variant or phase information for this assessment.

Still missing

• Genotype data for any additional CDKN2A variant in the same individual.
• Phase evidence demonstrating cis or trans configuration with a pathogenic/likely pathogenic variant.
• Phenotype and inheritance information needed to judge whether a trans observation would be compatible with BP2.

Sources

Reasoning

No affected individual with the exact variant and an independently established alternative molecular diagnosis was provided. BP5 cannot be assessed from aggregate database assertions or laboratory classifications.

Rule

Generic ACMG/AMP BP5 requires an alternate molecular basis for disease in an affected individual carrying the variant.

Evidence considered

• No applicable CDKN2A VCEP/ClinGen specification was retrieved; generic ACMG/AMP evaluation is required.
• The available case bundle contains no individual-level co-occurrence or alternative-diagnosis evidence.

Still missing

• Individual-level evidence of an independently established alternative molecular diagnosis that fully explains the observed phenotype in an exact-variant carrier.