IDH2 c.520G>A, p.(Ala174Thr)

PS1 not met: no established pathogenic p.Ala174Thr arising from a different nucleotide change exists in ClinVar (only the index variant, VUS, single submitter) or in the pm5_candidates/curated evidence; no comparator to satisfy the criterion.

PM1 not met: residue 174 is not in a statistically significant hotspot (Cancer Hotspots negative) and is unreported in COSMIC; no curated annotation places A174 in a critical, well-established functional domain; proximity to the R172 hotspot is insufficient under generic ACMG/AMP.

PM5 not assessed: the required alternate pathogenic missense at residue 174 was not identified, but the dedicated same-residue comparator search was skipped by the pipeline (pm5_candidates pm5_mode unknown, 0 candidates), so absence cannot be confirmed; flagged for human review.

PP2 not assessed: gene-level missense constraint (low benign missense rate) is not available for IDH2 in this case (no gnomAD constraint metric, HCI prior not supported), and no CSPEC defines PP2 applicability for this gain-of-function oncogene; the missense-mechanism component alone is insufficient.

BP1 not applicable: IDH2 disease is caused by missense gain-of-function variants (R140/R172 neomorphic 2-HG activity), not primarily truncating variants, so the BP1 premise does not apply.

Governing framework: no ClinGen CSPEC/VCEP exists for IDH2 (cspec found=false, framework_mode=generic_acmg, precedence_rank 3); all criteria were adjudicated against the generic ACMG/AMP 2015 rules (PMID:25741868).

No VCEP/CSPEC exists for IDH2 (cspec.found = false, gene_vcep_dir = null, final_classification_framework.source = generic_acmg_fallback), so the generic ACMG/AMP 2015 framework (Richards et al., PMID 25741868) governs all four criteria in this group.

The case contains no proband-parent trio data, no parental testing, no de novo event documentation, no pedigree/segregation information, and no non-segregation observations. PS2, PM6, PP1, and BS4 are therefore all not_assessed due to absence of evidence.

ClinVar VCV004621918 is a single-submitter germline VUS (Ambry Genetics, SCV007230384) with no criteria provided and no validated PMIDs; it contributes no usable de novo or segregation signal and must not be used to infer any of these criteria.

All 5 ClinVar-linked PMIDs retrieved (25626707, 25730230, 23169492, 22947299, 23881473) are newborn screening / expanded carrier screening / DTC genetic testing policy statements unrelated to IDH2; none mention c.520G>A and none were reviewed for this group.

Framework: generic ACMG/AMP 2015 (PMID:25741868) via generic_acmg_combination_rules; no CSPEC/VCEP specification exists for IDH2 (vcep_materials: 'no cspec url available').

PS3: not_assessed - no well-established functional study showing a deleterious effect of p.Ala174Thr was identified; no enzyme/cellular/splicing assay data exist in the case evidence.

BS3: not_assessed - no well-established functional study showing no damaging effect was identified; SpliceAI (max delta 0.00) is in silico and does not qualify.

Net functional-group contribution: none. Neither PS3 nor BS3 can be counted toward the pathogenic or benign tally; the variant remains functionally uncharacterized.

No criterion was inferred from the ClinVar single-submitter VUS classification or from any other source's final classification.

Governing framework: no ClinGen CSPEC/VCEP specification exists for IDH2 (cspec found=false, gene_vcep_dir=null), so the generic ACMG/AMP 2015 framework (Richards et al., PMID:25741868; generic_acmg_combination_rules) applies to all criteria in this group.

PS4 not met: no case-control or cohort enrichment evidence for c.520G>A exists in any retrieved source (ClinVar single-lab VUS, gnomAD AF 6.2e-07 with 1/1.6M alleles, COSMIC absent, OncoKB unknown effect, literature empty of variant mentions).

PP4 not assessed: no proband or carrier phenotype/family-history data is available; the only condition label ('Inborn genetic diseases') is generic and non-specific.

PP5 and BP6 not met by policy: both require an exact-variant ClinVar expert-panel classification (P/LP for PP5; B/LB for BP6); the only ClinVar assertion is a 1-star, single-laboratory VUS (Ambry Genetics) with expert_panel_submissions=0, and ordinary laboratory submissions must never trigger either criterion.

BP5 not met: no reported case carrying this variant with an alternate molecular basis of disease.

All retrieved publications (PMID:23169492, PMID:25626707, PMID:25730230) are policy/position statements with no IDH2 or c.520G>A content; they were consulted and excluded for this group's criteria.

PM3 not met: the variant was never observed in trans with a pathogenic IDH2 variant in an affected proband; the case bundle contains no second variant, no phase/segregation data, and no literature report of c.520G>A, so the required observation is entirely absent.

BP2 not met: no observation in trans with a pathogenic variant for a fully dominant disease (the germline IDH2 disorder is autosomal recessive; no fully dominant germline IDH2 disease is established) and no observation in cis with a pathogenic variant in any inheritance pattern.

Both allelic criteria rest on absence of observations: gnomAD v4.1 (1 heterozygous allele, 0 homozygotes), ClinVar single-submitter VUS with no criteria/PMIDs, and 5 screened publications with no variant mention. Neither criterion contributes points to the classification.

The ClinVar VUS classification was not used to infer or refute any criterion.

PP3 not met: REVEL 0.749 is below the calibrated PP3-supporting threshold (>= 0.773, Pejaver et al. 2022, PMID 36413997); SpliceAI max delta 0.00 predicts no splice impact; BayesDel excluded as insufficiently calibrated (no verified published threshold); aGVGD unavailable.

BP4 not met: REVEL 0.749 is above the calibrated BP4-supporting threshold (<= 0.128, PMID 36413997), arguing against benignity; SpliceAI null result covers only the splice sub-path and is insufficient alone for BP4; BayesDel excluded as insufficiently calibrated.

Net contribution of the in_silico_splicing group: neither PP3 nor BP4 is applied; no calibrated computational evidence (missense or splice path) is counted for this variant under the generic ACMG fallback.

No IDH2 VCEP/CSPEC or gene-specific PP3/BP4 lookup materials exist (vcep_materials.json empty, cspec.found = false), so the generic ACMG/AMP 2015 framework (PMID 25741868) with ClinGen SVI calibrated REVEL thresholds (PMID 36413997) governs.

Governing framework: no ClinGen CSPEC/VCEP or local gene-specific framework exists for IDH2 (cspec found=false, gene_vcep_dir=null, final_classification_framework fallback with precedence_rank 3), so generic ACMG/AMP 2015 (PMID 25741868) applies, with non-VCEP gnomAD frequency conventions PM2<0.1%, BS1>0.3%, BA1>1%.

Population frequency is extremely low: absent from gnomAD v2.1 and gnomAD-Canada v1.0; single allele in gnomAD v4.1 (total AF 6.195e-07, MAF 0.00006%, 1/1,614,196 alleles; highest subpopulation AF 8.474e-07 in NFE; 0 homozygotes).

BA1 not met (AF ~5 orders of magnitude below the >1%/>5% threshold); BS1 not met (AF far below the >0.3% threshold and any plausible disease-expected frequency); BS2 not met (0 homozygotes, no clinically confirmed healthy-adult observations, no established early-onset fully penetrant germline IDH2 disorder).

PM2 met at supporting strength: frequency far below the 0.1% low-frequency threshold, consistent with a rare/private allele; 0 homozygotes; supporting-strength calibration per ClinGen SVI point system (PMID 32720330).

Retrieved literature (5 papers from ClinVar citations) contains no IDH2 or c.520G>A content — all are newborn-screening, carrier-screening, or DTC-testing policy/ethics papers — and contributes no population-frequency evidence; none cited for this group.