Five-step scale from benign to pathogenic. This variant is uncertain. Select a step to read what it means.
This variantScenario
BenignPathogenic
Uncertain.The recorded classification is uncertain significance.
In plain language
Gene-level background from the recorded interpretation.
Words with a dotted underline have a short definition. Hover or tap them.
About ATM
ATM encodes a kinase that acts as a master controller of the cellular DNA damage response: it detects double-strand breaks and coordinates DNA repair, cell cycle arrest, or apoptosis through downstream targets including p53, BRCA1, and CHK2. Loss-of-function mutations in both copies of ATM cause ataxia telangiectasia, an autosomal recessive disorder characterized by neurological problems, immune deficiency, and cancer predisposition. ATM functions as a tumor suppressor: individuals with ataxia telangiectasia are prone to childhood lymphomas, leukemias, and breast cancer, while carriers of a single altered copy have increased risk of breast, pancreatic, prostate, and other cancers. Somatic ATM mutations also occur in lymphoid malignancies and solid tumors, and ATM-deficient cancers are often especially sensitive to DNA-damaging treatments.
LYFE Sciences is an AI system, and it can make mistakes.
Criteria may be applied incorrectly, sources may be misread, and a
confident-looking classification can still be wrong.
Double-check every criterion and its underlying evidence
before relying on any call.
01The reasoning
How it was classified
The recorded framework is ClinGen ATM VCEP v1.5. The criteria and strengths shown below come from the saved assessment; this page does not recalculate the classification.
Recorded applied criteria
The applied criteria and their strengths are shown as assessed under the stated framework.
Met (supporting): gnomAD v4.1 total allele frequency 0.000186% (3/1,613,222 alleles) is far below the 0.001% threshold.
Benign criteria applied · 1
BP4Supporting
BP4 · supporting
Met (supporting): REVEL 0.043 is well below the 0.249 missense threshold; SpliceAI 0.01 also rules out splice impact.
Recorded classification: Uncertain significance
Under the ClinGen HBOP ATM VCEP v1.5 framework exactly two criteria were met - PM2 supporting (rarity) and BP4 supporting (benign-leaning) - and their conflicting directions trigger the conflicting-evidence rule, yielding VUS (uncertain significance).
02The sources
Evidence at a glance
Evidence summaries and captured source records, where available. Missing entries should not be interpreted as negative findings.
gnomAD v4.1
gnomAD v2.1
Population frequency
3 / 1,613,222 alleles
gnomAD v4.1: 3 / 1,613,222 alleles. Recorded AF 1.85963e-06. Recorded homozygote count 0.
gnomAD v2.1
This variant is present in gnomAD v2.1 (AF= 3.99186e-06; MAF= 0.00040%, 1/250510 alleles, homozygotes = 0) and has highest observed frequency in the Admixed American population (AF= 2.8967e-05; MAF= 0.00290%, 1/34522 alleles, homozygotes = 0).
OncoKB did not identify variant-specific reviewed functional evidence for this variant; gene-level curated context is available for reviewer follow-up. ATM, a kinase involved in the DNA damage response, is mutated in various solid and hematologic malignancies.
ACMG/AMP criteria under ClinGen ATM VCEP v1.5, with the data and sources behind each one.
28 criteria with recorded statuses: 2 met · 5 not met · 5 not assessed · 16 not applicable. Open any row for the full reasoning, the rule and the evidence behind it.
Governing framework: ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ATM Version 1.5 (cspec doc 639508985, precedence rank 1, framework found). Used instead of generic ACMG/AMP; no generic fallback needed.
Variant: NM_000051.4:c.838A>G → p.(Ile280Val), a missense substitution in exon 7 of ATM (cDNA 813-1051), no protein length change, no premature termination codon, NMD not relevant, canonical splice consensus = false, SpliceAI max delta = 0.01 (no predicted splice impact).
PVS1 not_applicable: the ATM VCEP PVS1 decision tree covers only null variants (nonsense, frameshift, canonical +/-1,2 splice, initiation codon, single/multi-exon deletion); a missense SNV with no splice impact is outside the criterion's scope. Pipeline generic scaffold agrees (apply_generic_pvs1_framework = false).
PM4 not_applicable: the ATM VCEP v1.5 restricts PM4 to stop-loss variants; p.Ile280Val is a missense change with no protein length alteration.
BP3 not_applicable: the ATM VCEP v1.5 cspec marks BP3 'Not Applicable'; the variant is also not an in-frame indel and no repeat-region context exists.
All three criteria in this group are excluded on variant-class grounds with high confidence; no literature evidence is required and no publication in the case registry mentions the exact variant (verified by full-text search of all 5 available full texts).
No criterion in this group was inferred from any other source's final classification (ClinVar VUS status not used as evidence).
Governing framework: ClinGen HBOP ATM VCEP v1.5 (cspec.genome.network doc 639508985, precedence rank 1, framework_mode 'vcep'). Generic ACMG/AMP is not used for PM1/PP2/BP1 because the CSPEC explicitly marks all three 'Not Applicable'.
PS1 (same amino acid change as an established P/LP variant): not met - no P/LP variant producing p.Ile280Val exists in any collected source; ClinVar VCV000407462 is VUS (6 labs, no expert panel); pm5_candidates.json reports 0 same-residue candidates; no publication in the bundle mentions the variant or codon 280. An alternate-nucleotide comparator yielding the same amino acid is also structurally impossible from the ATT reference codon.
PM5: not applicable - the ATM VCEP v1.5 PM5 rule is a truncation-cutoff rule (frameshift/truncating/splice variants with PTC upstream of p.Arg3047), which does not cover this missense variant. Generic same-residue PM5 is not the VCEP rule; even if it were, no same-residue P/LP comparators were found.
PM1, PP2, BP1: not applicable per explicit CSPEC applicability flags; no further evidence collection required.
No criterion in this group is met or applied; this group contributes no pathogenic or benign weight to the final classification.
Governing framework: ClinGen HBOP ATM VCEP v1.5 (cspec found, framework_mode=vcep, precedence rank 1); generic ACMG/AMP 2015 is not used because an applicable VCEP specification exists and addresses each of PS2, PM6, PP1, and BS4.
PS2, PM6, and BS4 are all declared Not Applicable by the ATM VCEP v1.5 cspec, so they are recorded as not_applicable with no strength; no case-level de novo, parental-testing, or family-segregation data exists to revisit this.
PP1 is Applicable under the VCEP (AR condition, affected relatives must carry both variants identified in the proband), but the case bundle contains zero segregation data (no genotyped affected relatives, no pedigree/meioses), so PP1 is not_assessed and the specific gap is documented.
No publication was used for this group: none of the 5 retrieved full texts (PMID 25741868, 32761968, 34262154, 24418350, 25394175) mention c.838A>G or p.Ile280Val, and none of the 10 triaged papers has a segregation/de novo theme; PP1/PS2/PM6/BS4 therefore rest on the VCEP rule set only.
This group contributes no met pathogenic or benign evidence; the classification outcome for this variant must be driven by other criterion groups.
ATM c.838A>G (p.Ile280Val) is a missense variant in exon 7; functional criteria are applicable but the evidence base is insufficient to apply them.
Neither PS3 nor BS3 can be assessed under the governing ClinGen HBOP ATM VCEP v1.5 because no result from a VCEP-approved assay (Mitui 2009, Barone 2009, Scott 2002) can be confirmed for this variant: the approved-assay table contains no per-variant rows, the three assay papers are not in the case bundle, and the only direct functional measurement (Sun et al. 2025 prime-editing saturation screen, DeepATM_predicted=No) classified the variant 'Intermediate' with Combined_score -1.104 - an equivocal readout that meets neither the PS3 fail-to-rescue rule nor the BS3 rescue rule.
The Sun et al. 2025 (PMID 40580951) assay is not among the three VCEP v1.5-approved assays and postdates the specification; per the gene-framework index guidance, it must not be silently converted to a PS3/BS3 strength - flagged for human review instead.
Potential cross-source confusion ruled out: the 'c.838A>G (p.Arg280Gly)' entry in PMID 32761968 is in TP53, not ATM, and is not evidence for this variant.
Both criteria returned not_assessed with needs_human_review=true; the primary gap is an approved-assay functional readout (kinase activity and/or radiosensitivity) for this specific variant.
Governing framework: ClinGen HBOP ATM VCEP v1.5 (cspec doc 639508985, framework_mode=vcep) applies; generic ACMG/AMP 2015 (PMID:25741868) serves only as background since the VCEP supersedes it for this gene.
PS4: The VCEP defines PS4 (Strong) as case-control evidence with p<=0.05 AND OR/HR/RR>=2 or lower 95% CI >=1.5. No case-control study of the exact variant c.838A>G (p.Ile280Val) exists in the evidence bundle. The only case-control-format ATM study (PMID:34262154) was read in full text and does not contain the variant; its enrichment signals are aggregate gene-level (VUS OR 1.41, 95% CI 1.21-1.64; LOF OR 2.6) and cannot be attributed to this variant. PS4: not_met.
PP4: VCEP marks PP4 Not Applicable; A-T phenotype-specificity evidence is handled by the VCEP's PM3/BP2 points table, not PP4. PP4: not_applicable.
PP5/BP6: VCEP marks both Not Applicable, and independently the pipeline's exact-variant expert-panel rule does not trigger - ClinVar 407462 is Uncertain significance (2 stars, 6 laboratory submissions, 0 expert panel submissions), so there is no expert-panel P/LP or B/LB classification to support PP5 or BP6. Both: not_applicable.
BP5: VCEP marks BP5 Not Applicable; no alternate-molecular-basis evidence exists. BP5: not_applicable.
Net group result: no criterion in this group is met; all four benign/phenotype criteria are Not Applicable under the ATM VCEP and PS4 lacks any variant-specific case-control data.
Governing framework: ClinGen HBOP ATM VCEP v1.5 (cspec doc 639508985, precedence_rank 1, framework_mode=vcep) takes precedence over generic ACMG/AMP 2015. Both PM3 and BP2 are Applicable and are scored exclusively via the ATM PM3/BP2 points table (ATM_PM3_BP2_1.5.pdf); generic ACMG definitions were intentionally not applied.
PM3 = not_assessed: the evidence bundle contains zero A-T proband observations of c.838A>G - no homozygous carrier, no in-trans compound heterozygote, no phase data, no phenotype+second-variant documentation. The Natera submission labeled 'Ataxia-telangiectasia' carries no assertion criteria or phenotypic/zygosity detail. 0 points are obtainable under the VCEP table, so PM3 cannot be scored in either direction; population frequency (gnomAD AF ~4e-6) satisfies the <=0.01% prerequisite but grants no points.
BP2 = not_assessed: no unaffected (non-A-T, >=18y) carrier of c.838A>G with a P/LP ATM variant in trans (confirmed or phase-unknown) is documented in ClinVar or any literature source in the corpus; no in-cis observations exist. 0 (negative) points are obtainable, so BP2 cannot be scored in either direction.
No literature source reports this variant (NM_000051.4:c.838A>G / p.Ile280Val) at all: PMID:32761968 (Ryu et al. 2020) full text was searched and contains no mention of the variant; PMID:34262154 (Dalmasso et al. 2021) full text is unavailable in this case and the abstract contains no variant-level data. Neither paper contributes allelic evidence; both were consulted and are documented as negative.
Absence of allelic observations in the consulted corpus is not proof that none exist globally; PM3 and BP2 remain unresolvable (not_assessed) pending A-T proband sequencing reports and phase-resolved carrier data.
Governing framework: ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer (HBOP) Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ATM v1.5 (cspec, doc 639508985) - a VCEP exists, so generic ACMG/AMP is not used.
Variant: NM_000051.4:c.838A>G (p.Ile280Val), missense, exon 7 - assessed on both computational sub-paths per VCEP instructions (splice analysis considered for all variant types).
PP3: not met. REVEL 0.043 does not exceed 0.7333 (missense path) and SpliceAI max delta 0.01 does not reach 0.2 (splicing path); all corroborating predictors (AlphaMissense 0.0853, EVE 0.205, boostDM 0.121) are low.
BP4: met (Supporting). REVEL 0.043 <= 0.249 satisfies the VCEP missense BP4 threshold. Per the lab's BP4 policy override, the SpliceAI sub-path (0.01 <= 0.1) is not counted as an independent line for this missense variant; outcome is identical under the VCEP's literal text either way.
BP7: not applicable. Criterion scope is synonymous and deep intronic variants; this is a missense exonic variant, and no RNA data (BP7(RNA)) exists in the case evidence.
No double-counting: the benign prediction is applied once (BP4 via REVEL); SpliceAI is used only to rule out splice impact and BayesDel/Sun et al. predictors are non-binding corroboration.
No literature source in the case bundle mentions the exact variant c.838A>G (full texts of PMID:32761968 and PMID:34262154 checked; string absent), so no publication is cited as variant-specific evidence.
Governing framework: ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer (HBOP) ATM VCEP v1.5 (cspec document 639508985). Generic ACMG/AMP was not used because an applicable gene-specific specification exists.
BA1 not met: grpmax filtering AF 5.53e-06 (0.000553%) in gnomAD v4.1 is ~900-fold below the VCEP BA1 threshold of >0.5%.
BS1 not met: grpmax filtering AF 0.000553% is ~90-fold below the VCEP BS1 threshold of >0.05%.
PM2 met at supporting strength: gnomAD v4.1 total AF 0.000186% (3/1,613,222 alleles, 0 homozygotes) is <=0.001% as required; corroborated by gnomAD v2.1 (0.00040%, 0 homozygotes) and absence from gnomAD-Canada v1.0. VCEP instructs PM2_Supporting, not moderate.
BS2 not applicable: HBOP ATM VCEP v1.5 lists BS2 as Not Applicable; unaffected-carrier observations are scored through the PM3/BP2 points system instead.
No disease-prevalence or penetrance adjustment is needed: the VCEP grpmax filtering thresholds already encode the expected allele frequency for ATM-associated disease (recessive A-T, cancer susceptibility), and the observed frequency is orders of magnitude below the benign cutoffs and at/under the rarity cutoff.
No homozygotes observed in any gnomAD v4.1 or v2.1 population, consistent with a rare variant rather than a common benign polymorphism.
No criteria match.
Met · 2
PM2SupportingMet (supporting): gnomAD v4.1 total allele frequency 0.000186% (3/1,613,222 alleles) is far below the 0.001% threshold.
Details ▾Hide ▴
Reasoning
HBOP ATM VCEP v1.5 PM2 rule: frequency <=0.001% in gnomAD v4 dataset, used as PM2_Supporting (not moderate), with n=1 in a single subpopulation considered sufficiently rare. Observed gnomAD v4.1 total AF = 1.86e-06 (0.000186%, 3/1,613,222 alleles), which is well below the 0.001% threshold; grpmax FAF = 5.53e-06 (0.000553%) is also below 0.001%; homozygotes = 0. Rarity is corroborated by gnomAD v2.1 (AF 3.99e-06, 0.00040%, 1/250,510, 0 homozygotes) and by absence from gnomAD-Canada v1.0. Highest subpopulation AF (Admixed American 3.33e-05, 2/60,006) exceeds the absolute 0.001% figure if applied per-subpopulation, but the VCEP rule keys on gnomAD v4 dataset frequency with the n=1-per-subpopulation sufficiency clause, and the overall dataset frequency satisfies the threshold; this nuance is noted in warnings rather than blocking PM2_Supporting. Per VCEP instructions, PM2 is applied at supporting strength and is not counted as conflicting evidence for otherwise benign variants.
Rule
Frequency <=0.001% in gnomAD v4 dataset; use as PM2_Supporting (not moderate); n=1 in a single subpopulation is sufficiently rare (HBOP ATM VCEP v1.5)
BP4SupportingMet (supporting): REVEL 0.043 is well below the 0.249 missense threshold; SpliceAI 0.01 also rules out splice impact.
Details ▾Hide ▴
Reasoning
ATM HBOP VCEP v1.5 BP4 (Supporting) requires (missense path) REVEL <= 0.249 OR (splicing path) SpliceAI <= 0.1. Per the lab's BP4 missense/splicing policy override recorded in vcep_materials.json, exactly one mechanism is evaluated for a missense variant and only the missense sub-path (REVEL <= 0.249) counts; the SpliceAI sub-path is NOT applied as an independent BP4 line for missense variants. REVEL = 0.043 <= 0.249, so BP4 is met via the missense sub-path. Note: the override does not change the outcome here - under the VCEP's literal text, SpliceAI max delta 0.01 (<= 0.1) would also satisfy the splicing sub-path, so BP4 would be met either way; SpliceAI is therefore recorded as ruling out splice impact but is not double-counted as a second independent BP4 line. Corroboration: Sun et al. 2025 Suppl Table S1 row for c.838A>G shows REVEL 0.043 (matches pipeline lookup), AlphaMissense 0.0853, Combined score -1.10 with Classification 'Intermediate' (not Non-functional) and DeepATM_predicted=No (directly measured); BayesDel (not part of VCEP rules) = -0.43543, benign-leaning. All computational evidence is concordant for a benign-leaning missense effect.
Rule
ATM HBOP VCEP v1.5 BP4 (Supporting): Missense: REVEL score <=0.249; Splicing: No predicted impact via splicing (SpliceAI <=0.1). Lab override: for missense variants use ONLY the missense sub-path (REVEL), never both sub-paths independently.
PS1Not met: no established pathogenic or likely pathogenic variant producing p.Ile280Val exists; ClinVar lists only uncertain significance for this variant.
Details ▾Hide ▴
Reasoning
Variant is NM_000051.4:c.838A>G, p.(Ile280Val), a missense change in ATM. The governing framework is the ClinGen HBOP ATM VCEP v1.5 (CSPEC found, precedence rank 1). Its PS1 rule states: 'Use for missense changes as long as splicing is ruled-out for both alterations' (missense sub-path: same amino acid change as an established P/LP variant). No established P/LP variant producing p.Ile280Val exists in the collected evidence: ClinVar (VCV000407462) classifies c.838A>G itself as Uncertain significance with 6 clinical laboratory submissions and 0 expert-panel submissions; the PM5 comparator candidate collection (pm5_candidates.json) saw 0 same-residue candidates and 0 expert-panel P/LP candidates; no literature source in the bundle mentions c.838A>G, Ile280, or I280. Additionally, from the reference codon ATT (positions c.838-840), no single-nucleotide substitution other than c.838A>G itself can produce a Val codon (GTT/GTC/GTA/GTG all require A>G at position 838 plus additional changes), so an alternate-nucleotide same-amino-acid comparator is structurally impossible at this codon. The VCEP splicing caveat for the VUA itself is satisfied (SpliceAI max delta score 0.01, no splice impact predicted), but this does not overcome the absence of a P/LP reference variant. PS1 not met.
Rule
ATM VCEP v1.5 PS1: 'Use for missense changes as long as splicing is ruled-out for both alterations. Use ATM PS1 Splicing table for splicing variants with similar predictions or observations of splice defect.'
Evidence considered
cspec
clinvar
pm5_candidates
spliceai
Still missing
No systematic query of LOVD/HGMD or ClinVar for alternate nucleotide changes at codon 280 producing p.Ile280Val was run; a same-AA-change comparator is structurally impossible from the ATT reference codon, but an exhaustive database sweep would fully close this.
An expert-panel (ATM VCEP) classification of any variant at codon 280 would be needed before PS1 could ever be considered met.
PS4Not met: no case-control study of this exact variant exists; the only case-control ATM study in the literature does not list c.838A>G.
Details ▾Hide ▴
Reasoning
Governing framework: ClinGen HBOP ATM VCEP v1.5 (cspec) defines PS4 (Pathogenic Strong) as: 'Case-control studies; p-value <=.05 AND (Odds ratio, hazard ratio, or relative risk >=2 OR lower 95% CI >=1.5).' The evidence bundle contains no case-control study of the exact variant c.838A>G (p.Ile280Val). The only case-control-format ATM study present (PMID:34262154, GenoMEL/MelaNostrum melanoma consortium) was checked in full text: c.838A>G is not listed among the 1,004 unique ATM variants retained, and only the 13 individual VUS observed in >=3 cases were analyzed per-variant (c.838A>G is not one of them). Its reported enrichment is aggregate/gene-level (all VUS OR 1.41, 95% CI 1.21-1.64, p=1.03e-02; LOF OR 2.6, 95% CI 1.56-4.11), which per the exact-variant grounding rule cannot be attributed to this specific variant, and the aggregate VUS OR also fails the VCEP's OR>=2 threshold. PMID:32761968 (HBOC splicing cohort) is an uncontrolled case series and does not mention the variant. No other case-control source exists in the bundle (PMID:35534704 is an uncontrolled 1682-patient case series without full text). PS4 is therefore not met: no variant-specific case-control enrichment evidence.
Rule
Case-control studies; p-value <=.05 AND (Odds ratio, hazard ratio, or relative risk >=2 OR lower 95% CI >=1.5).
Evidence considered
cspec ATM VCEP v1.5 PS4 rule (case-control p<=0.05 AND OR/HR/RR>=2 or lower 95% CI >=1.5)
Full text PMID:34262154 - searched for c.838A>G / p.Ile280Val / Ile280 / I280 / 280Val: variant absent; only aggregate gene-level ORs reported
Full text PMID:32761968 - searched for c.838A>G / p.Ile280Val / Ile280 / I280 / 280Val: variant absent; uncontrolled HBOC splicing cohort
ClinVar record 407462 submission audit - no expert panel submissions; only laboratory VUS submissions
Still missing
No case-control study of the exact variant c.838A>G (p.Ile280Val) has been identified; gene-level aggregate ATM association data (PMID:34262154) cannot be applied to a single variant.
PP3Not met: REVEL 0.043 is far below the 0.7333 threshold, and SpliceAI 0.01 is far below 0.2.
Details ▾Hide ▴
Reasoning
ATM HBOP VCEP v1.5 PP3 requires (missense path) REVEL > 0.7333 OR (splicing path) SpliceAI >= 0.2 for silent/missense/in-frame and intronic variants outside donor/acceptor +/-1,2. Both sub-paths were assessed for this missense variant (NM_000051.4:c.838A>G, p.Ile280Val, exon 7): REVEL = 0.043 (well below 0.7333) and SpliceAI max delta = 0.01 (well below 0.2; DS_AG 0.01, all other deltas 0.0). Neither threshold is met. Corroborating predictors in Sun et al. 2025 Suppl Table S1 (AlphaMissense 0.0853, EVE 0.205, boostDM 0.121, CADD 16.54) are all low/benign-leaning and do not support a damaging effect. PP3 is therefore not met with high confidence.
Rule
ATM HBOP VCEP v1.5 PP3 (Supporting): Missense: REVEL >0.7333; Splicing: SpliceAI >=0.2 for silent, missense/in-frame and intronic variants outside donor and acceptor 1,2 sites.
BA1Not met: grpmax filtering allele frequency 0.000553% is roughly 900-fold below the 0.5% threshold.
Details ▾Hide ▴
Reasoning
Governing framework is the ClinGen HBOP ATM VCEP v1.5 (cspec 639508985), which defines BA1 as grpmax filtering AF >0.5% in gnomAD v4. The variant c.838A>G (p.Ile280Val) has grpmax FAF = 5.53e-06 (0.000553%) in gnomAD v4.1 (joint), roughly 900-fold below the 0.5% threshold, so BA1 is not met. Total AF is 1.86e-06 (0.00019%, 3/1,613,222 alleles) with 0 homozygotes; highest subpopulation AF is Admixed American 3.33e-05 (2/60,006), still far below threshold. Per SVI general guidance referenced by the VCEP, the grpmax filtering AF is the appropriate metric for BA1, and it does not approach the threshold.
Rule
Grpmax Filtering AF >0.5% in gnomAD v4 dataset (HBOP ATM VCEP v1.5)
BS1Not met: grpmax filtering allele frequency 0.000553% is about 90-fold below the 0.05% threshold.
Details ▾Hide ▴
Reasoning
HBOP ATM VCEP v1.5 defines BS1 as grpmax filtering AF >0.05% in gnomAD v4. The observed grpmax FAF = 5.53e-06 (0.000553%) is ~90-fold below the 0.05% threshold, so BS1 is not met. gnomAD v4.1 total AF 0.00019% (3/1,613,222 alleles, 0 homozygotes) and gnomAD v2.1 total AF 0.00040% (1/250,510) are both far below the threshold expected for a benign population polymorphism in ATM (recessive A-T / cancer susceptibility). No threshold artifact (e.g., population-specific enrichment or homozygote excess) is present that would argue for a higher effective frequency.
Rule
Grpmax Filtering AF >0.05% in gnomAD v4 dataset (HBOP ATM VCEP v1.5)
PS3Not assessed: the only functional measurement available (Sun 2025) was equivocal, 'Intermediate', not a confirmed failure to rescue an ATM function. Flagged for human review: the assay is not approved by the ATM expert panel.
Details ▾Hide ▴
Reasoning
Governing framework is the ClinGen HBOP ATM VCEP v1.5 (cspec doc 639508985): PS3_Supporting applies when a variant fails to rescue an ATM-specific feature (e.g. phosphorylation of ATM-specific targets); PS3_Moderate requires failure to rescue BOTH an ATM-specific feature AND radiosensitivity; PS3_Strong is explicitly prohibited ('Do not use as strong.'). Approved assays are defined only in the VCEP supplementary table (clingen_hbop_atm_supplementary_tables_1_and_2_v1.xlsx): ATM kinase activity (Mitui 2009/PMID 18634022, Barone 2009/PMID 19431188, Scott 2002/PMID 11805335; proposed PS3_Supporting) and ATM radiosensitivity (Mitui 2009 only; proposed PS3_Moderate, combinable only with a kinase-activity result). That table is a methodology/strength-calibration table with no per-variant rows and does not list c.838A>G (p.Ile280Val); none of the three cited assay papers' full texts are available in this case bundle, so it cannot be confirmed that this variant was ever tested in a VCEP-approved assay. The only direct functional measurement found is Sun et al. 2025 (PMID 40580951) Table S1 (source key vcep_suppl_tables1_pmid_40580951), which contains a row for c.838A>G with DeepATM_predicted=No (variant directly assayed in the prime-editing saturation screen, not model-extrapolated), Classification='Intermediate', Confidence='Medium-high', Combined_score=-1.104. 'Intermediate' is an equivocal readout that does not meet the VCEP PS3 rule (confirmed failure to rescue an ATM-specific feature; 'Non-functional' class), and this assay is NOT one of the three VCEP v1.5-approved assays and postdates the specification, so no VCEP-calibrated strength can be assigned. Per the gene-framework index guidance, when Sun 2025 is the only PS3/BS3 evidence the case must be flagged for human review rather than silently assigning an uncalibrated strength. PS3 is therefore not assessed; the VCEP PS3 rule cannot be applied on current evidence.
Rule
PS3: Do not use as strong. PS3_Moderate: Use when a variant fails to rescue both an ATM specific feature (e.g. phosphorylation of ATM-specific targets) AND radiosensitivity. PS3_Supporting: Use when a variant fails to rescue an ATM specific feature, only (e.g. phosphorylation of ATM-specific targets). Do not use for radiosensitivity-only as that is not a feature specific to ATM deficiency. (ClinGen HBOP ATM VCEP v1.5)
Evidence considered
ATM VCEP v1.5 PS3 rules (cspec): Supporting = fails to rescue ATM-specific feature only; Moderate = fails to rescue ATM-specific feature AND radiosensitivity; Strong = do not use
VCEP-approved assay calibration table (kinase activity: Mitui/Barone/Scott, proposed PS3_Supporting; radiosensitivity: Mitui 2009 only, PS3_Moderate combinable with kinase result) - no per-variant row for c.838A>G
Sun et al. 2025 Table S1 row c.838A>G: Classification=Intermediate, Confidence=Medium-high, DeepATM_predicted=No, Combined_score=-1.104 (directly assayed but equivocal and not VCEP-approved)
PMID:32761968 checked - its c.838A>G (p.Arg280Gly) is TP53, not ATM; no ATM functional data for this variant
PMID:34262154 checked - no variant mention, no functional assay data
Still missing
Confirm whether c.838A>G (p.Ile280Val) was tested in any VCEP-approved assay (Mitui 2009/PMID 18634022, Barone 2009/PMID 19431188, Scott 2002/PMID 11805335) - full texts of these papers are not in the case bundle
Obtain an approved-assay readout (e.g. phosphorylation of ATM substrates S1981/SMC1/NBS1/CHK2/p53 by Western blot in an ATM-null LCL) showing the variant fails to rescue an ATM-specific feature, optionally plus radiosensitivity (Mitui 2009) for PS3_Moderate
If relying on Sun et al. 2025 (PMID 40580951): human review needed on whether the prime-editing saturation screen result (Intermediate, Medium-high) qualifies as functional evidence under ATM VCEP v1.5 and how 'Intermediate' maps to the fail-to-rescue threshold (it currently does not meet the Non-functional bar)
PM3Not assessed: no ataxia-telangiectasia proband observations of this variant (homozygous or in trans) were available.
Details ▾Hide ▴
Reasoning
Governing framework is the ClinGen HBOP ATM VCEP v1.5 specification (cspec doc 639508985, framework_mode=vcep, precedence_rank 1), which makes PM3 Applicable and points-based per the ATM PM3/BP2 table (applies regardless of variant class): points per unrelated A-T proband, phenotype 'confident' = 4.0 (confirmed in trans) / 2.0 (phase unknown) / 1.0 (second variant unidentified or VUS) / 2.0 (homozygous, max 2 individuals); phenotype 'consistent' = half those values; PM3 Supporting=1.0, Moderate=2.0, Strong=4.0, Very Strong=8.0 points. The consolidated evidence bundle contains no A-T proband observation of c.838A>G whatsoever: no homozygous carrier, no compound-heterozygous/in-trans observation, no phase (parental/offspring) testing, and no A-T phenotype + second-variant data in any consulted source. All 6 ClinVar submissions (VCV000407462) are VUS with single-submitter criteria and no expert panel; their evidence signals are only PP3 (GeneDx, Ambry, Color), PS3-negative ('functional studies have not been reported', Color) and PM2 (Invitae) - none carry co-occurrence, zygosity, or phase information. The one submission labeled for the 'Ataxia-telangiectasia' condition (Natera SCV002089526) has no assertion criteria and no phenotype/zygosity/second-variant detail, so it cannot be counted as an A-T proband observation. Exhaustive search of the bundle and all available full texts (32761968 reviewed line-by-line; 34262154 full text unavailable) found no mention of c.838A>G/p.Ile280Val. The population-frequency prerequisite (variant may not exceed general population frequency >0.01%) is satisfied (gnomAD v2.1 total AF 3.99e-06; v4.1 AMR 2/60006, NFE 1/1179796, 0 homozygotes), but that is a prerequisite only and accrues no points. With 0 points obtainable, PM3 cannot be scored in either direction and is returned not_assessed pending A-T proband observations. Generic ACMG/AMP 2015 PM3 was deliberately not substituted because an applicable VCEP specification exists.
Rule
ClinGen HBOP ATM VCEP v1.5 (cspec doc 639508985) PM3: points per unrelated A-T proband via ATM PM3/BP2 table - phenotype confident: 4.0 (confirmed in trans) / 2.0 (phase unknown) / 1.0 (second variant unidentified or VUS) / 2.0 (homozygous, max 2 individuals); phenotype consistent: half; PM3 Supporting >= 1.0, Moderate >= 2.0, Strong = 4.0, Very Strong >= 8.0 points; variant may not exceed general population frequency >0.01%.
Evidence considered
ATM PM3/BP2 v1.5 table (vcep_atm_pm3_bp2_1_5): points per unrelated A-T proband (confident phenotype 4.0 confirmed in trans / 2.0 phase unknown / 1.0 second variant unidentified or VUS / 2.0 homozygous max 2; consistent phenotype half); PM3 Supporting=1.0, Moderate=2.0, Strong=4.0, Very Strong=8.0
cspec PM3 entry: Applicable, instructionsToUse references ATM PM3/BP2 table; 'Multiple unrelated cases are additive'; variant may not exceed general population frequency >0.01%
clinvar (VCV000407462) submission audit: 6 submissions all VUS (Natera, GeneDx, Ambry, Color, Labcorp/Invitae, Catlab); evidence signals PP3/PS3-negative/PM2 only; no co-occurrence, phase, zygosity, or A-T proband data; 0 expert panel submissions
PMID:32761968 full text: no mention of c.838A>G / p.Ile280Val / residue 280 anywhere; no A-T probands, no allelic/phase data for this variant
PMID:34262154: abstract only (full text unavailable in case corpus); no variant-level or allelic data verifiable
Still missing
No documented A-T proband observation of c.838A>G (homozygous, or compound heterozygous in trans with a P/LP ATM variant) in any consulted source; dedicated A-T proband sequencing/allelism literature beyond the 10 triaged papers was not searched
No phase information (parental or offspring testing) is available for any carrier of c.838A>G - trans vs cis unresolved
Natera SCV002089526 lists the 'Ataxia-telangiectasia' condition but provides no assertion criteria, phenotype detail, zygosity, or second-variant information; it cannot support point assignment
PP1Not assessed: no family segregation data for this variant were available (no affected relatives genotyped).
Details ▾Hide ▴
Reasoning
PP1 is Applicable under the ClinGen HBOP ATM VCEP v1.5 (cspec doc 639508985): AR condition only, requiring segregation in affected relatives who carry both variants identified in the proband, with strength by count of affected relatives (1 = Supporting, 2 = Moderate, >=3 = Strong); AD co-segregation is explicitly not to be used (PMID 32773770). However, the case bundle contains no segregation evidence whatsoever: no affected relatives genotyped for c.838A>G, no proband family pedigree, no meioses, and no non-segregation observations (term searches across all bundle sections returned only VCEP rule text and abstract snippets). With zero informative affected relatives, no PP1 strength can be assigned; status is not_assessed pending family segregation data.
Rule
ClinGen HBOP ATM VCEP v1.5 (cspec) PP1: AR Condition - affected relatives must have both variants identified in proband; Strong = segregation in >=3 affected relatives, Moderate = 2 affected relatives, Supporting = 1 affected relative; AD Condition - do not use (PMID 32773770).
Still missing
No segregation data: no affected relatives genotyped for c.838A>G / p.Ile280Val
No proband family pedigree or meioses data in the case bundle
No non-segregation (informative unaffected/affected family member) observations reported
BS3Not assessed: the only functional measurement available (Sun 2025) was equivocal, 'Intermediate', not a confirmed rescue of function. Flagged for human review: the assay is not approved by the ATM expert panel.
Details ▾Hide ▴
Reasoning
Governing framework is the ClinGen HBOP ATM VCEP v1.5 (cspec doc 639508985): BS3_Supporting applies when a variant rescues EITHER an ATM-specific feature OR radiosensitivity; BS3_Moderate applies when a variant rescues BOTH an ATM-specific feature AND radiosensitivity. Approved assays are defined only in the VCEP supplementary table (clingen_hbop_atm_supplementary_tables_1_and_2_v1.xlsx): kinase activity assays (Mitui 2009/PMID 18634022, Barone 2009/PMID 19431188, Scott 2002/PMID 11805335; proposed BS3_Supporting) and radiosensitivity (Mitui 2009 only; proposed BS3_Supporting). This is a calibration table with no per-variant rows; c.838A>G (p.Ile280Val) is not listed, and the three assay papers' full texts are not available in this case bundle, so no VCEP-approved assay result can be confirmed for this variant. The only direct functional measurement found is Sun et al. 2025 (PMID 40580951) Table S1 (source key vcep_suppl_tables1_pmid_40580951): c.838A>G, DeepATM_predicted=No (directly assayed), Classification='Intermediate', Confidence='Medium-high', Combined_score=-1.104. 'Intermediate' is not a rescue/'Functional' readout, so it does not satisfy the BS3 rule (rescue of an ATM-specific feature), and the assay is not VCEP v1.5-approved (postdates the spec), so no calibrated benign strength can be assigned. Supporting population context (gnomAD v4.1 AF 1.86e-06, 3 heterozygotes, 0 homozygotes; ClinVar VUS from 6 labs) is frequency/reporting data, not functional assay evidence, and does not substitute for an approved-assay rescue result. Per the gene-framework index guidance, when Sun 2025 is the only PS3/BS3 evidence the case must be flagged for human review. BS3 is therefore not assessed.
Rule
BS3: BS3_Supporting: Use when a variant rescues EITHER an ATM specific feature OR rescues radiosensitivity. BS3_Moderate: Use when a variant rescues both an ATM specific feature (e.g. phosphorylation of ATM-specific targets) AND radiosensitivity. (ClinGen HBOP ATM VCEP v1.5)
Evidence considered
ATM VCEP v1.5 BS3 rules (cspec): Supporting = rescues EITHER ATM-specific feature OR radiosensitivity; Moderate = rescues BOTH
VCEP-approved assay calibration table (kinase activity: proposed BS3_Supporting; radiosensitivity Mitui 2009: BS3_Supporting) - no per-variant row for c.838A>G
Sun et al. 2025 Table S1 row c.838A>G: Classification=Intermediate (not Functional), DeepATM_predicted=No, Combined_score=-1.104 - does not demonstrate rescue
PMID:32761968 checked - its c.838A>G (p.Arg280Gly) is TP53, not ATM; no ATM functional data for this variant
PMID:34262154 checked - no variant mention, no functional assay data
Still missing
Confirm whether c.838A>G (p.Ile280Val) was tested in any VCEP-approved assay (Mitui 2009/PMID 18634022, Barone 2009/PMID 19431188, Scott 2002/PMID 11805335) and whether the reported readout was normal (rescue)
Obtain an approved-assay readout showing the variant rescues an ATM-specific feature (and, for BS3_Moderate, also radiosensitivity)
If relying on Sun et al. 2025 (PMID 40580951): human review needed on whether the directly-assayed 'Intermediate' result (not 'Functional') can be interpreted as rescue evidence under ATM VCEP v1.5
BP2Not assessed: no unaffected carriers of this variant with a pathogenic ATM variant in trans were documented.
Details ▾Hide ▴
Reasoning
Under the governing ClinGen HBOP ATM VCEP v1.5 specification, BP2 is Applicable and points-based per the ATM PM3/BP2 table (applies regardless of variant class), repurposing generic BP2 (in trans with a pathogenic variant / in cis observations) entirely into a points system: points per unaffected (non-A-T) individual carrying the variant with a P/LP ATM variant in trans = -4.0 (confirmed trans) / -2.0 (phase unknown) / -2.0 lab setting or -1.0 database setting (homozygous, max -2.0); BP2 Supporting=-1.0, Moderate=-2.0, Strong<=-4.0; the CSPEC additionally requires the individual be >=18 years with no evidence of A-T. The case corpus contains no such observations: no unaffected carrier of c.838A>G with a P/LP ATM variant in trans (confirmed or phase-unknown) is documented in ClinVar (all 6 submissions VUS, no co-occurrence/phase/zygosity signals), and no literature source in the bundle reports this variant in any individual, unaffected or affected. No in-cis observations exist either. With 0 (negative) points obtainable, BP2 cannot be scored in either direction and is returned not_assessed pending phase-resolved unaffected-carrier observations. Generic ACMG/AMP 2015 BP2 was deliberately not substituted because the applicable VCEP specification supersedes it ('See ATM PM3/BP2 table for approach to assign points per proband, and final BP2 code assignment based on the sum of BP2-related points').
Rule
ClinGen HBOP ATM VCEP v1.5 (cspec doc 639508985) BP2: points per unaffected (non-A-T) individual carrying the variant with a P/LP ATM variant in trans -4.0 (confirmed trans) / -2.0 (phase unknown) / -2.0 lab setting or -1.0 database setting (homozygous, max -2.0); BP2 Supporting=-1.0, Moderate=-2.0, Strong<=-4.0; individual must be >=18 years with no evidence of A-T.
Evidence considered
ATM PM3/BP2 v1.5 table (vcep_atm_pm3_bp2_1_5): points per unaffected (non-A-T) individual with P/LP variant in trans -4.0 (confirmed trans) / -2.0 (phase unknown) / -2.0 lab or -1.0 database (homozygous, max -2.0); BP2 Supporting=-1.0, Moderate=-2.0, Strong<=-4.0
cspec BP2 entry: Applicable, instructionsToUse requires >=18 years and no evidence of A-T for homozygous/in-trans downgrade assessments; original ACMG summary (in trans with pathogenic variant / in cis) superseded by the VCEP points table
clinvar (VCV000407462) submission audit: 6 VUS submissions; no carrier-level co-occurrence, phase, or zygosity data; no unaffected-individual observations
PMID:32761968 full text and PMID:34262154 (abstract only): no report of c.838A>G in any individual, no in-trans carrier observations
Still missing
No documented unaffected (non-A-T, >=18 years) carrier of c.838A>G with a P/LP ATM variant in trans (confirmed or phase-unknown) in any consulted source
No phase-resolved (trans/cis) observation of c.838A>G exists anywhere in the corpus; parental/offspring phase testing data would be required to ever apply BP2 points
Not applicable · 16· recorded as not applicableShow reasons ▾Hide reasons ▴PVS1N/ANot applicable: c.838A>G is a missense change p.(Ile280Val), not a null variant, so the loss-of-function rule does not apply.
Details ▾Hide ▴
Reasoning
Governing framework is the ClinGen HBOP ATM VCEP v1.5 (cspec, precedence rank 1), whose PVS1 decision guide (ATM_PVS1_1.5.pdf) applies only to null variants: nonsense, frameshift, canonical +/-1,2 splice-site variants, initiation codon variants, and single/multi-exon deletions, using default transcript NM_000051.3/ENST00000278616.8. The variant under assessment, NM_000051.4:c.838A>G, is a missense substitution p.(Ile280Val) (NP_000042.3) located in exon 7 (cDNA coords 813-1051 of NM_000051.4 per Mutalyzer selector). It is not a nonsense, frameshift, or initiation codon change; it does not alter protein length, introduces no premature termination codon, and NMD is not relevant. canonical_splice_consensus = false, and SpliceAI predicts no splice impact (max delta score = 0.01; DS_AG 0.01, all others 0.0), so it is not a canonical splice-site or splice-impacting variant either. The pipeline's generic PVS1 scaffold independently confirms exclusion: pvs1_variant_assessment.json records consequence_class = 'missense', variant_bucket = 'other', and apply_generic_pvs1_framework = false ('This variant does not fall into the default generic PVS1 null-variant buckets'). PVS1 therefore cannot apply to this variant type and is marked not_applicable rather than met/not_met.
Rule
ATM VCEP v1.5 PVS1 Decision Tree (ATM_PVS1_1.5.pdf, adapted from PMID 30192042): PVS1 applies to null variants (nonsense, frameshift, canonical +/-1,2 splice sites, initiation codon, single/multi-exon deletion) in a gene where LOF is a known mechanism; default transcript NM_000051.3/ENST00000278616.8; caveats for variants at the extreme 3' end and for splice variants predicted to cause in-frame exon skipping leaving the remainder of the protein intact.
Evidence considered
variant consequence class = missense, p.(Ile280Val), no protein length change
exon 7 of 63 (cDNA position 838 within 813-1051) per Mutalyzer selector_short for NM_000051.4
PS2VCEPNot applicable: the ATM expert panel does not use de novo evidence for this recessive condition.
Details ▾Hide ▴
Reasoning
Governing framework is the ClinGen HBOP ATM VCEP v1.5 specification (cspec doc 639508985; framework_mode=vcep, precedence rank 1), which explicitly declares PS2 'Not Applicable' with no strength rules. An applicable VCEP specification exists, so generic ACMG/AMP PS2 is not applied. This is consistent with ATM being a recessive ataxia-telangiectasia gene where de novo events are not a disease mechanism under this specification. Independently, the case bundle contains no proband, parental-testing, or de novo data (variant-input-only cold check; no 'de novo', 'mother', 'father', or 'parent' evidence found), so PS2 could not have been met even under a generic framework.
PM1VCEPNot applicable: the ATM expert panel does not use the mutational-hotspot criterion for this gene.
Details ▾Hide ▴
Reasoning
The governing ClinGen HBOP ATM VCEP v1.5 (CSPEC, doc 639508985) explicitly marks PM1 as 'Not Applicable' for ATM. The VCEP does not use the generic ACMG/AMP PM1 (mutational hotspot / functional domain) criterion; per the user's standing policy, the ClinGen CSPEC takes precedence over generic ACMG/AMP when a specification exists. No PM1 evidence evaluation is performed. For completeness only: the Cancer Hotspots query for 'ATM I280' returned no statistically significant hotspot row, and the VCEP PVS1 guide notes FAT/PI3-K/FATC are treated as one contiguous domain (PMID 28508083) - neither observation would be routed to PM1 under this framework.
VCEP rule
Not Applicable - ATM VCEP v1.5 (CSPEC) marks PM1 as Not Applicable.
PM4N/ANot applicable: this rule covers only stop-loss variants, and p.Ile280Val is a missense change.
Details ▾Hide ▴
Reasoning
The governing ATM VCEP v1.5 (cspec) restricts PM4 (Pathogenic Moderate) to stop-loss variants: 'Use for stop-loss variants.' The variant under assessment is a missense single-nucleotide substitution c.838A>G / p.(Ile280Val): it is not a stop-loss change, involves no in-frame indel, and produces no protein length alteration (position_first 279, position_last_original/predicted 280 in the Mutalyzer protein prediction; predicted protein retains full length). Generic ACMG/AMP PM4 also targets in-frame insertions/deletions and stop-loss variants, none of which apply here. PM4 is therefore not_applicable to this variant type.
Rule
ATM VCEP v1.5 (cspec rule): 'PM4 — Use for stop-loss variants.' (Pathogenic Moderate).
Evidence considered
protein change p.(Ile280Val) is a single amino acid substitution — protein length unchanged
variant is not a stop-loss variant and not an in-frame indel (Mutalyzer normalization: substitution, point, A>G at c.838)
PM5N/ANot applicable: the ATM panel's PM5 applies only to truncating variants upstream of p.Arg3047; this is a missense change.
Details ▾Hide ▴
Reasoning
Under the governing ClinGen HBOP ATM VCEP v1.5, PM5 is NOT the classic same-residue missense criterion. The CSPEC PM5 rule is a truncation-cutoff rule: 'Apply to frameshifting or truncating variants with premature termination codons upstream of p.Arg3047. Apply to splice variants as with premature termination codons upstream of p.Arg3047 where PVS1_VS(RNA) is applied based on high quality observed splicing impact and must be NMD prone.' The variant under assessment, c.838A>G (p.Ile280Val), is a missense variant, so it falls entirely outside the scope of the VCEP's PM5 rule (no frameshift/truncation/PTC, not a splice variant). The pipeline artifact pm5_candidates.json confirms this interpretation (pm5_mode: 'truncation_cutoff_gene_specific', eligible_for_classic_pm5_search: false, recommendation level: 'not_applicable'). Even under a generic-ACMG same-residue PM5 interpretation, the collected evidence contains 0 same-residue P/LP comparator candidates (0 candidates seen, 0 expert-panel P/LP candidates), so PM5 would not be met on the merits either. PM5 not applicable.
Rule
ATM VCEP v1.5 PM5: 'Apply to frameshifting or truncating variants with premature termination codons upstream of p.Arg3047. Apply to splice variants as with premature termination codons upstream of p.Arg3047 where PVS1_VS(RNA) is applied based on high quality observed splicing impact and must be NMD prone.'
Evidence considered
cspec
pm5_candidates
Still missing
If a generic-ACMG same-residue PM5 evaluation is ever required outside the ATM VCEP, a dedicated ClinVar/LOVD search for P/LP missense variants at codon 280 (e.g., p.Ile280Leu, p.Ile280Phe, p.Ile280Thr) would need to be run; none was performed because the VCEP rule does not call for it.
PM6VCEPNot applicable: the ATM expert panel does not use presumed-de-novo evidence for this recessive condition.
Details ▾Hide ▴
Reasoning
The ClinGen HBOP ATM VCEP v1.5 specification (cspec doc 639508985) explicitly declares PM6 'Not Applicable' with no strength rules. Because an applicable VCEP specification exists and excludes this criterion, generic ACMG/AMP PM6 (presumed de novo without confirmation) is not applied. Consistent with PS2: ATM-related disease here is recessive, and de novo evidence is not weighted under this specification. No parental-testing or de novo evidence exists in the case bundle regardless.
PP2VCEPNot applicable: the ATM expert panel does not use the missense-mechanism criterion for this gene.
Details ▾Hide ▴
Reasoning
The governing ClinGen HBOP ATM VCEP v1.5 (CSPEC, doc 639508985) explicitly marks PP2 as 'Not Applicable' for ATM. The VCEP does not use the generic ACMG/AMP PP2 (low rate of benign missense variation / missense mechanism) criterion for this gene; per the user's standing policy, the ClinGen CSPEC takes precedence over generic ACMG/AMP when a specification exists. No PP2 evidence evaluation is performed.
VCEP rule
Not Applicable - ATM VCEP v1.5 (CSPEC) marks PP2 as Not Applicable.
PP4VCEPNot applicable: the ATM panel routes phenotype evidence through its PM3/BP2 points system instead of this criterion.
Details ▾Hide ▴
Reasoning
The governing ClinGen HBOP ATM VCEP v1.5 specification (cspec) explicitly marks PP4 as 'Not Applicable'. The VCEP channels phenotype-specificity evidence for Ataxia Telangiectasia through the PM3/BP2 phenotype points table (confident/consistent A-T diagnosis) rather than through PP4, and A-T is a recessive disorder for which this expert panel does not use PP4. Generic ACMG/AMP PP4 (phenotype or family history highly specific for a single-etiology disease) is therefore superseded and not evaluated. No phenotype-specificity evidence for this exact variant was identified in the bundle in any case.
Evidence considered
cspec ATM VCEP v1.5 PP4 applicability = Not Applicable
PP5VCEPNot applicable: ClinVar labels this variant uncertain significance with no expert-panel pathogenic classification.
Details ▾Hide ▴
Reasoning
Two independent grounds. (1) The governing ClinGen HBOP ATM VCEP v1.5 specification (cspec) marks PP5 as 'Not Applicable'. (2) The global PP5 rule applied by this pipeline requires an exact-variant ClinVar expert-panel classification of Pathogenic/Likely pathogenic; ClinVar record 407462 for c.838A>G (p.Ile280Val) is 'Uncertain significance' (2 stars, 'criteria provided, multiple submitters, no conflicts'), with 6 clinical-laboratory submissions, 0 expert panel submissions, and no expert-panel classification anywhere in the submission audit (expert_panel_submissions: 0). Ordinary laboratory submissions and aggregate ClinVar labels never trigger PP5. No exact-variant expert-panel P/LP classification exists, so PP5 is not applicable.
Evidence considered
cspec ATM VCEP v1.5 PP5 applicability = Not Applicable
BS2VCEPNot applicable: the ATM panel captures unaffected-carrier observations through its PM3/BP2 points system instead.
Details ▾Hide ▴
Reasoning
BS2 is marked Not Applicable in the ClinGen HBOP ATM VCEP v1.5 specification (cspec 639508985). For ATM, observations of the variant in unaffected individuals carrying a P/LP variant in trans (or homozygously) are instead captured by the VCEP's points-based PM3/BP2 table, which assigns negative points for unaffected (non-A-T) carriers. Therefore BS2 is not assessed as an independent criterion under this framework; no unaffected-individual observation data were required. Generic ACMG BS2 was not applied because an applicable gene-specific specification exists.
VCEP rule
Not Applicable per ClinGen HBOP ATM VCEP v1.5 (applicability: Not Applicable); unaffected-carrier observations handled by PM3/BP2 points table
BS4VCEPNot applicable: the ATM panel excludes lack-of-segregation evidence for this condition, and no family genotyping data existed.
Details ▾Hide ▴
Reasoning
The ClinGen HBOP ATM VCEP v1.5 specification (cspec doc 639508985) declares BS4 'Not Applicable' and instructs: for the AD condition, co-segregation analysis in low-penetrance genes can lead to false positives (PMID 32773770); for the AR condition, informative instances of lack of co-segregation in A-T families are too rare to be weighted at this time, and biallelic unaffected patients in an A-T family should instead be considered under BP2. Because an applicable VCEP specification exists and explicitly excludes BS4, generic ACMG/AMP BS4 is not applied. The case bundle additionally contains no family genotyping data (no unaffected relatives tested, no non-segregation observations), so BS4 could not have been met under any framework.
VCEP rule
ClinGen HBOP ATM VCEP v1.5 (cspec): BS4 applicability = 'Not Applicable'; AD condition not used (PMID 32773770); AR lack-of-co-segregation in A-T families too rare to weight, routed to BP2 instead.
BP1VCEPNot applicable: the ATM expert panel does not use the truncating-only-disease criterion for this gene.
Details ▾Hide ▴
Reasoning
The governing ClinGen HBOP ATM VCEP v1.5 (CSPEC, doc 639508985) explicitly marks BP1 as 'Not Applicable' for ATM. The VCEP does not use the generic ACMG/AMP BP1 (missense variant in a gene where primarily truncating variants cause disease) criterion for this gene; ATM disease mechanism (LOF via truncation and missense both recognized) is governed by the VCEP's own PVS1/PS1/PM3 machinery instead. Per the user's standing policy, the ClinGen CSPEC takes precedence over generic ACMG/AMP when a specification exists. No BP1 evidence evaluation is performed.
VCEP rule
Not Applicable - ATM VCEP v1.5 (CSPEC) marks BP1 as Not Applicable.
BP3VCEPNot applicable: the variant is a missense change with no insertion, deletion, or repeat-region context.
Details ▾Hide ▴
Reasoning
The governing ATM VCEP v1.5 (cspec) explicitly marks BP3 as 'Not Applicable' in its criteria applicability table. Independently, the generic BP3 definition (in-frame insertions/deletions in a repetitive region without a known function) does not apply: the variant is a missense SNV (c.838A>G / p.Ile280Val) with no insertion, deletion, or protein length change, and no repeat-region context is identified anywhere in the case evidence bundle (no repeat annotation for this position in prefetch/evidence data). BP3 is therefore not_applicable.
BP5VCEPNot applicable: the ATM expert panel does not use the alternate-molecular-basis criterion.
Details ▾Hide ▴
Reasoning
The governing ClinGen HBOP ATM VCEP v1.5 specification (cspec) explicitly marks BP5 as 'Not Applicable'. Generic ACMG/AMP BP5 (variant found in a case with an alternate molecular basis for disease) is superseded by the VCEP specification and is not evaluated. No evidence of an alternate molecular basis of disease in carriers of this variant was identified in the bundle.
Evidence considered
cspec ATM VCEP v1.5 BP5 applicability = Not Applicable
BP6VCEPNot applicable: ClinVar labels this variant uncertain significance with no expert-panel benign classification.
Details ▾Hide ▴
Reasoning
Two independent grounds. (1) The governing ClinGen HBOP ATM VCEP v1.5 specification (cspec) marks BP6 as 'Not Applicable'. (2) The global BP6 rule applied by this pipeline requires an exact-variant ClinVar expert-panel classification of Benign/Likely benign; ClinVar record 407462 for c.838A>G (p.Ile280Val) is 'Uncertain significance' (2 stars) with 0 expert panel submissions and no benign/likely-benign expert-panel assertion anywhere in the submission audit. Ordinary laboratory submissions and aggregate ClinVar labels never trigger BP6. No exact-variant expert-panel B/LB classification exists, so BP6 is not applicable.
Evidence considered
cspec ATM VCEP v1.5 BP6 applicability = Not Applicable
BP7N/ANot applicable: this rule covers synonymous and deep-intronic variants, and p.Ile280Val is a missense exonic change.
Details ▾Hide ▴
Reasoning
ATM HBOP VCEP v1.5 BP7 (Supporting) applies to synonymous variants and deep intronic variants defined as further than (but not including) +7 and further than (but not including) -21 at donor and acceptor sites, respectively; the BP7(RNA) sub-path requires an observed lack of aberrant RNA defect for silent substitutions and intronic variants. NM_000051.4:c.838A>G is a missense exonic variant (p.Ile280Val, exon 7, position 838 well inside the exon, ~175 nt from the exon 7 5' boundary and ~63 nt from its 3' boundary) - it is neither synonymous nor intronic, so the computational BP7 sub-path is out of scope. No RNA assay data for this variant is present in the case evidence (ClinVar submissions are VUS without functional/RNA evidence; none of the full-text publications in the bundle mention c.838A>G), so the BP7(RNA) sub-path cannot be applied either. BP7 is therefore not applicable for this variant; the benign computational signal is fully captured by BP4 (REVEL 0.043).
Rule
ATM HBOP VCEP v1.5 BP7 (Supporting): Use for synonymous and deep intronic variants defined as further than (but not including) +7 and further than (but not including) -21 at donor and acceptor sites, respectively. BP7(RNA): observed lack of aberrant RNA defect for silent substitutions and intronic variants.
Evidence considered
Variant is missense (p.Ile280Val), exonic (exon 7) - outside BP7 scope (synonymous and deep intronic variants only)
No RNA/splicing functional study on c.838A>G in the case evidence bundle
Still missing
No RNA functional data (BP7(RNA) sub-path) available for this variant; not required for adjudication since BP7's scope (synonymous/deep intronic) excludes this missense exonic variant
AF 1.9e-06 · 3 / 1,613,222 alleles · 0 hom · FAF 5.5e-06
Admixed American2 / 60,006
3.3e-05
European (non-Finnish)1 / 1,179,796
8.5e-07
0 alternate alleles recorded in 8 other groups: Remaining individuals, European (Finnish), Amish, East Asian, Middle Eastern, South Asian, Ashkenazi Jewish, African/African American.
gnomAD v2.1
AF 4e-06 · 1 / 250,510 alleles · 0 hom
Admixed American1 / 34,522
2.9e-05
0 alternate alleles recorded in 7 other groups: African/African American, Ashkenazi Jewish, East Asian, European (Finnish), European (non-Finnish), Remaining individuals, South Asian.
Statements
gnomAD v4.1
This variant is present in gnomAD v4.1 (AF= 1.85963e-06; MAF= 0.00019%, 3/1613222 alleles, homozygotes = 0) and has highest observed frequency in the Admixed American population (AF= 3.333e-05; MAF= 0.00333%, 2/60006 alleles, homozygotes = 0); grpmax FAF= 5.53e-06.
gnomAD v2.1
This variant is present in gnomAD v2.1 (AF= 3.99186e-06; MAF= 0.00040%, 1/250510 alleles, homozygotes = 0) and has highest observed frequency in the Admixed American population (AF= 2.8967e-05; MAF= 0.00290%, 1/34522 alleles, homozygotes = 0).
gnomAD-Canada
Absent from gnomAD-Canada v1.0.
SpliceAI and Pangolin
SpliceAI acceptor lossat +43 bp
0.000
SpliceAI acceptor gainat -28 bp
0.010
SpliceAI donor lossat -88 bp
0.000
SpliceAI donor gainat +63 bp
0.000
Pangolin splice lossmagnitude shown
-0.000
Pangolin splice gain
0.010
Other predictors
REVEL
0.043
BayesDel
-0.435
How the predictions were used
BP4SupportingMet (supporting): REVEL 0.043 is well below the 0.249 missense threshold; SpliceAI 0.01 also rules out splice impact.
Details ▾Hide ▴
Reasoning
ATM HBOP VCEP v1.5 BP4 (Supporting) requires (missense path) REVEL <= 0.249 OR (splicing path) SpliceAI <= 0.1. Per the lab's BP4 missense/splicing policy override recorded in vcep_materials.json, exactly one mechanism is evaluated for a missense variant and only the missense sub-path (REVEL <= 0.249) counts; the SpliceAI sub-path is NOT applied as an independent BP4 line for missense variants. REVEL = 0.043 <= 0.249, so BP4 is met via the missense sub-path. Note: the override does not change the outcome here - under the VCEP's literal text, SpliceAI max delta 0.01 (<= 0.1) would also satisfy the splicing sub-path, so BP4 would be met either way; SpliceAI is therefore recorded as ruling out splice impact but is not double-counted as a second independent BP4 line. Corroboration: Sun et al. 2025 Suppl Table S1 row for c.838A>G shows REVEL 0.043 (matches pipeline lookup), AlphaMissense 0.0853, Combined score -1.10 with Classification 'Intermediate' (not Non-functional) and DeepATM_predicted=No (directly measured); BayesDel (not part of VCEP rules) = -0.43543, benign-leaning. All computational evidence is concordant for a benign-leaning missense effect.
Rule
ATM HBOP VCEP v1.5 BP4 (Supporting): Missense: REVEL score <=0.249; Splicing: No predicted impact via splicing (SpliceAI <=0.1). Lab override: for missense variants use ONLY the missense sub-path (REVEL), never both sub-paths independently.
PP3Not met: REVEL 0.043 is far below the 0.7333 threshold, and SpliceAI 0.01 is far below 0.2.
Details ▾Hide ▴
Reasoning
ATM HBOP VCEP v1.5 PP3 requires (missense path) REVEL > 0.7333 OR (splicing path) SpliceAI >= 0.2 for silent/missense/in-frame and intronic variants outside donor/acceptor +/-1,2. Both sub-paths were assessed for this missense variant (NM_000051.4:c.838A>G, p.Ile280Val, exon 7): REVEL = 0.043 (well below 0.7333) and SpliceAI max delta = 0.01 (well below 0.2; DS_AG 0.01, all other deltas 0.0). Neither threshold is met. Corroborating predictors in Sun et al. 2025 Suppl Table S1 (AlphaMissense 0.0853, EVE 0.205, boostDM 0.121, CADD 16.54) are all low/benign-leaning and do not support a damaging effect. PP3 is therefore not met with high confidence.
Rule
ATM HBOP VCEP v1.5 PP3 (Supporting): Missense: REVEL >0.7333; Splicing: SpliceAI >=0.2 for silent, missense/in-frame and intronic variants outside donor and acceptor 1,2 sites.
BP7N/ANot applicable: this rule covers synonymous and deep-intronic variants, and p.Ile280Val is a missense exonic change.
Details ▾Hide ▴
Reasoning
ATM HBOP VCEP v1.5 BP7 (Supporting) applies to synonymous variants and deep intronic variants defined as further than (but not including) +7 and further than (but not including) -21 at donor and acceptor sites, respectively; the BP7(RNA) sub-path requires an observed lack of aberrant RNA defect for silent substitutions and intronic variants. NM_000051.4:c.838A>G is a missense exonic variant (p.Ile280Val, exon 7, position 838 well inside the exon, ~175 nt from the exon 7 5' boundary and ~63 nt from its 3' boundary) - it is neither synonymous nor intronic, so the computational BP7 sub-path is out of scope. No RNA assay data for this variant is present in the case evidence (ClinVar submissions are VUS without functional/RNA evidence; none of the full-text publications in the bundle mention c.838A>G), so the BP7(RNA) sub-path cannot be applied either. BP7 is therefore not applicable for this variant; the benign computational signal is fully captured by BP4 (REVEL 0.043).
Rule
ATM HBOP VCEP v1.5 BP7 (Supporting): Use for synonymous and deep intronic variants defined as further than (but not including) +7 and further than (but not including) -21 at donor and acceptor sites, respectively. BP7(RNA): observed lack of aberrant RNA defect for silent substitutions and intronic variants.
Evidence considered
Variant is missense (p.Ile280Val), exonic (exon 7) - outside BP7 scope (synonymous and deep intronic variants only)
No RNA/splicing functional study on c.838A>G in the case evidence bundle
Still missing
No RNA functional data (BP7(RNA) sub-path) available for this variant; not required for adjudication since BP7's scope (synonymous/deep intronic) excludes this missense exonic variant
Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology.
Richards S et al. · 2015
ACMG/AMP 2015 standards and guidelines for sequence variant interpretation. Defines the generic PS1, PM1, PM5, PP2, and BP1 criteria that this group's assessment is based on conceptually, but the ATM VCEP v1.5 specification supersedes the generic definitions for PM1/PP2/BP1 (Not Applicable) and PM5 (repurposed as truncation-cutoff rule). The paper does not mention c.838A>G, p.Ile280Val, or codon 280 of ATM in any form.
While the previous ACMG recommendations provided interpretative categories of sequence variants and an algorithm for interpretation, the recommendations did not provide defined terms or detailed variant classification guidance. This report describes updated standards and guidelines for classification of sequence variants using criteria informed by expert opinion and the available evidence.
Full text searched for variant mentions; no occurrence of c.838A>G / Ile280 / I280 / p.Ile280Val found
Exon splicing analysis of intronic variants in multigene cancer panel testing for hereditary breast/ovarian cancer.
Ryu JS et al. · 2020
Ryu et al. 2020, exon splicing analysis of intronic variants in multigene cancer panel testing for hereditary breast/ovarian cancer. Reports P/LP variants in a 23-gene HBOC panel (ATM P/LP frequency 0.4%) and RT-PCR splicing analyses of selected intronic variants. The paper does not mention c.838A>G, p.Ile280Val, or codon 280 of ATM; it provides no same-residue or same-amino-acid comparator relevant to PS1/PM5 for this variant.
Full text searched for variant mentions; no occurrence of c.838A>G / Ile280 / I280 / p.Ile280Val found
Germline ATM variants predispose to melanoma: a joint analysis across the GenoMEL and MelaNostrum consortia.
Dalmasso B et al. · 2021
Dalmasso et al. 2021, germline ATM variants and melanoma risk across GenoMEL/MelaNostrum consortia (2,104 cases). Enrichment analysis of ATM LOF and VUS variants versus gnomAD NFE. The paper does not mention c.838A>G, p.Ile280Val, or codon 280 of ATM; it provides no variant-specific evidence for this residue.
EFNS/ENS Consensus on the diagnosis and management of chronic ataxias in adulthood.
van de Warrenburg BP et al. · 2014
van de Warrenburg et al. 2014, EFNS/ENS consensus on diagnosis and management of chronic ataxias in adulthood. General clinical guideline mentioning ataxia telangiectasia among recessive ataxias; no variant-level data. The paper does not mention c.838A>G, p.Ile280Val, or codon 280 of ATM and is not a source of same-residue evidence for this group.
A practice guideline from the American College of Medical Genetics and Genomics and the National Society of Genetic Counselors: referral indications for cancer predisposition assessment.
Hampel H et al. · 2015
Hampel et al. 2015, ACMG/NSGC practice guideline on referral indications for cancer predisposition assessment. General referral guideline; no variant-level data. The paper does not mention c.838A>G, p.Ile280Val, or codon 280 of ATM and is not a source of same-residue evidence for this group.
The evidence statements recorded by the interpretation pipeline. They have not been reinterpreted by the page.
gnomAD v4.1
This variant is present in gnomAD v4.1 (AF= 1.85963e-06; MAF= 0.00019%, 3/1613222 alleles, homozygotes = 0) and has highest observed frequency in the Admixed American population (AF= 3.333e-05; MAF= 0.00333%, 2/60006 alleles, homozygotes = 0); grpmax FAF= 5.53e-06.
gnomAD v2.1
This variant is present in gnomAD v2.1 (AF= 3.99186e-06; MAF= 0.00040%, 1/250510 alleles, homozygotes = 0) and has highest observed frequency in the Admixed American population (AF= 2.8967e-05; MAF= 0.00290%, 1/34522 alleles, homozygotes = 0).
gnomAD-Canada
Absent from gnomAD-Canada v1.0.
ClinVar
This variant has been reported in ClinVar as Uncertain significance (6 clinical laboratories). (ClinVarID = 407462)
In silico
SpliceAI predicts no significant splice impact for this variant (max delta score = 0.01). REVEL score = 0.043. BayesDel score = -0.43543.
Functional
OncoKB did not identify variant-specific reviewed functional evidence for this variant; gene-level curated context is available for reviewer follow-up. ATM, a kinase involved in the DNA damage response, is mutated in various solid and hematologic malignancies.
Somatic
This variant does not lie in a statistically significant hotspot. This variant has not previously been reported in somatic cancers (COSMIC).
Cancer Hotspots
This variant does not lie in a statistically significant hotspot.
This variant was interpreted by pipeline 7.1.0.
The current version is 11.0.07.
Re-running queues a fresh interpretation; the result replaces this page when complete.
Re-runs are not free. Blank draws on today's public interpretation allowance; with a code, it uses one of that code's uses.