ATM c.1039G>T, p.(Glu347Ter)

Reasoning

ATM c.1039G>T is a nonsense variant, p.(Glu347Ter), in exon 8 of the MANE Select NM_000051.4 transcript. It introduces a premature termination codon at residue 347 of the 3057-residue reference protein, far upstream of the final coding exon, and is therefore predicted to undergo nonsense-mediated decay rather than produce a distal truncated protein. The ATM HBOP VCEP specifies PVS1 for null variants in ATM, identifies NM_000051.3 as the default transcript with constitutive exons, and cautions only for extreme 3′ LOF variants or splice variants with in-frame exon skipping; neither caveat applies. Full-strength PVS1 is appropriate.

Rule

ATM HBOP VCEP v1.5 PVS1 decision guide: apply PVS1 to a null variant in ATM when loss of function is a known disease mechanism; use caution only at the extreme 3′ end or for splice variants predicted to cause in-frame exon skipping with the remainder of the protein intact.

Evidence considered

• Variant normalization identifies NM_000051.4:c.1039G>T as NP_000042.3:p.(Glu347Ter), a nonsense change in exon 8; the reference protein is 3057 residues and the predicted protein terminates at residue 347.
• The ATM HBOP VCEP PVS1 guide defines PVS1 for nonsense, frameshift, canonical splice, initiation-codon, and exon-deletion null variants in ATM, states that the default NM_000051.3 transcript exons are constitutive, and limits its stated cautions to extreme 3′ LOF variants and splice variants predicted to skip an exon in-frame while leaving the remainder intact.
• The gene-level PVS1 eligibility assessment identifies ATM loss of function as an established germline disease mechanism under the applicable VCEP framework.

Sources

Reasoning

The ATM VCEP v1.5 PM2 Supporting threshold is frequency <=0.001% in gnomAD v4. The aggregate gnomAD v4 allele frequency is 1.858980237800752e-06 (0.000185898%), which meets this threshold, so PM2 Supporting is met. However, the maximum observed subpopulation frequency is 4.998833605492052e-05 (0.0049988%) in Admixed American individuals, above the stated PM2 cutoff; this overall-versus-subpopulation threshold discrepancy warrants human review rather than being silently resolved.

Rule

Frequency <=0.001% in gnomAD v4 dataset; if n=1 in a single subpopulation, PM2 Supporting applies

Evidence considered

• ATM VCEP v1.5 specifies PM2 Supporting at frequency <=0.001% in the gnomAD v4 dataset and notes that n=1 in a single subpopulation is sufficient for PM2 Supporting.
• gnomAD v4 reports total AF = 1.858980237800752e-06 (0.000185898%), based on 3/1,613,788 alleles, with 0 homozygotes.
• The highest observed subpopulation is Admixed American: 3/60,014 alleles, AF = 4.998833605492052e-05 (0.0049988%), with 0 homozygotes; grpmax FAF = 1.327e-05.

Still missing

• The VCEP PM2 wording states a frequency threshold but does not explicitly state whether the aggregate AF or the maximum subpopulation AF must control when they differ.

Sources

Reasoning

The ATM VCEP v1.5 PM5 specification explicitly permits PM5 at Supporting strength for frameshifting or truncating variants with premature termination codons upstream of p.Arg3047. NM_000051.4:c.1039G>T produces the premature termination codon p.Glu347Ter, which is upstream of p.Arg3047; the criterion is therefore met at Supporting strength.

Rule

ATM VCEP v1.5 PM5: apply to frameshifting or truncating variants with premature termination codons upstream of p.Arg3047; Supporting strength is specified for this pathway.

Evidence considered

• The case normalization identifies NM_000051.4:c.1039G>T as the truncating variant NP_000042.3:p.(Glu347Ter), at residue 347.
• The ATM VCEP v1.5 PM5 rule states: apply to frameshifting or truncating variants with premature termination codons upstream of p.Arg3047.

Sources

Reasoning

The ATM VCEP v1.5 BA1 threshold is grpmax Filtering AF >0.5% in gnomAD v4. The observed grpmax FAF is 1.327e-05 (0.001327%), far below 0.5%; therefore BA1 is not met.

Rule

Grpmax Filtering AF >0.5% in gnomAD v4 dataset

Evidence considered

• ATM VCEP v1.5 specifies BA1 when grpmax Filtering AF is >0.5% in gnomAD v4.
• gnomAD v4 reports grpmax FAF = 1.327e-05, with total AF = 1.858980237800752e-06 and 3/1,613,788 alleles observed; the maximum subpopulation frequency is 4.998833605492052e-05 in Admixed American individuals.

Sources

Reasoning

The ATM VCEP v1.5 BS1 threshold is grpmax Filtering AF >0.05% in gnomAD v4. The observed grpmax FAF is 1.327e-05 (0.001327%), below 0.05%; therefore BS1 is not met.

Rule

Grpmax Filtering AF >0.05% in gnomAD v4 dataset

Evidence considered

• ATM VCEP v1.5 specifies BS1 when grpmax Filtering AF is >0.05% in gnomAD v4.
• gnomAD v4 reports grpmax FAF = 1.327e-05 (0.001327%); the highest observed subpopulation AF is 4.998833605492052e-05 (0.00500%) in Admixed American individuals, also below the BS1 threshold.

Sources

Reasoning

The ATM VCEP permits splice-impact assessment across variant types and assigns BP4_Supporting only when SpliceAI is <=0.1. The observed maximum delta score is 0.307, so the no-predicted-splice-impact requirement is not met. The laboratory missense-only BP4 override does not change this outcome because this is a nonsense, not missense, variant.

Rule

ATM VCEP v1.5 BP4 splicing sub-path: SpliceAI <=0.1.

Evidence considered

• The governing ATM VCEP specification uses SpliceAI <=0.1 for BP4 splicing evidence and states that splice analysis should be considered for all variant types.
• SpliceAI reports DS_AL 0.243 and DS_DL 0.307, with maximum delta score 0.307; this exceeds the BP4 no-impact threshold.

Sources

Reasoning

The ClinGen HBOP ATM VCEP v1.5 framework states that PS3 must not be used at Strong strength; PS3_Moderate requires failure to rescue both an ATM-specific feature and radiosensitivity, while PS3_Supporting requires failure of an ATM-specific feature and cannot be based on radiosensitivity alone. The variant NM_000051.4:c.1039G>T (p.Glu347Ter) has a directly measured non-functional result in Sun et al. 2025 Table S1 (DeepATM_predicted=No, Confidence=High), indicating genuine prime-editing functional-screen evidence rather than a computational prediction. However, this assay is not one of the three ATM functional assays calibrated in the current VCEP specification, and the available evidence does not document the required ATM-specific phosphorylation and radiosensitivity readouts for this variant. Therefore PS3 cannot be assigned under the current VCEP without expert review of assay validity and calibration.

Rule

ClinGen HBOP ATM VCEP v1.5 PS3: do not use Strong; Moderate requires failure to rescue both an ATM-specific feature and radiosensitivity; Supporting requires failure to rescue an ATM-specific feature only and is not assigned for radiosensitivity alone.

Evidence considered

• ClinGen HBOP ATM VCEP v1.5 specifies that PS3 is not used as Strong; PS3_Moderate requires failure to rescue both an ATM-specific feature and radiosensitivity, and PS3_Supporting requires failure of an ATM-specific feature and excludes radiosensitivity-only evidence.
• Sun et al. 2025 Table S1 directly matches NM_000051.4:c.1039G>T / p.Glu347Ter and reports Classification=Non-functional, Confidence=High, and DeepATM_predicted=No; the latter indicates a direct prime-editing functional measurement rather than a model extrapolation.
• The ATM VCEP functional-assay calibration document identifies its approved assays as the Mitui 2009, Barone 2009, and Scott 2002 assays and explicitly indicates that calibration is not per-variant evidence.

Still missing

• No variant-specific result from one of the VCEP-approved ATM kinase-activity assays (PMID:18634022, PMID:19431188, or PMID:11805335) was identified.
• The available Sun et al. result does not provide the VCEP-required paired ATM-specific feature and radiosensitivity readouts needed for PS3_Moderate, nor a VCEP-calibrated strength for this newer assay.
• Human review is required to determine whether the direct prime-editing assay can be accepted and what PS3 strength, if any, is appropriate.

Sources

Reasoning

No qualifying case-control study was provided for ATM c.1039G>T (p.Glu347Ter).

Rule

ATM HBOP VCEP v1.5: Case-control studies; p-value ≤0.05 AND (odds ratio, hazard ratio, or relative risk ≥2 OR lower 95% CI ≥1.5). PS4_Moderate is not to be used.

Evidence considered

• The ATM HBOP VCEP v1.5 permits PS4 only for case-control studies with p-value ≤0.05 and OR, HR, or RR ≥2, or lower 95% CI ≥1.5; it explicitly does not permit PS4_Moderate based on proband counting.
• The exact-variant ClinVar submissions contain no validated publication trail or criterion lead, and the pre-extracted candidate paper PMID:17508274 does not mention ATM or c.1039G>T.

Still missing

• A case-control study for this exact variant reporting p-value and an odds ratio, hazard ratio, or relative risk meeting the ATM VCEP PS4 threshold is needed.

Sources

Reasoning

The applicable ClinGen HBOP ATM VCEP v1.5 specification evaluates PM3 for the autosomal-recessive ataxia-telangiectasia phenotype using weighted observations in unrelated affected probands, with the score determined by phenotype confidence, zygosity, and phase with a VCEP-classified pathogenic or likely pathogenic ATM variant. The case-evidence bundle contains no affected ataxia-telangiectasia proband observation for NM_000051.4:c.1039G>T, no documented second ATM variant, and no parental or offspring testing establishing trans or cis phase. Therefore, no PM3 points can be assigned.

Rule

ClinGen HBOP ATM VCEP v1.5: PM3 points are assigned 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; consistent phenotype values are half these values; homozygous observations are capped at two individuals). PM3 Supporting = 1 point, Moderate = 2 points, Strong = 4 points, Very Strong = at least 8 points.

Evidence considered

• ClinGen HBOP ATM VCEP v1.5 specifies PM3 as a points-based criterion for ataxia-telangiectasia and requires assessment of affected-proband phenotype, zygosity, and phase with a VCEP-classified pathogenic or likely pathogenic ATM variant.
• Review of the consolidated case-evidence bundle found no qualifying affected-proband, second-variant, parental, offspring, trans-phase, or cis-phase observation for NM_000051.4:c.1039G>T (NP_000042.3:p.(Glu347Ter)).
• The available ClinVar classification and general ATM loss-of-function information do not establish PM3 because PM3 requires allelic observations in affected individuals rather than a final classification from another source.

Still missing

• Affected-proband observations for the variant in individuals with ataxia-telangiectasia or an atypical/variant ataxia-telangiectasia phenotype are not documented.
• A second ATM variant classified under the ATM VCEP specification is not documented.
• Phase information from parental or offspring testing is not documented.
• The phenotype confidence category and any qualifying laboratory results are not documented.

Sources

Reasoning

PP1 is applicable only through the ATM VCEP autosomal-recessive segregation pathway, which requires affected relatives to have both variants identified in the proband. The case-evidence bundle contains no usable family pedigree, affected-relative genotypes, parental/offspring testing, phase information, or meioses establishing segregation of NM_000051.4:c.1039G>T. Therefore no PP1 strength can be assigned.

Rule

AR condition only: affected relatives must have both variants identified in the proband; Supporting for 1 affected relative, Moderate for 2, Strong for at least 3 affected relatives.

Evidence considered

• The ATM VCEP PP1 rule permits Supporting for segregation in 1 affected relative, Moderate for 2, and Strong for at least 3 affected relatives under the autosomal-recessive condition.
• The ATM VCEP instructions require affected relatives to have both variants identified in the proband.
• The consolidated case-evidence bundle provides no case-specific segregation, parental testing, affected-relative genotype, pedigree, or meiosis data for this variant.

Still missing

• Documented affected relatives with phenotype consistent with the relevant ATM autosomal-recessive condition
• Genotype results for the variant in each affected relative
• Confirmation that affected relatives carry both variants identified in the proband
• Pedigree and independent meioses sufficient to count segregation

Sources

Reasoning

The ClinGen HBOP ATM VCEP v1.5 framework defines BS3_Supporting when a variant rescues either an ATM-specific feature or radiosensitivity, and BS3_Moderate when it rescues both. The available direct functional result for NM_000051.4:c.1039G>T (p.Glu347Ter) is classified as non-functional, not functional or rescuing, in Sun et al. 2025 Table S1. Thus it does not support BS3. Because no variant-specific normal-function result from a VCEP-approved assay was identified, and the newer direct assay is not calibrated by the current VCEP, BS3 is not assessed rather than assigned as met or not_met on the basis of an unvalidated assay framework.

Rule

ClinGen HBOP ATM VCEP v1.5 BS3: Supporting requires rescue of either an ATM-specific feature or radiosensitivity; Moderate requires rescue of both.

Evidence considered

• ClinGen HBOP ATM VCEP v1.5 specifies BS3_Moderate for rescue of both an ATM-specific feature and radiosensitivity, and BS3_Supporting for rescue of either one.
• Sun et al. 2025 Table S1 directly matches NM_000051.4:c.1039G>T / p.Glu347Ter and reports Classification=Non-functional, Confidence=High, and DeepATM_predicted=No; this is evidence against normal function, not BS3 evidence.
• The ATM VCEP functional-assay calibration document lists the approved assay systems but does not report a variant-specific result for c.1039G>T.

Still missing

• No variant-specific normal-function or rescue result from a VCEP-approved ATM kinase-activity or radiosensitivity assay was identified.
• The available direct prime-editing result is non-functional and therefore does not provide benign functional evidence.
• Human review is required to determine whether any additional validated functional data exist and whether the newer assay can be incorporated under the VCEP framework.

Sources

Reasoning

The applicable ClinGen HBOP ATM VCEP v1.5 specification evaluates BP2 using negative points for an unaffected, non-ataxia-telangiectasia individual carrying the variant with a VCEP-classified pathogenic or likely pathogenic ATM variant in trans, with age and clinical status requirements for the relevant downgrade scenario. The case-evidence bundle contains no unaffected adult observation with this variant and a qualifying ATM variant in confirmed or unknown phase, and no valid cis co-occurrence observation. Therefore, no BP2 points can be assigned.

Rule

ClinGen HBOP ATM VCEP v1.5: BP2 points per unaffected non-A-T individual are -4.0 for confirmed trans, -2.0 for phase unknown, and for homozygous observations -2.0 in a laboratory setting or -1.0 in a database setting, with homozygous observations capped at -2.0. BP2 Supporting = -1 point, Moderate = -2 points, Strong = at least -4 points.

Evidence considered

• ClinGen HBOP ATM VCEP v1.5 specifies BP2 as a points-based criterion based on unaffected non-ataxia-telangiectasia individuals carrying the variant with a pathogenic or likely pathogenic ATM variant, with separate weights for confirmed trans, phase unknown, and laboratory/database settings.
• Review of the consolidated case-evidence bundle found no qualifying unaffected adult, co-occurring ATM pathogenic/likely pathogenic variant, trans-phase observation, or cis-phase observation for NM_000051.4:c.1039G>T (NP_000042.3:p.(Glu347Ter)).
• The variant's rare population frequency and reported classifications do not establish BP2 because BP2 requires an unaffected-individual allelic observation.

Still missing

• An unaffected individual aged 18 years or older with no evidence of ataxia-telangiectasia carrying the variant is not documented.
• A co-occurring ATM pathogenic or likely pathogenic variant classified under the ATM VCEP specification is not documented in such an individual.
• Trans phase or cis phase for any co-occurring variant is not documented.
• The number and setting of qualifying unaffected observations are not documented, so the laboratory/database point allocation cannot be determined.

Sources