Five-step scale from benign to pathogenic. This variant is pathogenic. Select a step to read what it means.
This variantScenario
BenignPathogenic
Pathogenic.A pathogenic classification under the recorded framework.
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 (Very Strong): nonsense change p.(Glu450Ter) at ~15% of the protein is predicted to trigger nonsense-mediated decay, upstream of critical ATM domains.
PM2Supporting
PM2 · supporting
Met (Supporting): absent from gnomAD v4.1 and v2.1, meeting the VCEP's 0.001% allele-frequency threshold.
PM5Supporting
PM5 · supporting
Met (Supporting): premature termination codon at residue 450 lies upstream of the VCEP's p.Arg3047 truncating-variant threshold.
Benign criteria applied · 0
No benign criterion marked as applied in the recorded assessment.
Recorded classification: Pathogenic
Pathogenic under the ClinGen HBOP ATM VCEP v1.5 Rule4: one Very Strong (PVS1) plus two Supporting criteria (PM2, PM5).
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
Not recorded
gnomAD v4.1 allele counts were not recorded for this variant.
ACMG/AMP criteria under ClinGen ATM VCEP v1.5, with the data and sources behind each one.
28 criteria with recorded statuses: 3 met · 2 not met · 6 not assessed · 17 not applicable. Open any row for the full reasoning, the rule and the evidence behind it.
PVS1 applied at full (Very Strong) strength: nonsense variant p.(Glu450Ter) in ATM, a gene with VCEP-recognized LOF disease mechanism, positioned well upstream of the last exon and of the most 3' pathogenic residue (p.Arg3047), predicted NMD substrate, disrupting the FATKIN and HEAT-repeat domains; neither VCEP caveat (extreme 3' position; in-frame exon-skip rescue) applies.
PM4 not applicable: ATM VCEP restricts PM4 to stop-loss variants only; this is a stop-gain (nonsense) variant.
BP3 not applicable: ATM VCEP CSPEC explicitly disables BP3 for this gene regardless of variant type.
Governing framework: ClinGen HBOP ATM VCEP v1.5 (cspec), which is used in preference to generic ACMG/AMP for all five criteria in this group.
Variant NM_000051.3:c.1348G>T / NP_000042.3:p.(Glu450Ter) is a nonsense (stop-gain) variant, not missense, which limits applicability of missense-oriented codon/residue/domain criteria under this VCEP.
PM1, PP2, and BP1 are all explicitly flagged 'Not Applicable' by the ATM VCEP v1.5 specification itself, independent of variant type.
PS1's standard pathway is missense-only per the VCEP text, and the splicing sub-path could not be executed for lack of a matched reference P/LP splice variant, so PS1 is not_applicable.
PM5 is met at Supporting strength via the ATM VCEP's own repurposed truncating-variant rule (PTC upstream of p.Arg3047), which is a gene-specific deviation from the generic ACMG PM5 definition.
For this segregation/de novo group, the governing ClinGen HBOP ATM VCEP v1.5 specification was used. PS2, PM6, and BS4 are not applicable under that specification; PP1 is potentially applicable for autosomal recessive disease but is not assessed because no informative family segregation data are present.
ATM VCEP v1.5 (cspec) restricts PS3/BS3 to a defined set of approved functional assays (kinase-activity Western blot readouts from Mitui 2009, Barone 2009, Scott 2002; radiosensitivity assays from Mitui 2009 and Scott 2002, combinable with kinase results only). No evidence in this case's bundle confirms NM_000051.3:c.1348G>T (p.(Glu450Ter)) was tested in any of these approved assays, and no per-variant Sun et al. 2025 (PMID 40580951) prime-editing screen data was extracted for this variant. Both PS3 and BS3 are therefore not_assessed pending retrieval of that per-variant data.
The governing framework is the ClinGen HBOP ATM VCEP v1.5 specification, which is applicable to this ATM variant and marks PP4, PP5, BP5, and BP6 as Not Applicable while providing a defined PS4 case-control rule.
No criterion was inferred from another source's final classification; the ClinVar single-submitter Likely pathogenic assertion was excluded from PP5 and BP6 because it is not an expert-panel assertion.
NM_000051.3:c.1348G>T / NP_000042.3:p.(Glu450Ter) is a nonsense (stop-gain) variant, which falls outside the variant-type scope the ATM VCEP v1.5 specification assigns to the PP3/BP4 missense sub-path (REVEL) and the PP3/BP4/BP7 splicing sub-path (SpliceAI), both of which are explicitly limited to silent, missense/in-frame, and qualifying intronic variants.
The VCEP text also explicitly bars combining a splicing-based PP3 code with a PVS1 code for the same variant, reinforcing that this nonsense variant's computational splicing signal (SpliceAI DS_DG=0.955) belongs to the PVS1/null-variant adjudication path, not to this group's PP3/BP4/BP7 remit.
REVEL is unavailable and BayesDel lacks a citable verified publication/threshold in this pipeline, so the missense sub-path is independently non-actionable even setting aside variant type.
All three criteria in this group (PP3, BP4, BP7) are therefore returned as not_applicable for this variant.
No criteria match.
Met · 3
PVS1Very strongMet (Very Strong): nonsense change p.(Glu450Ter) at ~15% of the protein is predicted to trigger nonsense-mediated decay, upstream of critical ATM domains.
Details ▾Hide ▴
Reasoning
NM_000051.3:c.1348G>T produces a nonsense change, NP_000042.3:p.(Glu450Ter)/p.(E450*), in ATM, a gene for which the ClinGen HBOP ATM VCEP v1.5 CSPEC provides an official PVS1 decision tree and for which germline loss-of-function is an established disease mechanism (Ataxia-Telangiectasia and predisposition to breast/pancreatic/other cancers). p.Glu450 falls at roughly codon 450 of a ~3056-residue protein (~15% into the coding sequence), far upstream of the penultimate exon/last-50-bp region and far upstream of the gene's most 3' pathogenic residue, p.Arg3047, per the VCEP PVS1 guide. The premature termination codon is therefore predicted to trigger nonsense-mediated decay (NMD) rather than escape it, and even if translated, the truncated product would lack the entire HEAT repeat region and the FAT/PI3K/FATC (FATKIN) domains that the VCEP explicitly designates as functionally critical. Neither of the VCEP's PVS1 caveats applies: this is not a variant at the extreme 3' end of the gene, and it is not a splice variant predicted to cause an in-frame exon skip that preserves the remainder of the protein. Per the VCEP PVS1/PVS1(RNA) decision tree, a null (nonsense) variant predicted to undergo NMD, well upstream of the terminal pathogenic residue and disrupting critical domains, is assigned PVS1 at full (Very Strong) strength.
Rule
ATM VCEP v1.5 PVS1 decision tree: null variant (nonsense) with LOF as an established disease mechanism, NMD-predicted (well upstream of p.Arg3047 and the terminal exon), not meeting either 3'-extreme or exon-skip-rescue caveat -> PVS1 (Very Strong).
Evidence considered
Case summary normalization: NM_000051.3:c.1348G>T predicts NP_000042.3:p.(Glu450Ter) / p.(E450*), a nonsense variant.
pvs1_gene_context.json: ATM has an official CSPEC/VCEP PVS1 decision tree and germline loss-of-function is an established disease mechanism for ATM (lof_mechanism_supported=true, pvs1_gene_gate=eligible).
pvs1_variant_assessment.json: consequence_class classified as 'nonsense'; suggested_default_strength 'PVS1' (i.e. full/Very Strong) under the generic PVS1 framework, with the downgrade considerations (NMD escape, non-critical distal region, extreme 3' position) explicitly not triggered here given the early truncation position.
ATM_PVS1_1.5.pdf (VCEP decision guide): null variants (including nonsense) in ATM are assigned PVS1 per the decision tree; caveats are (1) caution for LOF variants at the extreme 3' end of the gene and (2) caution for splice variants causing in-frame exon skipping that leaves the remainder of the protein intact - neither applies to this early nonsense variant.
ATM_PVS1_1.5.pdf: the VCEP identifies p.Arg3047 as the most 3' residue considered pathogenic; p.Glu450 lies far upstream of this position, consistent with NMD-competent status and disruption of the FATKIN and HEAT-repeat domains that the VCEP designates as critical/important for ATM function.
Still missing
Exact constitutive exon number for c.1348 was not resolved in the retrieved case bundle (pvs1_variant_assessment.exon = null); position was inferred from coding coordinate (c.1348 of a ~9168 nt CDS) and protein coordinate (residue 450 of 3056), which is sufficient to place the variant far from the terminal exon/last-50bp NMD-escape zone but a formal exon-map lookup would further corroborate NMD prediction.
PM2SupportingMet (Supporting): absent from gnomAD v4.1 and v2.1, meeting the VCEP's 0.001% allele-frequency threshold.
Details ▾Hide ▴
Reasoning
The ClinGen HBOP ATM VCEP v1.5 permits PM2 only at Supporting strength for a frequency at or below 0.001% in gnomAD v4, including an allele count of one in a single subpopulation. The variant is absent from gnomAD v4.1, with corroborating absence from gnomAD v2.1 and gnomAD-Canada, and therefore meets the PM2 Supporting requirement.
Rule
ATM VCEP v1.5: PM2 Supporting for frequency <=0.001% in gnomAD v4 dataset; absence qualifies.
Evidence considered
The ATM VCEP v1.5 specifies PM2 Supporting for frequency <=0.001% in gnomAD v4; the VCEP notes that n=1 in a single subpopulation is sufficiently rare.
gnomAD v4.1 reports the variant as absent, which is below the VCEP frequency threshold.
The variant is also reported as absent from gnomAD v2.1 and gnomAD-Canada v1.0, providing concordant population-data support.
PM5SupportingMet (Supporting): premature termination codon at residue 450 lies upstream of the VCEP's p.Arg3047 truncating-variant threshold.
Details ▾Hide ▴
Reasoning
The ATM VCEP v1.5 specification repurposes PM5 (default strength Pathogenic Moderate) with a Supporting-strength rule for truncating variants: 'Apply to frameshifting or truncating variants with premature termination codons upstream of p.Arg3047.' NP_000042.3:p.(Glu450Ter) is a nonsense variant creating a premature termination codon at residue 450, which is well upstream of the p.Arg3047 threshold defined by the VCEP (near the extreme 3' end of the 3056-residue ATM protein). This variant therefore satisfies the VCEP's PM5 rule for truncating variants at Supporting strength. Note this is a VCEP-specific repurposing of the PM5 code label distinct from the generic ACMG PM5 definition (different missense change at a residue with an established pathogenic missense change); it is applied here strictly per the governing ATM cspec rule text, not the generic ACMG PM5 mechanism.
Rule
ATM VCEP v1.5 PM5_Supporting: truncating variant with PTC upstream of p.Arg3047
Evidence considered
ATM VCEP v1.5 (cspec) PM5 rule: 'Apply to frameshifting or truncating variants with premature termination codons upstream of p.Arg3047' at Supporting strength.
Variant NP_000042.3:p.(Glu450Ter) creates a premature termination codon at residue 450, upstream of the p.Arg3047 cutoff specified by the VCEP.
BA1Not met: absent from gnomAD v4.1, below the >0.5% BA1 allele-frequency threshold.
Details ▾Hide ▴
Reasoning
The ClinGen HBOP ATM VCEP v1.5 BA1 threshold is group maximum filtering allele frequency greater than 0.5% in gnomAD v4. The variant is absent from gnomAD v4.1, so the BA1 threshold is not met.
Rule
ATM VCEP v1.5: BA1 Stand Alone for grpmax filtering AF >0.5% in gnomAD v4 dataset.
Evidence considered
The ATM VCEP v1.5 specifies BA1 as applicable at a gnomAD v4 group maximum filtering AF >0.5%.
gnomAD v4.1 reports the variant as absent; therefore no allele frequency supports BA1.
BS1Not met: absent from gnomAD v4.1, below the >0.05% BS1 allele-frequency threshold.
Details ▾Hide ▴
Reasoning
The ClinGen HBOP ATM VCEP v1.5 BS1 threshold is group maximum filtering allele frequency greater than 0.05% in gnomAD v4. The variant is absent from gnomAD v4.1, so the BS1 threshold is not met.
Rule
ATM VCEP v1.5: BS1 Strong for grpmax filtering AF >0.05% in gnomAD v4 dataset.
Evidence considered
The ATM VCEP v1.5 specifies BS1 Strong for grpmax filtering AF >0.05% in gnomAD v4 dataset.
gnomAD v4.1 reports the variant as absent; therefore no allele frequency supports BS1.
PS3Not assessed: insufficient evidence - this variant was not confirmed to have been tested in any VCEP-approved functional assay (Mitui 2009, Barone 2009, Scott 2002).
Details ▾Hide ▴
Reasoning
ATM VCEP v1.5 defines PS3/BS3 via a VCEP-approved functional assay 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) confer PS3_Supporting if abnormal, and radiosensitivity assays confer PS3_Moderate but only combinable with an abnormal kinase result. Applying this requires confirming the variant under assessment (NM_000051.3:c.1348G>T, p.(Glu450Ter)) was actually tested in one of these three named papers. The bundle contains no evidence that this variant appears in Mitui 2009, Barone 2009, or Scott 2002, and none of these three papers' full text is present in this case's retrieved literature set to check directly. The pipeline's high-priority literature triage (PMID 27413114, 30348496, 30553448) consists of general ATM reviews/mouse-model mechanistic papers; none of these mention this specific variant by cHGVS or protein change (checked full text for '1348', 'E450', 'Glu450', '450Ter', '450*' - no hits, the '1348' hit in PMID 27413114 is an unrelated page-range citation). Additionally, the Sun et al. 2025 prime-editing/deep-learning saturation screen (PMID 40580951, Suppl_TableS1) covers essentially all ATM SNVs and could plausibly include this stop-gain variant, but no per-variant row data (cHGVS match, DeepATM_predicted flag, Combined_score/Classification) for c.1348G>T was extracted into this evidence bundle, so it cannot be confirmed whether this variant was directly assayed there, and even if it were, per VCEP index guidance that paper is not one of the three VCEP-approved assays and would require flagging needs_human_review rather than being scored directly. No functional assay evidence for this specific variant is available in the bundle to support PS3.
Still missing
No confirmation that NM_000051.3:c.1348G>T / p.(Glu450Ter) was tested in any of the three VCEP-approved kinase-activity or radiosensitivity assay papers (Mitui 2009 PMID 18634022, Barone 2009 PMID 19431188, Scott 2002 PMID 11805335) - full text of these three papers not present in this case's retrieved literature to check directly
No Sun et al. 2025 (PMID 40580951) Table S1 per-variant row (cHGVS match, DeepATM_predicted flag, Combined_score/Classification) was extracted into the evidence bundle for this variant
None of the three high-priority triaged papers (PMID 27413114, 30348496, 30553448) mention this specific variant
PS4Not assessed: no case-control enrichment data (odds ratio, confidence interval, or p-value) for this variant was available.
Details ▾Hide ▴
Reasoning
The ATM VCEP v1.5 PS4 rule requires a case-control study with p-value <= 0.05 and either an odds ratio, hazard ratio, or relative risk >= 2, or a lower 95% confidence interval >= 1.5. The case-evidence bundle contains no qualifying variant-specific case-control enrichment data, counts, effect estimate, confidence interval, or p-value for NM_000051.3:c.1348G>T.
Rule
ATM 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).
Evidence considered
The governing ATM VCEP v1.5 specification defines PS4 as met at Strong when a case-control study has p-value <= 0.05 and either an odds ratio, hazard ratio, or relative risk >= 2, or a lower 95% confidence interval >= 1.5.
The available evidence bundle does not report qualifying case-control enrichment statistics for the exact variant NM_000051.3:c.1348G>T.
Still missing
A qualifying variant-specific case-control study with analyzable case and control counts, p-value, effect estimate, and confidence interval was not identified.
PM3Not assessed: no affected-proband or in-trans evidence with a pathogenic ATM variant was available.
Details ▾Hide ▴
Reasoning
The applicable ATM VCEP PM3 framework requires documented Ataxia-Telangiectasia proband observations and evidence of the variant in trans with a pathogenic or likely pathogenic ATM variant, with points assigned according to phenotype confidence and phase. The case-evidence bundle contains no affected-proband, parental, phase, trans/cis, homozygosity, or compound-heterozygosity observation for NM_000051.3:c.1348G>T; therefore no PM3 points or strength can be assigned.
Rule
ATM VCEP v1.5 PM3: PM3 Supporting = 1 point, Moderate = 2 points, Strong = 4 points, Very Strong = at least 8 points; points require eligible A-T proband evidence under the ATM PM3/BP2 table.
Evidence considered
The ATM VCEP v1.5 PM3 rule is points-based for A-T and distinguishes confirmed trans, phase-unknown, unidentified/VUS second variants, and homozygous observations.
The consolidated case-evidence bundle provides normalization and population/variant records but no affected-proband or allelic-phase observations for this variant.
Still missing
No documented A-T affected proband carrying NM_000051.3:c.1348G>T.
No documented pathogenic/likely pathogenic ATM variant in trans, parental testing, or other phase evidence.
No documented phenotype and laboratory findings permitting confident or consistent A-T classification.
PP1Not assessed: no affected relatives or segregation data were available to support co-segregation.
Details ▾Hide ▴
Reasoning
PP1 is applicable for ATM autosomal recessive disease when an affected relative segregates with both variants identified in the proband. The ATM VCEP v1.5 strength calibration is Supporting for one affected relative, Moderate for two, and Strong for at least three. The case bundle contains no affected relatives, segregation observations, meioses, phase information, or family testing, so no PP1 strength can be assigned.
The ATM VCEP v1.5 PP1 instructions state: "AR Condition: Affected relatives must have both variants identified in proband."
The ATM VCEP v1.5 strength rules specify Supporting for segregation in one affected relative, Moderate for two affected relatives, and Strong for at least three affected relatives.
The consolidated case-evidence bundle contains no segregation or family-testing evidence and therefore cannot establish any informative affected-relative meioses.
Still missing
Affected-relative count is not provided.
Whether affected relatives carry both ATM variants identified in the proband is not provided.
Family genotypes, phase, phenotype consistency, and informative meioses are not provided.
BS3Not assessed: insufficient evidence - this variant was not confirmed to have been tested in any VCEP-approved functional assay.
Details ▾Hide ▴
Reasoning
Same VCEP-approved functional assay framework applies to BS3 as to PS3 (clingen_hbop_atm_supplementary_tables_1_and_2_v1.xlsx: an ATM-specific rescue in kinase-activity or radiosensitivity assays from Mitui 2009/PMID 18634022, Barone 2009/PMID 19431188, or Scott 2002/PMID 11805335 confers BS3_Supporting, or BS3_Moderate if both an ATM-specific feature and radiosensitivity are rescued). No evidence was found that NM_000051.3:c.1348G>T / p.(Glu450Ter) was functionally tested in any of these three papers, and this variant is a nonsense/stop-gain change producing a premature truncation, making a benign functional rescue biologically implausible a priori but this cannot be confirmed or excluded from the available evidence. No Sun et al. 2025 (PMID 40580951) per-variant functional classification was extracted for this variant in the bundle either. Insufficient evidence to assess BS3.
Still missing
No confirmation that this variant was tested (with a normal/rescued readout) in any of the three VCEP-approved assay papers
No Sun et al. 2025 (PMID 40580951) Table S1 per-variant functional classification data available for this variant in the bundle
BP2Not assessed: no unaffected-carrier, trans-phase, or co-occurrence observations were available.
Details ▾Hide ▴
Reasoning
The applicable ATM VCEP BP2 framework requires documented unaffected non-A-T individuals carrying the variant with a pathogenic or likely pathogenic ATM variant in trans, or qualifying homozygous observations, with negative points assigned according to phase and setting. The case-evidence bundle contains no unaffected individual, trans/cis, phase, homozygosity, or co-occurrence observation for NM_000051.3:c.1348G>T; therefore no BP2 points or strength can be assigned.
Rule
ATM VCEP v1.5 BP2: Supporting = -1 point, Moderate = -2 points, Strong = at most -4 points; points require eligible unaffected non-A-T co-occurrence evidence under the ATM PM3/BP2 table.
Evidence considered
The ATM VCEP v1.5 BP2 rule is points-based and requires eligible unaffected non-A-T co-occurrence evidence, with confirmed-trans, phase-unknown, laboratory-setting, database-setting, and homozygous categories.
The consolidated case-evidence bundle provides no unaffected co-occurrence or allelic-phase observations for this variant.
Still missing
No documented unaffected non-A-T individual carrying the variant with a pathogenic/likely pathogenic ATM variant in trans.
No phase or parental testing data establishing trans or cis configuration.
No eligible homozygous unaffected observation and no documented age/clinical assessment relevant to A-T.
Not applicable · 17· recorded as not applicableShow reasons ▾Hide reasons ▴PS1N/ANot applicable: this is a nonsense (stop-gain) variant, not a missense change, so there is no amino acid substitution to compare against a reference variant.
Details ▾Hide ▴
Reasoning
The ATM VCEP v1.5 PS1 specification restricts the standard PS1 pathway to missense changes ('Use for missense changes as long as splicing is ruled-out for both alterations') and offers a separate splicing-region PS1 table for variants assessed on a splice-prediction basis. This variant, NM_000051.3:c.1348G>T / NP_000042.3:p.(Glu450Ter), is a nonsense (stop-gain) substitution, not a missense change, so the standard missense PS1 pathway (same amino-acid change as an established P/LP variant) does not apply - there is no amino acid substitution to compare against a reference variant. The splicing-region PS1 table requires identifying a known P/LP reference variant at the same position with a matching predicted splice event and comparable-or-higher prediction strength; no such reference variant or comparison data is present in the case bundle, so the splicing sub-path cannot be executed even though SpliceAI flags a possible donor-gain effect (DS_DG=0.955) at this position.
Rule
ATM VCEP v1.5 PS1: missense-only standard pathway; splicing sub-path requires a matched reference P/LP variant, none identified.
Evidence considered
ATM VCEP v1.5 PS1 guide restricts the standard missense PS1 pathway to missense changes only, and provides a separate splicing-region table requiring a matching reference P/LP variant at the same position.
Variant is NM_000051.3:c.1348G>T, NP_000042.3:p.(Glu450Ter), a nonsense/stop-gain substitution, not a missense change.
SpliceAI predicts a possible donor-gain effect (DS_DG=0.955) but no reference P/LP splice variant at the same position was available in the bundle to execute the VCEP splicing sub-path.
Still missing
No PS1(Splicing) reference P/LP variant at this position was identified in the case materials to run the VCEP's splicing sub-path comparison.
PS2VCEPNot applicable: the ATM VCEP does not use PS2, since de novo occurrences are not informative for autosomal recessive disease.
Details ▾Hide ▴
Reasoning
The ClinGen HBOP ATM VCEP v1.5 specification marks PS2 as not applicable for ATM-related autosomal dominant and autosomal recessive disease because informative de novo occurrences have not been observed and de novo events in an autosomal recessive condition are not informative for phase. The case bundle also contains no proband-level parental testing or confirmed de novo observation.
Evidence considered
The ATM VCEP v1.5 framework states: "Do not use for AD or AR disease: Informative de novo occurrences have not yet been observed and de novo AR conditions are unlikely to be informed by phase."
The consolidated case-evidence bundle provides variant normalization and database observations but no pedigree, parental genotype, maternity confirmation, paternity confirmation, or de novo case report.
Still missing
No proband-level maternity and paternity confirmation or de novo testing evidence is present, although such evidence would not make PS2 applicable under the governing ATM VCEP specification.
PM1VCEPNot applicable: the ATM VCEP explicitly marks PM1 (mutational hotspot) as not applicable for this gene.
Details ▾Hide ▴
Reasoning
The ATM VCEP v1.5 specification explicitly lists PM1 as 'Not Applicable' for this gene, superseding the generic ACMG PM1 (mutational hotspot/functional domain) criterion.
VCEP rule
ATM VCEP v1.5: PM1 = Not Applicable
Evidence considered
ATM VCEP v1.5 (cspec) lists PM1 applicability as 'Not Applicable'.
PM4N/ANot applicable: PM4 applies only to stop-loss variants, and this is a stop-gain (nonsense) change.
Details ▾Hide ▴
Reasoning
The ATM VCEP v1.5 CSPEC restricts PM4 to stop-loss variants only ('Use for stop-loss variants,' Pathogenic Moderate default). NM_000051.3:c.1348G>T is a nonsense (stop-gain) substitution, NP_000042.3:p.(Glu450Ter), producing a premature termination codon rather than a stop-loss, in-frame indel, or repeat-region expansion/contraction. It therefore does not fall within any category to which the ATM VCEP's PM4 rule applies, and no protein-length-increase or in-frame deletion/insertion scenario exists to evaluate.
Rule
ATM VCEP v1.5 PM4: applies only to stop-loss variants; not applicable to nonsense (stop-gain) variants.
Evidence considered
case_summary.json CSPEC PM4 entry: rule text restricts PM4 to 'stop-loss variants' at Pathogenic Moderate strength; this variant is a stop-gain (nonsense) change, not a stop-loss, in-frame indel, or repeat-region variant, so the rule's condition is not met.
PM6VCEPNot applicable: the ATM VCEP marks PM6 as not applicable, and no de novo occurrence is documented.
Details ▾Hide ▴
Reasoning
The ClinGen HBOP ATM VCEP v1.5 specification marks PM6 as not applicable and instructs that it not be used for autosomal dominant or autosomal recessive disease. No assumed de novo occurrence is documented in the case bundle.
Evidence considered
The ATM VCEP v1.5 framework marks PM6 as not applicable and states: "Do not use for AD or AR disease: Informative de novo occurrences have not yet been observed and de novo AR conditions are unlikely to be informed by phase."
The case bundle contains no parental genotypes, family history, or proband-level evidence supporting an assumed de novo event.
Still missing
No parental testing or assumed de novo observation is documented, although PM6 is not permitted by the governing ATM VCEP specification.
PP2VCEPNot applicable: the ATM VCEP marks PP2 as not applicable, and this is not a missense variant.
Details ▾Hide ▴
Reasoning
The ATM VCEP v1.5 specification explicitly lists PP2 as 'Not Applicable' for this gene, superseding the generic ACMG PP2 (missense-in-a-gene-where-missense-is-common-mechanism-of-disease) criterion. Additionally, this variant is a nonsense/stop-gain change, not a missense variant, so PP2's missense mechanism does not apply regardless.
VCEP rule
ATM VCEP v1.5: PP2 = Not Applicable
Evidence considered
ATM VCEP v1.5 (cspec) lists PP2 applicability as 'Not Applicable'.
Variant is a nonsense (stop-gain) substitution, not missense.
PP3N/ANot applicable: the VCEP's SpliceAI-based PP3 excludes nonsense variants, which also cannot be combined with PVS1.
Details ▾Hide ▴
Reasoning
NM_000051.3:c.1348G>T produces NP_000042.3:p.(Glu450Ter) - a nonsense (stop-gain) variant, not a missense, silent, or intronic change. The ATM VCEP v1.5 PP3 strength rule explicitly enumerates only 'silent, missense/in-frame and ... intronic variants outside of donor and acceptor 1,2 sites' as eligible for the SpliceAI-based splicing sub-path; nonsense variants are absent from that list. The missense sub-path (REVEL >0.7333) also does not apply since this is not a missense substitution. Separately, the cspec's own PP3 instructionsToUse states 'PP3 for splice predictions may not be applied in addition to PVS1 or PVS1_Variable(RNA) codes' - this variant is a canonical null/nonsense change that is the PVS1 sub-group's variant type to adjudicate, so applying PP3 here would double-count the same loss-of-function mechanism. SpliceAI does show a high donor-gain delta (DS_DG=0.955) at this position, but per the VCEP's explicit typing/exclusivity rules this signal is not routed through PP3 for a nonsense variant.
Rule
ATM VCEP v1.5 PP3: Missense REVEL>0.7333; Splicing SpliceAI>=0.2 for silent/missense-in-frame/qualifying intronic variants only; not combinable with PVS1 codes for the same variant.
Evidence considered
case_summary.json cspec PP3 strength rule text limits the SpliceAI sub-path to 'silent, missense/in-frame and for intronic variants outside of donor and acceptor 1,2 sites' - nonsense/stop-gain is not among the listed eligible types.
vcep_materials.json ATM cspec PP3 instructionsToUse: 'PP3 for splice predictions may not be applied in addition to PVS1 or PVS1_Variable(RNA) codes', establishing mutual exclusivity between the splicing-based PP3 signal and a PVS1 call for a null/nonsense variant.
case_summary.json protein normalization: NP_000042.3:p.(Glu450Ter), confirming this is a nonsense/stop-gain variant, not missense/silent/intronic.
prefetch spliceai evidence_sentence: SpliceAI predicts possible splice impact (max delta score = 0.955, DS_DG=0.955 donor gain) - noted for completeness but not actionable under PP3 for this variant type per the VCEP's own typing rule.
Still missing
Confirm with the PVS1/null-variant group whether PVS1 (or PVS1_Variable RNA) is ultimately applied to this variant, since VCEP text bars combining PP3(splicing) with any PVS1 code for the same variant.
PP4VCEPNot applicable: the ATM VCEP specification marks PP4 as not applicable.
Details ▾Hide ▴
Reasoning
The applicable ATM VCEP v1.5 specification marks PP4 as Not Applicable. No generic PP4 assessment is substituted because a gene-specific specification exists.
VCEP rule
ATM VCEP v1.5: PP4 Not Applicable.
Evidence considered
The ATM VCEP v1.5 criterion specification explicitly marks PP4 as Not Applicable.
PP5VCEPNot applicable: the ATM VCEP marks PP5 as not applicable, and the ClinVar record is a single laboratory submission.
Details ▾Hide ▴
Reasoning
The applicable ATM VCEP v1.5 specification marks PP5 as Not Applicable. The exact-variant ClinVar record is a single laboratory submission with criteria provided and no expert-panel submission; therefore it cannot trigger PP5 under the requested rule.
VCEP rule
ATM VCEP v1.5: PP5 Not Applicable; requested override permits PP5 only for an exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification.
Evidence considered
The ATM VCEP v1.5 criterion specification explicitly marks PP5 as Not Applicable.
ClinVar variation 545995 exactly matches NM_000051.3:c.1348G>T / p.Glu450Ter but has only one GeneDx submission classified Likely pathogenic with review status criteria provided, single submitter; expert-panel submissions are zero.
Under the requested adjudication rule, an ordinary laboratory submission or non-expert exact-variant assertion does not support PP5.
BS2VCEPNot applicable: ATM cancer predisposition has incomplete penetrance, so observation in healthy individuals cannot be used as BS2.
Details ▾Hide ▴
Reasoning
BS2 is explicitly not applicable in the ATM VCEP v1.5 because ATM-related cancer predisposition has incomplete penetrance. Therefore, observation in an apparently healthy individual cannot be used as BS2 under this framework.
VCEP rule
ATM VCEP v1.5: BS2 not applicable; do not use because ATM has incomplete penetrance.
Evidence considered
The ATM VCEP v1.5 states: do not use BS2 because ATM has incomplete penetrance.
No healthy-individual or homozygote observation is needed to adjudicate BS2 because the governing VCEP makes the criterion not applicable.
BS4VCEPNot applicable: the ATM VCEP marks BS4 as not applicable for autosomal recessive ataxia-telangiectasia.
Details ▾Hide ▴
Reasoning
The ClinGen HBOP ATM VCEP v1.5 specification marks BS4 as not applicable. For autosomal recessive ataxia-telangiectasia, it states that informative instances of lack of co-segregation are too rare to receive weight at this time; no non-segregation evidence is present in the case bundle.
Evidence considered
The ATM VCEP v1.5 framework marks BS4 as not applicable and states: "AR Condition: Informative instances of lack of co-segregation in A-T families are too rare to be considered for weight at this time."
The case bundle contains no family genotypes, affected relatives lacking the variant, or other non-segregation observations.
Still missing
No affected-family-member non-segregation observations are documented, although BS4 is not permitted by the governing ATM VCEP specification.
BP1VCEPNot applicable: the ATM VCEP marks BP1 as not applicable, and this variant is itself truncating, not missense.
Details ▾Hide ▴
Reasoning
The ATM VCEP v1.5 specification explicitly lists BP1 as 'Not Applicable' for this gene, superseding the generic ACMG BP1 (missense variant in a gene where truncating variants are the only known mechanism) criterion. This variant is itself a truncating (nonsense) change, which is orthogonal to BP1's missense-benign logic in any case.
VCEP rule
ATM VCEP v1.5: BP1 = Not Applicable
Evidence considered
ATM VCEP v1.5 (cspec) lists BP1 applicability as 'Not Applicable'.
Variant is a nonsense (truncating) substitution, not missense.
BP3VCEPNot applicable: the ATM VCEP explicitly marks BP3 as not applicable for all ATM variants.
Details ▾Hide ▴
Reasoning
The ATM VCEP v1.5 CSPEC explicitly marks BP3 as 'Not Applicable' for ATM, independent of variant class or repeat-region context. This case is governed by the ATM CSPEC (VCEP mode), so the generic ACMG BP3 (in-frame indel in a repetitive region without a known functional effect) is not used for this gene; BP3 is not assessed for this or any other ATM variant under this specification.
VCEP rule
ATM VCEP v1.5: BP3 = Not Applicable (gene-specific exclusion, applies to all variant types including in-frame/repeat-region variants).
Evidence considered
case_summary.json CSPEC criteria list: BP3 entry has 'applicability': 'Not Applicable' with no rules, confirming the ATM VCEP does not permit BP3 application under any circumstance.
BP4N/ANot applicable: the SpliceAI-based BP4 sub-path excludes nonsense variants, and no calibrated missense predictor applies.
Details ▾Hide ▴
Reasoning
Same variant-type reasoning as PP3: c.1348G>T is a nonsense (stop-gain) change, p.(Glu450Ter), not missense or a qualifying splicing-sub-path type. The VCEP's BP4 splicing sub-path (SpliceAI<=0.1) is likewise specified only for silent/missense-in-frame/qualifying intronic variants, and this pipeline's curator override note on BP4 (gene_notes) explicitly restricts the SpliceAI-based BP4 sub-path to non-missense variants for which splicing is the primary plausible mechanism - a nonsense variant's primary plausible mechanism is nonsense-mediated decay/truncation (PVS1 territory), not splicing. The missense sub-path (REVEL<=0.249) does not apply since REVEL was not found for this variant and it is not a missense substitution regardless. BayesDel (0.652534) is available but, per this pipeline's standing policy, has no verified published calibration threshold and cannot be used for PP3/BP4 in either direction.
Rule
ATM VCEP v1.5 BP4: Missense REVEL<=0.249; Splicing SpliceAI<=0.1, restricted by curator override to variants whose primary plausible mechanism is splicing (not applicable to nonsense/stop-gain variants).
Evidence considered
case_summary.json cspec BP4 strength rule limits the SpliceAI sub-path (<=0.1) to the same variant-type list as PP3 (silent, missense/in-frame, qualifying intronic); nonsense is excluded.
vcep_materials.json gene_notes curator override: for BP4, missense variants must use ONLY the missense REVEL sub-path, and non-missense variants use the splicing sub-path only when splicing is the primary plausible mechanism (synonymous/qualifying intronic) - a nonsense/stop-gain variant's primary mechanism is truncation/NMD, not splicing, so neither BP4 sub-path is applicable here.
case_summary.json compact_evidence.revel: score null, found=false - REVEL not available for the missense sub-path (moot, variant is not missense).
case_summary.json compact_evidence.bayesdel: score 0.652534, found=true, but no verified published PP3/BP4 threshold/PMID can be cited for BayesDel in this pipeline, so it is treated as not_available for PP3/BP4 regardless of the numeric value.
Still missing
REVEL score not available (found=false) - moot here since the variant is not missense, but noted for completeness.
BayesDel score present (0.652534) but not usable: no verified published threshold citable for PP3/BP4 per pipeline policy.
BP5VCEPNot applicable: the ATM VCEP marks BP5 as not applicable, and no separate disease-causing variant is documented.
Details ▾Hide ▴
Reasoning
The applicable ATM VCEP v1.5 specification marks BP5 as Not Applicable. No separate disease-causing variant in the same individual is documented in the available case-evidence bundle for assessment under BP5.
VCEP rule
ATM VCEP v1.5: BP5 Not Applicable.
Evidence considered
The ATM VCEP v1.5 criterion specification explicitly marks BP5 as Not Applicable.
The available case-evidence bundle does not document a qualifying alternate molecular diagnosis in the same individual for BP5 assessment.
BP6VCEPNot applicable: the ATM VCEP marks BP6 as not applicable, and the ClinVar record is a single laboratory Likely pathogenic submission.
Details ▾Hide ▴
Reasoning
The applicable ATM VCEP v1.5 specification marks BP6 as Not Applicable. The exact-variant ClinVar record is not benign and is not an expert-panel assertion; it is a single laboratory Likely pathogenic submission, so it cannot trigger BP6 under the requested rule.
VCEP rule
ATM VCEP v1.5: BP6 Not Applicable; requested override permits BP6 only for an exact-variant ClinVar expert-panel Benign or Likely benign classification.
Evidence considered
The ATM VCEP v1.5 criterion specification explicitly marks BP6 as Not Applicable.
ClinVar variation 545995 exactly matches NM_000051.3:c.1348G>T / p.Glu450Ter and reports Likely pathogenic from one clinical laboratory with no expert-panel submissions.
Under the requested adjudication rule, only an exact-variant ClinVar expert-panel Benign or Likely benign classification can support BP6; none is present.
BP7N/ANot applicable: BP7 applies only to synonymous or deep intronic variants, and this is a coding nonsense change.
Details ▾Hide ▴
Reasoning
BP7 under the ATM VCEP v1.5 specification is restricted to synonymous (silent) substitutions and deep intronic variants (beyond +7 donor / -21 acceptor) that lack splicing prediction impact. NM_000051.3:c.1348G>T is a nonsense (stop-gain) coding substitution, NP_000042.3:p.(Glu450Ter), not a synonymous change and not intronic, so BP7 is categorically not applicable regardless of any SpliceAI score.
Rule
ATM VCEP v1.5 BP7: applicable only to synonymous and deep intronic variants; not applicable to nonsense/missense coding substitutions.
Evidence considered
case_summary.json cspec BP7 rule text restricts applicability to 'synonymous and deep intronic variants defined as further than +7 and further than -21 at donor and acceptor sites respectively'.
case_summary.json protein normalization confirms the variant is a nonsense substitution (p.Glu450Ter), not synonymous or intronic, excluding BP7 by variant type alone.
Every interpreted variant in ATM, mapped onto the exon structure of NM_000051.3. Select a row to find it on the plot.
ATMNM_000051.3
Fetching transcript structure from UCSC…
Exons
—
Transcript span
—
Strand
—
Variants mapped
—
Source
UCSC ncbiRefSeqCurated
All variants in ATM—Select a row to find it on the plot; use the link to open its page.
Protein
Location
Classification
Link
gnomAD v4.1Allele count data not recorded
gnomAD v2.1Allele count data not recorded
gnomAD-Canada v1.0Allele count data not recorded
Statements
gnomAD v4.1
Absent from gnomAD v4.1.
gnomAD v2.1
Absent from gnomAD v2.1.
gnomAD-Canada
Absent from gnomAD-Canada v1.0.
SpliceAI and Pangolin
SpliceAI acceptor lossat -9 bp
0.002
SpliceAI acceptor gainat -112 bp
0.051
SpliceAI donor lossat +259 bp
0.378
SpliceAI donor gainat -2 bp
0.955
Pangolin splice lossmagnitude shown
-0.285
Pangolin splice gain
0.793
Other predictors
REVEL
Not available
BayesDel
0.653
How the predictions were used
PP3N/ANot applicable: the VCEP's SpliceAI-based PP3 excludes nonsense variants, which also cannot be combined with PVS1.
Details ▾Hide ▴
Reasoning
NM_000051.3:c.1348G>T produces NP_000042.3:p.(Glu450Ter) - a nonsense (stop-gain) variant, not a missense, silent, or intronic change. The ATM VCEP v1.5 PP3 strength rule explicitly enumerates only 'silent, missense/in-frame and ... intronic variants outside of donor and acceptor 1,2 sites' as eligible for the SpliceAI-based splicing sub-path; nonsense variants are absent from that list. The missense sub-path (REVEL >0.7333) also does not apply since this is not a missense substitution. Separately, the cspec's own PP3 instructionsToUse states 'PP3 for splice predictions may not be applied in addition to PVS1 or PVS1_Variable(RNA) codes' - this variant is a canonical null/nonsense change that is the PVS1 sub-group's variant type to adjudicate, so applying PP3 here would double-count the same loss-of-function mechanism. SpliceAI does show a high donor-gain delta (DS_DG=0.955) at this position, but per the VCEP's explicit typing/exclusivity rules this signal is not routed through PP3 for a nonsense variant.
Rule
ATM VCEP v1.5 PP3: Missense REVEL>0.7333; Splicing SpliceAI>=0.2 for silent/missense-in-frame/qualifying intronic variants only; not combinable with PVS1 codes for the same variant.
Evidence considered
case_summary.json cspec PP3 strength rule text limits the SpliceAI sub-path to 'silent, missense/in-frame and for intronic variants outside of donor and acceptor 1,2 sites' - nonsense/stop-gain is not among the listed eligible types.
vcep_materials.json ATM cspec PP3 instructionsToUse: 'PP3 for splice predictions may not be applied in addition to PVS1 or PVS1_Variable(RNA) codes', establishing mutual exclusivity between the splicing-based PP3 signal and a PVS1 call for a null/nonsense variant.
case_summary.json protein normalization: NP_000042.3:p.(Glu450Ter), confirming this is a nonsense/stop-gain variant, not missense/silent/intronic.
prefetch spliceai evidence_sentence: SpliceAI predicts possible splice impact (max delta score = 0.955, DS_DG=0.955 donor gain) - noted for completeness but not actionable under PP3 for this variant type per the VCEP's own typing rule.
Still missing
Confirm with the PVS1/null-variant group whether PVS1 (or PVS1_Variable RNA) is ultimately applied to this variant, since VCEP text bars combining PP3(splicing) with any PVS1 code for the same variant.
BP4N/ANot applicable: the SpliceAI-based BP4 sub-path excludes nonsense variants, and no calibrated missense predictor applies.
Details ▾Hide ▴
Reasoning
Same variant-type reasoning as PP3: c.1348G>T is a nonsense (stop-gain) change, p.(Glu450Ter), not missense or a qualifying splicing-sub-path type. The VCEP's BP4 splicing sub-path (SpliceAI<=0.1) is likewise specified only for silent/missense-in-frame/qualifying intronic variants, and this pipeline's curator override note on BP4 (gene_notes) explicitly restricts the SpliceAI-based BP4 sub-path to non-missense variants for which splicing is the primary plausible mechanism - a nonsense variant's primary plausible mechanism is nonsense-mediated decay/truncation (PVS1 territory), not splicing. The missense sub-path (REVEL<=0.249) does not apply since REVEL was not found for this variant and it is not a missense substitution regardless. BayesDel (0.652534) is available but, per this pipeline's standing policy, has no verified published calibration threshold and cannot be used for PP3/BP4 in either direction.
Rule
ATM VCEP v1.5 BP4: Missense REVEL<=0.249; Splicing SpliceAI<=0.1, restricted by curator override to variants whose primary plausible mechanism is splicing (not applicable to nonsense/stop-gain variants).
Evidence considered
case_summary.json cspec BP4 strength rule limits the SpliceAI sub-path (<=0.1) to the same variant-type list as PP3 (silent, missense/in-frame, qualifying intronic); nonsense is excluded.
vcep_materials.json gene_notes curator override: for BP4, missense variants must use ONLY the missense REVEL sub-path, and non-missense variants use the splicing sub-path only when splicing is the primary plausible mechanism (synonymous/qualifying intronic) - a nonsense/stop-gain variant's primary mechanism is truncation/NMD, not splicing, so neither BP4 sub-path is applicable here.
case_summary.json compact_evidence.revel: score null, found=false - REVEL not available for the missense sub-path (moot, variant is not missense).
case_summary.json compact_evidence.bayesdel: score 0.652534, found=true, but no verified published PP3/BP4 threshold/PMID can be cited for BayesDel in this pipeline, so it is treated as not_available for PP3/BP4 regardless of the numeric value.
Still missing
REVEL score not available (found=false) - moot here since the variant is not missense, but noted for completeness.
BayesDel score present (0.652534) but not usable: no verified published threshold citable for PP3/BP4 per pipeline policy.
BP7N/ANot applicable: BP7 applies only to synonymous or deep intronic variants, and this is a coding nonsense change.
Details ▾Hide ▴
Reasoning
BP7 under the ATM VCEP v1.5 specification is restricted to synonymous (silent) substitutions and deep intronic variants (beyond +7 donor / -21 acceptor) that lack splicing prediction impact. NM_000051.3:c.1348G>T is a nonsense (stop-gain) coding substitution, NP_000042.3:p.(Glu450Ter), not a synonymous change and not intronic, so BP7 is categorically not applicable regardless of any SpliceAI score.
Rule
ATM VCEP v1.5 BP7: applicable only to synonymous and deep intronic variants; not applicable to nonsense/missense coding substitutions.
Evidence considered
case_summary.json cspec BP7 rule text restricts applicability to 'synonymous and deep intronic variants defined as further than +7 and further than -21 at donor and acceptor sites respectively'.
case_summary.json protein normalization confirms the variant is a nonsense substitution (p.Glu450Ter), not synonymous or intronic, excluding BP7 by variant type alone.
SpliceAI predicts possible splice impact for this variant (max delta score = 0.95). BayesDel score = 0.652534.
3 paper records listed. No publication is explicitly linked to an applied criterion in the recorded metadata. None of the listed paper records explicitly indicates that the variant is named.
Recorded publications
ATM Mutations in Cancer: Therapeutic Implications.
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.