ATM c.5503del, p.(Thr1835LeufsTer11)

Reasoning

Governing framework is the ClinGen HBOP ATM VCEP v1.5 specification (cspec), which directs PVS1 through the ATM PVS1 decision tree (adapted from PMID 30192042 / ClinGen SVI PVS1 recommendations, PMC6185798). The variant NM_000051.4:c.5503del is a single-nucleotide frameshift predicted to produce NP_000042.3:p.(Thr1835LeufsTer11), a premature termination codon (PTC) at codon 1845 (11 residues into the shifted frame). (1) LOF mechanism gate: germline loss of function is an established disease mechanism for ATM — the VCEP itself provides PVS1 guidance and pvs1_gene_context marks the gene eligible. (2) Transcript relevance: NM_000051.4 is MANE Select (variant_validator: mane_select=true, refseq_select=true) and shares the coding sequence with the VCEP default transcript NM_000051.3/ENST00000278616.8; the VCEP states all exons are constitutive without alternate isoforms that could rescue presumed LoF events. (3) NMD: the frameshift lies in exon 37 of 63 (total numbering) and the PTC at aa 1845 is >50 nt upstream of the final exon-exon junction, so NMD is predicted; this is not an extreme-3'-end variant — the VCEP's most C-terminal residue considered pathogenic is p.Arg3047 (PMIDs 8755918, 19691550, 18560558, 10980530, 26628246), and the PTC here (1845) is far upstream of it. (4) Exon/domain importance: truncation at 1845 removes the entire FAT/PI3K/FATC (FATKIN) region, which the VCEP treats as one contiguous domain critical for ATM function (PMID 28508083); per VCEP PVS1 guide note 2, even a NMD-escaping truncation adversely affecting FATKIN warrants full PVS1 rather than a downgrade. (5) No competing splice effect: SpliceAI max delta 0.05 predicts no significant splice impact, so the splice-related caution does not apply. (6) No population-LoF tolerance signal: variant is absent from gnomAD v2.1, v4.1 and gnomAD-Canada. No decision-tree downgrade trigger applies; PVS1 is met at Very Strong.

Rule

Use ATM PVS1 Decision Tree: frameshift null variant with predicted NMD in a gene where LOF is a known mechanism of disease, not at the extreme 3' end, no splice-escape or population-LoF-tolerance downgrade triggers -> PVS1 (Very Strong).

Evidence considered

• ClinGen HBOP ATM VCEP v1.5 CSPEC: PVS1 rule 'Use ATM PVS1 Decision Tree' with Very Strong default; PM5 truncation boundary at p.Arg3047 cited as the most C-terminal pathogenic residue context.
• ATM PVS1 v1.5 decision guide (ATM_PVS1_1.5.pdf): frameshift is a PVS1-eligible null variant class; caveats are extreme-3'-end variants and splice variants leaving remainder of protein intact, neither of which applies; FAT/PI3K/FATC (FATKIN) treated as one contiguous critical domain (PMID 28508083); all exons of NM_000051.3 considered constitutive; NMD-escaping alterations adversely affecting FATKIN can still be granted PVS1.
• pvs1_variant_assessment: variant classified as frameshift (variant_bucket 'frameshift'), generic PVS1 framework applicable, default strength PVS1, no exon-skipping/3'-end downgrade triggers identified; transcript NM_000051.4:c.5503del normalized to p.(Thr1835LeufsTer11) (Mutalyzer).
• pvs1_gene_context: germline loss of function is an established ATM disease mechanism per the VCEP (cspec), gene gate 'eligible'.
• Transcript structure from Mutalyzer normalization: c.5503 falls in exon 37 of 63 (c.5497-5674, total exon numbering; 36th coding exon); PTC at aa 1845 is therefore mid-gene and >50 nt upstream of the final exon-exon junction, supporting predicted NMD and excluding an extreme-3'-end scenario.
• Variant is absent from gnomAD v2.1, gnomAD v4.1 and gnomAD-Canada, consistent with an LoF-intolerant gene and a true null allele.
• SpliceAI max delta score 0.05 (DS_AG 0.053): no significant predicted splice impact, so the PVS1 splice-caution does not apply.

Still missing

• No RNA-based confirmation of NMD was available; the Very Strong call rests on the standard predicted-NMD decision-tree path for a mid-gene frameshift, which is consistent with VCEP guidance.
• No functional assay readout for this specific variant was available to corroborate loss of function (PS3 territory, not required for PVS1).

Sources

Reasoning

ClinGen HBOP ATM VCEP v1.5 (cspec) defines PM2 as frequency <=0.001% in the gnomAD v4 dataset, with the note that n=1 in a single subpopulation is already sufficiently rare for PM2_supporting. The variant is absent from gnomAD v4.1 (0 alleles, AF=0), which is below the 0.001% threshold, and absence is corroborated by gnomAD v2.1 (exomes) and gnomAD-Canada v1.0 (genomes). No homozygotes are observed in any queried dataset. PM2 is therefore met at Supporting strength per the VCEP rule (the cspec PM2 rule maps to PM2_Supporting). This is a frameshift/null variant (p.Thr1835LeufsTer11), and population absence is consistent with the recessive A-T disease mechanism; however, the PM2 call itself rests solely on the VCEP frequency threshold and the gnomAD absence data. No other source's final classification was used.

Rule

Frequency <=0.001% in gnomAD v4 dataset; if n=1 in a single sub population, that is sufficiently rare and PM2_supporting applies

Evidence considered

• ClinGen HBOP ATM VCEP v1.5 (cspec, svi doc 639508985) specifies PM2 as: 'Frequency <=.001% in gnomAD v4 dataset. If n=1 in a single sub population, that is sufficiently rare and PM2_supporting would apply.' The cspec rule payload maps this to PM2_Supporting.
• gnomAD v4.1 (exome) direct variant lookup for chr11-108304677-GA-G returned 'Absent from gnomAD v4.1.' (0 alleles; no homozygotes).
• gnomAD v2.1 (exome) direct variant lookup for 11-108175404-GA-G returned 'Absent from gnomAD v2.1.'
• gnomAD-Canada v1.0 (HostSeq genomes) API lookup for 11-108304677-GA-G returned 'Absent from gnomAD-Canada v1.0.'
• An allele count of 0 (AF=0) is below the <=0.001% PM2 threshold, so PM2_Supporting is met.

Still missing

• gnomAD source_data is empty (no per-population allele counts captured); absence (n=0) is reported as a direct-lookup result rather than a parsed count, so a per-subpopulation count breakdown is unavailable. This does not affect the determination because total absence satisfies the <=0.001% threshold.
• Only the gnomAD v4.1 exome dataset was queried for v4 (label 'gnomAD v4.1 (exome)'); a separate gnomAD v4 genome-only query was not performed, though gnomAD-Canada v1.0 genomes corroborate absence.

Sources

Reasoning

PM5 Supporting is met under the ATM VCEP v1.5 gene-specific rule. The variant is a frameshift with a premature termination codon after p.Thr1835, upstream of p.Arg3047, satisfying the specified truncating-variant cutoff. No classic same-residue missense comparator search is applicable because the governing ATM rule uses this truncation-cutoff pathway for PM5.

Rule

ATM VCEP v1.5 PM5: apply PM5 Supporting to frameshifting or truncating variants with premature termination codons upstream of p.Arg3047.

Evidence considered

• The ATM VCEP v1.5 PM5 rule states that PM5 Supporting applies to frameshifting or truncating variants with premature termination codons upstream of p.Arg3047.
• The case normalization identifies NM_000051.4:c.5503del as p.(Thr1835LeufsTer11), which terminates upstream of p.Arg3047.
• The PM5 candidate artifact reports that classic same-residue missense searching is ineligible for this non-missense, truncation-cutoff framework; it does not provide independent classification evidence.

Sources

Reasoning

ClinGen HBOP ATM VCEP v1.5 (cspec, found=true, precedence rank 1) defines BA1 as Grpmax Filtering AF >0.5% in the gnomAD v4 dataset. The variant NM_000051.4:c.5503del (p.Thr1835LeufsTer11), chr11-108304677-GA-G (GRCh38)/11-108175404-GA-G (GRCh37), is absent from gnomAD v4.1 (search_status 'absent', found=false), i.e. allele frequency 0, which does not exceed the 0.5% threshold. Absence is corroborated by gnomAD v2.1 (exomes) and gnomAD-Canada v1.0 (HostSeq genomes), both also reporting the variant absent. BA1 is therefore not met. No ClinVar or other final classification was used to reach this determination.

Rule

Grpmax Filtering AF >0.5% in gnomAD v4 dataset

Evidence considered

• ClinGen HBOP ATM VCEP v1.5 (cspec, svi doc 639508985) specifies BA1 as: 'Grpmax Filtering AF >.5% in gnomAD v4 dataset' (Benign Stand Alone).
• gnomAD v4.1 (exome) direct variant lookup for chr11-108304677-GA-G (dataset=gnomad_r4) returned search_status 'absent', found=false, evidence_sentence 'Absent from gnomAD v4.1.'
• gnomAD v2.1 (exome) direct variant lookup for 11-108175404-GA-G returned 'Absent from gnomAD v2.1.'
• gnomAD-Canada v1.0 (HostSeq genomes) API lookup for 11-108304677-GA-G returned 'Absent from gnomAD-Canada v1.0.'
• An allele frequency of 0 does not satisfy the >0.5% grpmax filtering AF requirement, so BA1 is not met.

Still missing

• gnomAD source_data is empty (per-population allele counts/grpmax AF not captured); only the absent/not-found determination from the direct variant lookup is available.

Sources

Reasoning

ClinGen HBOP ATM VCEP v1.5 (cspec) defines BS1 as Grpmax Filtering AF >0.05% in the gnomAD v4 dataset. The variant is absent from gnomAD v4.1 (search_status 'absent', found=false), so its allele frequency is 0, well below the 0.05% threshold. Absence is corroborated by gnomAD v2.1 and gnomAD-Canada v1.0. BS1 is therefore not met. No other source's final classification was used.

Rule

Grpmax Filtering AF >0.05% in gnomAD v4 dataset

Evidence considered

• ClinGen HBOP ATM VCEP v1.5 (cspec, svi doc 639508985) specifies BS1 as: 'Grpmax Filtering AF >.05% in gnomAD v4 dataset' (Benign Strong).
• gnomAD v4.1 (exome) direct variant lookup for chr11-108304677-GA-G returned 'Absent from gnomAD v4.1.'
• gnomAD v2.1 (exome) direct variant lookup for 11-108175404-GA-G returned 'Absent from gnomAD v2.1.'
• gnomAD-Canada v1.0 (HostSeq genomes) API lookup for 11-108304677-GA-G returned 'Absent from gnomAD-Canada v1.0.'
• An allele frequency of 0 does not satisfy the >0.05% grpmax filtering AF requirement, so BS1 is not met.

Still missing

• gnomAD source_data is empty (per-population allele counts/grpmax AF not captured); only the absent/not-found determination from the direct variant lookup is available.

Sources

Reasoning

The ATM VCEP permits PS3 only for validated functional evidence showing that the variant fails to rescue an ATM-specific feature, with PS3_Moderate when both an ATM-specific feature and radiosensitivity fail and PS3_Supporting when only the ATM-specific feature fails; PS3 Strong is not used. The consolidated evidence bundle contains no variant-specific result for NM_000051.4:c.5503del in an approved ATM assay. The reviewed literature extractions do not mention this variant. The Sun et al. 2025 supplementary table concerns ATM SNVs and does not provide a direct assay result for this frameshift variant. Therefore, PS3 cannot be assigned.

Rule

ATM VCEP v1.5: do not use PS3 as Strong; use PS3_Moderate only when both an ATM-specific feature and radiosensitivity fail to rescue, and PS3_Supporting when an ATM-specific feature alone fails to rescue; do not use radiosensitivity alone.

Evidence considered

• The ATM VCEP specification states that PS3 Strong must not be used; PS3_Moderate requires failure to rescue both an ATM-specific feature and radiosensitivity, while PS3_Supporting requires failure of an ATM-specific feature alone and must not be based on radiosensitivity alone.
• The ATM VCEP functional-assay table calibrates kinase-activity assays at proposed PS3_Supporting strength and allows radiosensitivity to contribute only when combined with kinase-activity evidence; it is a methodology/calibration table, not a per-variant result table.
• The bundle's paper_extracts entries for PMID:27413114, PMID:30348496, PMID:30553448, and PMID:25394175 all state that c.5503del and p.(T1835Lfs*11) were not mentioned, so none supplies variant-specific PS3 evidence.
• The bundle identifies the variant as a frameshift, whereas the Sun et al. 2025 supplementary dataset is described as an ATM SNV table; no direct functional result for this variant is available there.

Still missing

• A primary report or validated assay record must be located that directly tests NM_000051.4:c.5503del or its equivalent p.(Thr1835LeufsTer11).
• The report must provide interpretable ATM-specific functional readout(s), appropriate controls, and, for PS3_Moderate, concordant radiosensitivity evidence.
• Assay results must be checked against the ATM VCEP-approved assay methodology and calibrated thresholds before assigning strength.

Sources

Reasoning

No exact-variant case-control enrichment dataset was identified. The ATM HBOP VCEP specification requires a case-control study with p-value <=0.05 and odds ratio, hazard ratio, or relative risk >=2, or a lower 95% confidence bound >=1.5; the available paper provides only gene-level familial-pancreatic-cancer frequencies and does not mention this variant.

Rule

ATM HBOP VCEP v1.5 PS4: case-control study with p-value <=0.05 and OR/HR/RR >=2 or lower 95% CI >=1.5.

Evidence considered

• The applicable ATM HBOP VCEP PS4 rule requires case-control evidence with p-value <=0.05 and an odds ratio, hazard ratio, or relative risk >=2, or lower 95% CI >=1.5.
• PMID:25394175 reports gene-level ATM mutation frequencies in familial pancreatic cancer, but the extraction confirms that NM_000051.4:c.5503del / p.(Thr1835Lfs*11) is not mentioned; it cannot establish variant-specific enrichment.

Still missing

• An exact-variant case-control analysis reporting effect size, confidence interval, and p-value is required.

Sources

Reasoning

The ATM VCEP v1.5 PM3 point-based specification applies to Ataxia-Telangiectasia and requires affected-proband observations with a pathogenic/likely pathogenic variant in trans, phase-unknown co-occurrence, an unidentified/VUS second allele, or qualifying homozygosity. The case bundle contains no affected-proband genotype, second-allele identity or classification, phase, parental data, or qualifying homozygous observation for NM_000051.4:c.5503del; therefore no PM3 points or strength can be assigned.

Rule

ATM VCEP v1.5 PM3 points: confident A-T phenotype with confirmed-in-trans P/LP variant = 4.0 points; phase unknown = 2.0; second variant unidentified or VUS = 1.0; homozygous = 2.0 with a maximum of 2 individuals. Consistent phenotype values are half these amounts. PM3 strength thresholds: Supporting 1.0, Moderate 2.0, Strong 4.0, Very Strong 8.0.

Evidence considered

• The ClinGen HBOP ATM VCEP v1.5 framework specifies PM3 for recessive A-T using points per unrelated A-T proband: 4 points for a confident phenotype with a confirmed-in-trans pathogenic variant, 2 points when phase is unknown, 1 point when the second variant is unidentified or a VUS, and 2 points for homozygosity subject to the framework cap; strength thresholds are 1 supporting, 2 moderate, 4 strong, and 8 very strong points.
• The case evidence bundle documents the variant as NM_000051.4:c.5503del (NP_000042.3:p.(Thr1835LeufsTer11)) but provides no proband-level allelic, phase, parental, or homozygosity evidence for PM3.

Still missing

• Affected-proband observations for this variant are not available.
• No qualifying pathogenic/likely pathogenic second ATM variant and no trans or cis phase information are documented.
• No parental testing, consanguinity evidence, or qualifying homozygous A-T observation is documented.

Sources

Reasoning

The ATM VCEP v1.5 PP1 rule is applicable only under the AR condition (ataxia telangiectasia): segregation in >=3 affected relatives = Strong, 2 affected relatives = Moderate, 1 affected relative = Supporting, with the instruction that affected relatives must have both variants identified in the proband; the AD condition (ATM-related cancer predisposition) is explicitly not used (PMID 32773770, co-segregation in low-penetrance genes can produce false positives). No segregation data for c.5503del exists anywhere in the collected evidence: no A-T proband family with affected relatives tested for this variant, no meioses counted, no informative co-segregation or non-segregation events; the single ClinVar submission carries no segregation details (0 criterion leads), literature_pass.json surfaced no segregation-themed papers, and all four fetched full texts have variant_mentioned=false in paper_extracts.json. Because the evidence base is absent rather than contradictory, PP1 is not_assessed with the gap identified, not met or not_met.

Rule

VCEP v1.5 PP1 (AR only): >=3 affected relatives = Strong; 2 = Moderate; 1 = Supporting; affected relatives must have both proband variants identified

Evidence considered

• ClinGen HBOP ATM VCEP v1.5 (cspec) PP1 rules: 'AR Condition: Segregation in >=3 affected relatives' (Strong), 'Segregation in 2 affected relatives' (Moderate), 'Segregation in 1 affected relative' (Supporting); instructions: 'AR Condition: Affected relatives must have both variants identified in proband. AD Condition - Do not use: Co-segregation analysis in lower-penetrance genes can lead to false positive results (PMID 32773770).'
• No cosegregation information for c.5503del in the ClinVar submission (single submitter, criteria provided single submitter, no criterion leads in the submission audit).
• paper_extracts.json shows all four fetched papers (PMID:25394175, PMID:27413114, PMID:30348496, PMID:30553448) do not mention c.5503del or p.(Thr1835LeufsTer11) (variant_mentioned=false), so none contribute segregation data.

Still missing

• No segregation data for c.5503del: no affected relatives of an A-T proband carrying the variant with both proband variants identified were reported in ClinVar, literature_pass, or any fetched full text; segregation testing in an A-T family is required before PP1 can be evaluated.

Sources

Reasoning

No exact-variant ClinVar expert-panel Pathogenic/Likely pathogenic assertion is present. The exact match has one Pathogenic clinical-laboratory submission from Ambry Genetics with review status 'criteria provided, single submitter'; it is not an expert-panel assertion and therefore cannot support PP5.

Rule

PP5 may be applied only from an exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification.

Evidence considered

• ClinVar variation 825788 exactly matches NM_000051.4:c.5503del, but its sole submission is Ambry Genetics (Pathogenic; criteria provided, single submitter), and the submission audit identifies zero expert-panel submissions.

Still missing

• An exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic assertion is required.

Sources

Reasoning

The ATM VCEP permits BS3 only for validated functional evidence showing rescue of an ATM-specific feature, radiosensitivity, or both. The consolidated evidence bundle contains no variant-specific result for NM_000051.4:c.5503del in an approved ATM assay. The reviewed literature extractions do not mention this variant, and the Sun et al. 2025 SNV supplementary table does not provide a direct assay result for this frameshift variant. Consequently, no calibrated benign functional evidence is available and BS3 cannot be assigned.

Rule

ATM VCEP v1.5: use BS3_Moderate when both an ATM-specific feature and radiosensitivity are rescued, and BS3_Supporting when either an ATM-specific feature or radiosensitivity is rescued.

Evidence considered

• The ATM VCEP specification states that BS3_Moderate requires rescue of both an ATM-specific feature and radiosensitivity, while BS3_Supporting may be used when either the ATM-specific feature or radiosensitivity is rescued.
• The ATM VCEP functional-assay table provides approved assay methodology and calibration but does not report results for the current variant; its presence alone cannot establish BS3.
• The bundle's paper_extracts entries for PMID:27413114, PMID:30348496, PMID:30553448, and PMID:25394175 all state that c.5503del and p.(T1835Lfs*11) were not mentioned, so none supplies variant-specific BS3 evidence.
• The bundle identifies the variant as a frameshift, and the Sun et al. 2025 supplementary dataset is described as an ATM SNV table without a direct result for this variant.

Still missing

• A primary report or validated assay record must be located that directly tests NM_000051.4:c.5503del or its equivalent p.(Thr1835LeufsTer11).
• The report must document rescue of an ATM-specific feature and/or radiosensitivity with adequate positive and negative controls.
• Assay results must be checked against the ATM VCEP-approved assay methodology and calibrated thresholds before assigning BS3 strength.

Sources

Reasoning

The ATM VCEP v1.5 BP2 point-based specification requires unaffected, non-A-T individuals carrying NM_000051.4:c.5503del with a VCEP-classified pathogenic/likely pathogenic ATM variant in trans, or qualifying phase-unknown/homozygous observations. The case bundle contains no such unaffected-individual co-occurrence, second-allele classification, phase, or homozygosity evidence; therefore no BP2 points or strength can be assigned.

Rule

ATM VCEP v1.5 BP2 points: unaffected non-A-T individual with a VCEP-classified P/LP variant in confirmed trans = -4.0; phase unknown = -2.0; homozygous maximum -2.0, with laboratory-setting -2.0 and database-setting -1.0. BP2 strength thresholds: Supporting -1.0, Moderate -2.0, Strong at or below -4.0.

Evidence considered

• The ClinGen HBOP ATM VCEP v1.5 framework specifies BP2 for unaffected non-A-T individuals with a P/LP variant in trans: -4 points for confirmed trans, -2 points for phase unknown, and homozygous observations capped at -2 points; laboratory-setting and database-setting homozygous values are -2 and -1 respectively. BP2 strength thresholds are Supporting -1, Moderate -2, and Strong at or below -4 points.
• The case evidence bundle contains no unaffected-individual co-occurrence, phase, parental, homozygosity, or VCEP-classified second-variant evidence for BP2.

Still missing

• No unaffected non-A-T individuals carrying this variant with a pathogenic/likely pathogenic ATM variant are documented.
• No confirmed-in-trans, phase-unknown, or qualifying homozygous observations are documented.
• The classification of any potential co-occurring ATM variant under the ATM VCEP specification is unavailable.

Sources

Reasoning

No exact-variant ClinVar expert-panel Benign/Likely benign assertion is present. The exact match has only one non-expert Pathogenic laboratory submission, which cannot support BP6.

Rule

BP6 may be applied only from an exact-variant ClinVar expert-panel Benign or Likely benign classification.

Evidence considered

• ClinVar variation 825788 exactly matches NM_000051.4:c.5503del, but the submission audit identifies one non-expert Pathogenic Ambry Genetics submission and zero expert-panel submissions.

Still missing

• An exact-variant ClinVar expert-panel Benign or Likely benign assertion is required.

Sources