ATM c.5890A>G, p.Lys1964Glu

Reasoning

The VCEP BP4 rule applies to this missense variant. REVEL is 0.207, meeting the missense threshold of ≤.249, so BP4 Supporting is met. SpliceAI maximum delta is also 0.021, meeting the VCEP no-predicted-impact threshold of ≤0.1; this is concordant support within the single BP4 criterion, not a second code. Sun et al. 2025 Table S1 contains the exact c.5890A>G row with REVEL 0.207 and a computational Functional prediction; because DeepATM_predicted=Yes and Confidence=Computational prediction, that table row is corroboration only, not a functional assay result.

VCEP rule

* Missense: REVEL score **≤.249** * Splicing: No predicted impact via splicing (SpliceAI **≤0.1**).

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 BP4 rule: missense REVEL ≤.249; splicing no predicted impact by SpliceAI ≤0.1.
• Local REVEL lookup for NM_000051.4:c.5890A>G / p.(K1964E): REVEL score 0.207.
• SpliceAI result for NM_000051.4:c.5890A>G: maximum delta score 0.021, which is ≤0.1.
• Sun et al. 2025 Supplementary Table S1, row ATM c.5890A>G / K1964E: REVEL 0.207, Classification=Functional, Confidence=Computational prediction, DeepATM_predicted=Yes; this is computational corroboration and not an assayed functional result.

Sources

Reasoning

The ATM VCEP permits PM2 at Supporting strength when frequency is ≤0.001% in the gnomAD subpopulation with the highest frequency, with an exception only when frequency is >0.001% but n=1 in a single subpopulation and absent from every other subpopulation. The highest all-comers subpopulation frequency is 0.0822368% in gnomAD v4.1 (Middle Eastern, 5/6,080 alleles), and gnomAD v2.1 is 0.0555247% (Remaining individuals, 4/7,204 alleles); neither meets the rarity threshold, and neither is the n=1 exception. The non-cancer v2.1 exome highest subpopulation frequency is 0.0299222% (2/6,684 alleles), also above the threshold. Therefore PM2 is not met.

VCEP rule

Frequency **≤.001%** in gnomAD subpopulation with the highest frequency ... Exception: If >0.001% but n=1 in a single subpopulation (not present in any other subpopulation), that is sufficiently rare and PM2\_supporting would apply.

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 CSPEC PM2 rule: Frequency ≤.001% in gnomAD subpopulation with the highest frequency. Exception: If >0.001% but n=1 in a single subpopulation (not present in any other subpopulation), that is sufficiently rare and PM2_supporting would apply.
• gnomAD v4.1 all-comers: highest subpopulation is Middle Eastern, AF 0.000822368 (0.0822368%), 5/6,080 alleles, homozygotes 0; the variant is also observed in other subpopulations.
• gnomAD v2.1 all-comers: highest subpopulation is Remaining individuals, AF 0.000555247 (0.0555247%), 4/7,204 alleles, homozygotes 1; the variant is also observed in other subpopulations.
• GNOMAD_V2_1_NON_CANCER exomes: highest subpopulation is Remaining individuals, AF 0.000299222 (0.0299222%), 2/6,684 alleles, homozygotes 1; this is above 0.001% and is not a single-observation-only occurrence.

Sources

Reasoning

The VCEP PP3 rule applies to this missense variant through either the missense REVEL path or the splicing path. REVEL is 0.207, below the required >.7333 threshold, and SpliceAI maximum delta is 0.021, below the required ≥0.2 threshold for missense/in-frame variants. The governing Table S1 row for c.5890A>G reports REVEL 0.207 and classifies the DeepATM result as Functional with Confidence=Computational prediction and DeepATM_predicted=Yes; this is computational corroboration only and does not create an additional PP3 code.

VCEP rule

* Missense: REVEL **\>.7333** * Splicing: Predicted impact via splicing (SpliceAI **≥0.2**) for silent, missense/in-frame and for intronic variants outside of donor and acceptor 1,2 sites.

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 PP3 rule: missense REVEL >.7333; splicing SpliceAI ≥0.2 for silent, missense/in-frame and intronic variants outside donor and acceptor 1,2 sites.
• Local REVEL lookup for NM_000051.4:c.5890A>G / p.(K1964E): REVEL score 0.207.
• SpliceAI result for NM_000051.4:c.5890A>G: DS_AG 0.021, DS_AL 0.0, DS_DG 0.0, DS_DL 0.001; maximum delta score 0.021.
• Sun et al. 2025 Supplementary Table S1, row ATM c.5890A>G / K1964E: REVEL 0.207, Classification=Functional, Confidence=Computational prediction, DeepATM_predicted=Yes; the row is a prediction rather than an assayed result and is used only as corroboration here.

Sources

Reasoning

The ATM VCEP BA1 rule requires grpmax filtering AF >0.5% in gnomAD. The default all-comers gnomAD v4.1 grpmax FAF is 0.00032328 (0.032328%), and gnomAD v2.1 is 0.00007515 (0.007515%); both are below 0.5%. The available gnomAD v2.1 non-cancer exome data also have grpmax FAF 0.00007575 (0.007575%), below the threshold. Therefore BA1 is not met.

VCEP rule

Grpmax Filtering AF **>.5%** in gnomAD

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 CSPEC BA1 rule: Grpmax Filtering AF >.5% in gnomAD.
• gnomAD v4.1 all-comers: grpmax FAF 0.00032328 (0.032328%); total AF 7.56149e-05 (122/1,613,438 alleles), homozygotes 1.
• gnomAD v2.1 all-comers: grpmax FAF 7.515e-05 (0.007515%); total AF 9.20576e-05 (26/282,432 alleles), homozygotes 1.
• GNOMAD_V2_1_NON_CANCER exomes: AF 9.72467e-05 (23/236,512 alleles), homozygotes 1; grpmax FAF 7.575e-05 (0.007575%).

Sources

Reasoning

The ATM VCEP BS1 rule requires grpmax filtering AF >0.05% in gnomAD. The default all-comers gnomAD v4.1 grpmax FAF is 0.00032328 (0.032328%), and gnomAD v2.1 is 0.00007515 (0.007515%); both are below 0.05%. The available gnomAD v2.1 non-cancer exome grpmax FAF is 0.00007575 (0.007575%), also below the threshold. Therefore BS1 is not met.

VCEP rule

Grpmax Filtering AF **>.05%** in gnomAD

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 CSPEC BS1 rule: Grpmax Filtering AF >.05% in gnomAD.
• gnomAD v4.1 all-comers: grpmax FAF 0.00032328 (0.032328%); highest named subpopulation is Middle Eastern at AF 0.000822368 (0.0822368%).
• gnomAD v2.1 all-comers: grpmax FAF 7.515e-05 (0.007515%); highest named subpopulation is Remaining individuals at AF 0.000555247 (0.0555247%).
• GNOMAD_V2_1_NON_CANCER exomes: grpmax FAF 7.575e-05 (0.007575%); highest named subpopulation is Remaining individuals at AF 0.000299222 (0.0299222%).

Sources

Reasoning

The ATM VCEP permits PS1 for a missense change only when the same amino-acid change has been established as pathogenic and splicing has been ruled out for both alterations. The case is a missense variant, but searching ATM_PS1_1.6.pdf.txt for c.5890A>G, p.Lys1964Glu, and p.K1964E found no governing PS1-table entry, and the available evidence does not identify a previously established pathogenic ATM variant with the same amino-acid change. Therefore PS1 cannot be assessed from the VCEP-governed evidence.

VCEP rule

Use for missense changes as long as splicing is ruled-out for both alterations.

Evidence considered

• The VCEP case normalization identifies NM_000051.4:c.5890A>G as the missense substitution NP_000042.3:p.(Lys1964Glu) / p.(K1964E).
• ATM_PS1_1.6.pdf.txt was searched for c.5890A>G, p.Lys1964Glu, and p.K1964E; no entry for this variant was found.
• The VCEP PS1 guide requires a same-amino-acid pathogenic comparator and, for missense changes, splice defects ruled out for both alterations; neither comparator evidence nor a governing pre-assigned PS1 code is available here.

Still missing

• No entry for c.5890A>G, p.Lys1964Glu, or p.K1964E was found in ATM_PS1_1.6.pdf.txt; a previously established pathogenic same-amino-acid comparator is not available in the consulted evidence.

Sources

Reasoning

The governing ATM VCEP rule permits PS3 at Supporting when a variant fails to rescue an ATM-specific feature, but not for radiosensitivity alone. The exact variant was not found in the ClinGen HBOP ATM approved functional-assay supplementary tables. Sun et al. 2025 Supplementary Table S1 contains the exact c.5890A>G / K1964E row, but it is marked Classification=Functional, Confidence=Computational prediction, and DeepATM_predicted=Yes; under that file's governing policy, this is a model prediction rather than a direct prime-editing assay and cannot be used as PS3. No qualifying variant-specific ATM functional assay result is therefore available to assess PS3.

VCEP rule

Use when a variant fails to rescue an ATM specifc 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

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 PS3 rule: use Supporting when a variant fails to rescue an ATM-specific feature only, and do not use radiosensitivity-only evidence.
• ClinGen HBOP ATM approved functional-assay supplementary tables: searched c.5890A>G, p.Lys1964Glu, p.K1964E, and K1964E; no exact variant entry was found.
• Sun et al. 2025 Supplementary Table S1, exact row ATM c.5890A>G / K1964E: Classification=Functional; Confidence=Computational prediction; DeepATM_predicted=Yes. The row is explicitly a prediction and not a direct functional assay result.

Still missing

• No entry for c.5890A>G, p.Lys1964Glu, p.K1964E, or K1964E was found in clingen_hbop_atm_supplementary_tables_1_and_2_v1.xlsx.txt.
• The available S1 row is computationally predicted rather than directly assayed; a qualifying variant-specific ATM functional assay result is still needed.

Sources

Reasoning

The ATM VCEP permits PS4 only for a case-control study with p-value ≤.05 and either an odds ratio, hazard ratio, or relative risk ≥2 or a lower 95% CI ≥1.5. The reviewed case-control papers evaluated ATM at the gene, variant-group, or different-variant level, but none of the available full-text extracts reports NM_000051.4:c.5890A>G (p.Lys1964Glu) in affected cases and controls with variant-specific counts, p-value, effect estimate, or confidence interval. Therefore the VCEP PS4 rule cannot be evaluated for this exact variant; no generic ATM or disease-level enrichment result was substituted.

VCEP rule

Case-control studies; p-value ≤.05 **and** (Odds ratio, hazard ratio, or relative risk ≥2 **or** lower 95% CI ≥1.5).

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 PS4 rule: case-control studies require p-value ≤.05 and (odds ratio, hazard ratio, or relative risk ≥2 or lower 95% CI ≥1.5).
• Tavtigian et al. (PMID:19781682) pooled 1,544 cases and 1,224 controls from seven studies and screened an additional 987 cases and 1,021 controls, but the extracted full text does not mention c.5890A>G or p.K1964E; its evidence is for rare ATM missense subsets at gene level, not this variant.
• Goldgar et al. (PMID:21787400) analyzed 76 rare ATM variants in 2,570 cases and 1,448 controls, but the extracted full text does not mention c.5890A>G or p.K1964E and reports no significant association for the analyzed variant groups.
• Haiman et al. (PMID:23555315) analyzed 62 ATM coding variants in a multiethnic case-control study; the reported ATM OR=4.13, p=0.0065 was for a different variant, rs56009889 (p.Phe2307Leu), not c.5890A>G.
• PMID:17508274 and PMID:26324357 were checked because literature_pass indexed them for PS4; their extracts do not mention ATM c.5890A>G or p.K1964E and provide no variant-specific case-control evidence.

Still missing

• A case-control report specifically identifying ATM c.5890A>G or p.Lys1964Glu with affected-case and control counts, or a directly reported p-value and effect estimate with 95% CI, is required.

Sources

Reasoning

PM3 requires evidence for this recessive ATM disorder that the variant is detected in trans with a pathogenic variant, with the ATM VCEP PM3/BP2 table assigning points according to affected-proband phenotype and phase. The governing table ATM_PM3_BP2_1.6.pdf contains no entry for c.5890A>G, p.Lys1964Glu, p.K1964E, or K1964E, and the available case evidence does not document an affected A-T proband, a pathogenic variant in trans, phase, or qualifying homozygosity. Therefore no PM3 points or strength can be assigned.

VCEP rule

For recessive disorders, detected in trans with a pathogenic variant

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 PM3 rule: For recessive disorders, detected in trans with a pathogenic variant.
• ATM_PM3_BP2_1.6.pdf.txt was searched for c.5890A>G, p.Lys1964Glu, p.K1964E, and K1964E; no exact variant entry was found.
• The ATM VCEP PM3/BP2 table assigns affected A-T proband points by phenotype confidence and zygosity/phase, but the available case evidence supplies no qualifying affected-proband observation, pathogenic variant in trans, phase result, or qualifying homozygous observation.

Still missing

• No entry for c.5890A>G, p.Lys1964Glu, p.K1964E, or K1964E was found in ATM_PM3_BP2_1.6.pdf.txt.
• A qualifying affected A-T proband observation with phenotype classification and the other ATM allele's pathogenicity and phase is still needed.

Sources

Reasoning

The ATM VCEP provides PP1 only for an autosomal-recessive condition, with strength determined by segregation in affected relatives: supporting for 1, moderate for 2, and strong for at least 3 affected relatives. The available case evidence does not establish an autosomal-recessive ATM phenotype, identify affected relatives carrying the variant, or provide segregation results, so no PP1 strength can be assigned.

VCEP rule

AR Condition: Segregation in 1 affected relative; AR Condition: Segregation in 2 affected relatives; AR Condition: Segregation in ≥3 affected relatives

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 PP1 rules: AR condition; supporting for segregation in 1 affected relative, moderate for 2, and strong for at least 3 affected relatives.
• Goldgar et al. 2011 (PMID:21787400) evaluated family-based ATM analyses, but the paper does not explicitly mention c.5890A>G or p.(K1964E), so it supplies no variant-specific segregation result for this case.

Still missing

• Qualifying autosomal-recessive disease context is not documented.
• No variant-positive affected relatives or segregation count is available.

Sources

Reasoning

The governing ATM VCEP rule permits BS3 at Supporting when a variant rescues either an ATM-specific feature or radiosensitivity, with Moderate requiring rescue of both. The exact variant was not found in the ClinGen HBOP ATM approved functional-assay supplementary tables. Sun et al. 2025 Supplementary Table S1 contains the exact c.5890A>G / K1964E row, but it is marked Classification=Functional, Confidence=Computational prediction, and DeepATM_predicted=Yes; under that file's governing policy, this is a model prediction rather than a direct prime-editing assay and cannot be used as BS3. No qualifying variant-specific ATM functional assay result is therefore available to assess BS3.

VCEP rule

Use when a variant rescues EITHER an ATM specifc feature OR rescues radiosensitivity.

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 BS3 rule: use Supporting when a variant rescues either an ATM-specific feature or radiosensitivity; Moderate requires rescue of both.
• ClinGen HBOP ATM approved functional-assay supplementary tables: searched c.5890A>G, p.Lys1964Glu, p.K1964E, and K1964E; no exact variant entry was found.
• Sun et al. 2025 Supplementary Table S1, exact row ATM c.5890A>G / K1964E: Classification=Functional; Confidence=Computational prediction; DeepATM_predicted=Yes. The row is explicitly a prediction and not a direct functional assay result.

Still missing

• No entry for c.5890A>G, p.Lys1964Glu, p.K1964E, or K1964E was found in clingen_hbop_atm_supplementary_tables_1_and_2_v1.xlsx.txt.
• The available S1 row is computationally predicted rather than directly assayed; a qualifying variant-specific ATM functional assay result is still needed.

Sources

Reasoning

BP2 requires an observation in trans with a pathogenic variant for a fully penetrant dominant disorder or in cis with a pathogenic variant in any inheritance pattern; the ATM VCEP table then assigns benign points for qualifying unaffected non-A-T adults. The governing table ATM_PM3_BP2_1.6.pdf contains no entry for c.5890A>G, p.Lys1964Glu, p.K1964E, or K1964E, and the available case evidence does not document an unaffected adult, a pathogenic or likely pathogenic ATM variant, or cis/trans phase. Therefore no BP2 points or strength can be assigned.

VCEP rule

Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.

Evidence considered

• ClinGen HBOP ATM VCEP v1.6 BP2 rule: Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.
• ATM_PM3_BP2_1.6.pdf.txt was searched for c.5890A>G, p.Lys1964Glu, p.K1964E, and K1964E; no exact variant entry was found.
• The ATM VCEP BP2 table assigns points per unaffected non-A-T adult according to pathogenic/likely pathogenic co-occurrence and phase, but the available case evidence supplies no qualifying unaffected-adult observation, co-occurring pathogenic/likely pathogenic ATM variant, or cis/trans phase result.

Still missing

• No entry for c.5890A>G, p.Lys1964Glu, p.K1964E, or K1964E was found in ATM_PM3_BP2_1.6.pdf.txt.
• A qualifying unaffected adult observation, the co-occurring ATM variant's pathogenicity, and documented cis/trans phase are still needed.

Sources