NOTCH1 c.5885G>A, p.(Arg1962His)

Reasoning

The variant is rare in large population reference datasets: gnomAD v4.1 reports 5/1,611,804 alleles (AF 0.000310%; joint grpmax FAF 0.000443%) and no homozygotes. The highest reported dataset-specific point estimate, 0.01704% in 2/11,736 gnomAD-Canada NFE alleles, remains rare and has FAF95 0.00301%. This supports PM2 at supporting strength under the generic ACMG/AMP fallback.

Rule

Generic ACMG/AMP PM2, supporting: absent or rare in population controls; no NOTCH1-specific VCEP specification was retrieved.

Evidence considered

• gnomAD v4.1 observed 5/1,611,804 alleles (AF 0.000310%), joint grpmax FAF 0.000443%, and zero homozygotes.
• gnomAD v2.1 observed 3/247,012 alleles (AF 0.00121%) and zero homozygotes.
• gnomAD-Canada v1.0 observed 2/18,414 alleles (AF 0.01086% overall); its NFE stratum had AF 0.01704% with FAF95 0.00301% and zero homozygotes. The record is RF-flagged, so this higher point estimate is interpreted cautiously.

Still missing

• No NOTCH1-specific PM2 frequency cutoff was available; assessment uses the generic ACMG/AMP rare-in-controls framework.

Sources

Reasoning

No governing VCEP framework applies to NOTCH1 (cspec.found=false, vcep_materials.final_classification_framework.found=false; generic ACMG fallback in effect). NM_017617.5:c.5885G>A is a missense substitution (p.(Arg1962His)); computational-evidence assessment therefore proceeds on the missense path (REVEL) with SpliceAI used to check for a cryptic-splice hypothesis. REVEL score = 0.5, which falls between the ClinGen SVI-calibrated PP3 supporting threshold (REVEL >= 0.644) and BP4 supporting threshold (REVEL <= 0.290) (Pejaver et al. 2022, PMID 36413997) - an intermediate/inconclusive score that does not meet PP3. BayesDel score is present in source data (0.105612) but per this pipeline's calibration policy no verified published threshold/table citation is available for BayesDel, so it is treated as not_available and not used to score PP3. SpliceAI predicts negligible splice impact (max delta score = 0.003-0.004, with all predicted splice-altering positions 130-200+ bp away from the variant), arguing against a splice-disrupting mechanism, but a low SpliceAI score alone is evidence against splice disruption (relevant to BP4), not evidence toward PP3 deleteriousness. No multiple concordant lines of evidence support a deleterious computational prediction.

Rule

ClinGen SVI REVEL calibration (PMID 36413997): PP3 supporting requires REVEL >= 0.644; observed REVEL = 0.5 does not meet this threshold.

Evidence considered

• REVEL score = 0.5 for NM_017617.5:c.5885G>A (source: local REVEL v1.3 lookup, source_registry key 'revel'), which is below the PP3 supporting threshold of REVEL >= 0.644 established by Pejaver et al. 2022 (PMID 36413997, ClinGen SVI calibration of REVEL for PP3/BP4).
• SpliceAI predicts max delta score = 0.003-0.004 (DS_AG=0.0, DS_AL=0.004, DS_DG=0.0, DS_DL=0.003), well below any splice-altering threshold, with nearest predicted splice positions 130-206 bp away, indicating no cryptic splice site creation/loss - this rules out a splice-disruption mechanism but does not itself support PP3 deleteriousness for this missense change.
• BayesDel score = 0.105612 is present in source_registry but is excluded from scoring because no verified published calibration threshold/table for BayesDel is available to this pipeline.

Still missing

• REVEL score (0.5) falls in the intermediate/inconclusive zone between calibrated PP3 and BP4 thresholds
• BayesDel score present but lacks a citable published calibration threshold in this pipeline, so cannot be used

Sources

Reasoning

PP5 is not met because ClinVar has no exact-variant record for NOTCH1 c.5885G>A (p.Arg1962His), and therefore no exact-variant expert-panel Pathogenic/Likely pathogenic assertion.

Rule

Apply PP5 only for an exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification.

Evidence considered

• ClinVar exact-variant search returned no result; the submission audit is empty.

Sources

Reasoning

The observed population frequencies are far below a stand-alone benign frequency range. The largest robust reference-dataset frequency is 0.00645% in gnomAD v2.1 African/African American samples; gnomAD v4.1 reports a joint grpmax filtering allele frequency of 0.000443%.

Rule

Generic ACMG/AMP BA1: allele frequency is not high enough to independently establish benignity.

Evidence considered

• gnomAD v2.1 observed 3/247,012 alleles (AF 0.00121%), with the highest population AF of 1/15,512 (0.00645%) in African/African American samples and no homozygotes.
• gnomAD v4.1 observed 5/1,611,804 alleles (AF 0.000310%), with joint grpmax FAF 0.000443% and no homozygotes.

Sources

Reasoning

Observed frequencies do not exceed the generic benign-supporting frequency threshold used in the absence of a NOTCH1-specific specification. The highest observed frequency is 0.01704% in the gnomAD-Canada NFE stratum (2/11,736), with a 95% filtering allele frequency of 0.00301%; this is not sufficiently high to be greater than expected for NOTCH1-related disease.

Rule

Generic ACMG/AMP BS1: not met because the allele frequency is not greater than expected for the disorder.

Evidence considered

• gnomAD v4.1 observed AF 0.000310% overall and a joint grpmax FAF of 0.000443%, with no homozygotes.
• gnomAD-Canada v1.0 observed 2/18,414 alleles (AF 0.01086% overall); its NFE stratum had 2/11,736 alleles (AF 0.01704%) and FAF95 0.00301%, with no homozygotes.
• The gnomAD-Canada record is flagged RF; its higher point estimate was not used as a benign-frequency basis.

Still missing

• No NOTCH1 VCEP or other gene-specific BS1 maximum credible allele-frequency specification was available.

Sources

Reasoning

No governing VCEP framework applies to NOTCH1; generic ACMG fallback in effect. REVEL score = 0.5 for this missense variant does not fall below the ClinGen SVI-calibrated BP4 supporting threshold (REVEL <= 0.290) (Pejaver et al. 2022, PMID 36413997), so the missense computational-evidence path does not support BP4. SpliceAI predicts negligible splice impact (max delta score = 0.003-0.004), consistent with no splice disruption, but this single line of splice-prediction evidence is not sufficient on its own to satisfy BP4's 'multiple lines of computational evidence' requirement for a missense variant whose primary in-silico signal (REVEL) is inconclusive rather than benign-supporting. BayesDel score is present but excluded per calibration policy (no citable published threshold available).

Rule

ClinGen SVI REVEL calibration (PMID 36413997): BP4 supporting requires REVEL <= 0.290; observed REVEL = 0.5 does not meet this threshold.

Evidence considered

• REVEL score = 0.5 (source_registry key 'revel') is above the BP4 supporting threshold of REVEL <= 0.290 established by Pejaver et al. 2022 (PMID 36413997), so REVEL does not support BP4.
• SpliceAI max delta score = 0.003-0.004 (source_registry key 'spliceai') indicates no predicted splice disruption, but this alone is insufficient to establish the multiple concordant computational lines of evidence generic ACMG BP4 requires when the missense predictor (REVEL) is inconclusive.
• BayesDel score = 0.105612 is excluded from BP4 scoring because no verified published calibration threshold is available to this pipeline for BayesDel.

Still missing

• REVEL score (0.5) is intermediate and does not reach the BP4 supporting threshold
• Only one calibrated computational line (SpliceAI, benign-supporting) is usable; BayesDel cannot be scored, so 'multiple lines of evidence' for BP4 is not established

Sources

Reasoning

BP6 is not met because ClinVar has no exact-variant record for NOTCH1 c.5885G>A (p.Arg1962His), and therefore no exact-variant expert-panel Benign/Likely benign assertion.

Rule

Apply BP6 only for an exact-variant ClinVar expert-panel Benign or Likely benign classification.

Evidence considered

• ClinVar exact-variant search returned no result; the submission audit is empty.

Sources

Reasoning

PS1 requires a different nucleotide change at the same codon resulting in the identical amino acid change (p.Arg1962His) with an established pathogenic classification. No ClinGen VCEP (cspec) exists for NOTCH1 in this case (cspec.found = false), and the variant itself is absent from ClinVar (no_result). No source in the evidence bundle identifies any other nucleotide substitution encoding p.(Arg1962His) with a prior pathogenic/likely pathogenic classification, and no literature was returned (literature_pass.json: total_pmids = 0). There is no basis in the available evidence to assess PS1.

Evidence considered

• case_summary.json clinvar.evidence_sentence: 'This variant is absent from ClinVar.'
• literature_pass.json reports total_pmids = 0, no PMIDs found across evidence sources

Still missing

• No ClinVar record for any nucleotide variant producing p.Arg1962His to compare against
• No VCEP-specific PS1 rule available for NOTCH1
• No literature identified mentioning this exact amino-acid change

Sources

Reasoning

PS2 cannot be assessed: the available case-evidence bundle contains no proband phenotype, parental genotypes, parentage confirmation, or evidence that the variant arose de novo.

Rule

Generic ACMG/AMP PS2 requires a de novo occurrence with parental testing; no such case-level evidence is present.

Evidence considered

• No VCEP/gene-specific specification was retrieved; generic ACMG/AMP evidence interpretation applies.
• The consolidated bundle contains variant normalization and population/database evidence but no case-level parental testing, parentage, or proband phenotype/diagnostic information.

Still missing

• Obtain the proband's phenotype and diagnostic context.
• Obtain validated maternal and paternal testing for NM_017617.5:c.5885G>A, including parentage confirmation if available.

Reasoning

No VCEP/gene-specific framework is available for NOTCH1 in this case (framework_mode: generic_acmg, vcep_materials.json notes 'no cspec url available'). No functional/biochemical assay data for NM_017617.5:c.5885G>A (p.Arg1962His) is present anywhere in the evidence bundle: literature_pass.json found 0 PMIDs ('No PMIDs found across evidence sources'), fulltext_map.json is empty, and OncoKB explicitly states it 'did not identify variant-specific reviewed functional evidence for this variant.' There is no basis to assess PS3.

Evidence considered

• literature_pass.json: total_pmids=0, note='No PMIDs found across evidence sources.'
• vcep_materials.json: note='no cspec url available'; final_classification_framework.found=false, source='generic_acmg_fallback'.
• fulltext_map.json is empty ({}), confirming no full-text papers were retrieved for this case.
• OncoKB evidence_sentence: 'OncoKB did not identify variant-specific reviewed functional evidence for this variant; gene-level curated context is available for reviewer follow-up.'

Still missing

• No published functional/validated assay data (e.g., NOTCH signaling reporter assay, luciferase, biochemical activity assay) identified for this specific variant.
• No VCEP-specific PS3 assay calibration or approved assay list available for NOTCH1.
• Literature search returned zero PMIDs; a targeted search for functional characterization of p.Arg1962His would be needed to assess this criterion.

Reasoning

No variant-specific affected-case series, case-control dataset, or enrichment analysis was retrieved; PS4 cannot be assessed.

Rule

Generic ACMG/AMP PS4 requires statistically significant enrichment in affected individuals compared with controls.

Evidence considered

• The literature pass contains no PMIDs and no case-control or affected-cohort evidence for NOTCH1 c.5885G>A (p.Arg1962His).

Still missing

• Variant-specific affected and control allele counts, phenotype definition, ancestry matching, and statistical enrichment analysis are needed.

Reasoning

PM1 requires the variant to fall in a mutational hot spot or a well-established functional domain without benign variation. No ClinGen VCEP domain framework is available for NOTCH1 (cspec.found = false, vcep_materials.json note = 'no cspec url available'). The automated CancerHotspots.org lookup for NOTCH1 R1962 returned no_result ('This variant does not lie in a statistically significant hotspot'), and no Pfam/UniProt/InterPro domain annotation for residue 1962 was retrieved in the evidence bundle. OncoKB returned only gene-level, non-variant-specific context describing NOTCH1 as a receptor/transcription factor that can act as oncogene or tumor suppressor, with no domain- or hotspot-specific statement about residue 1962. Without domain-boundary or hotspot-enrichment data, PM1 cannot be assessed.

Evidence considered

• evidence.json hotspots.evidence_sentence: 'This variant does not lie in a statistically significant hotspot.' (CancerHotspots.org, search_status=no_result)
• evidence.json oncokb.evidence_sentence: OncoKB did not identify variant-specific reviewed functional evidence; only gene-level description of NOTCH1 as receptor/transcription factor with dual oncogene/tumor-suppressor roles

Still missing

• No protein domain/structural annotation for residue 1962 retrieved
• CancerHotspots.org query returned no_result for NOTCH1 R1962
• No VCEP-defined critical domain/hotspot list available for NOTCH1

Sources

Reasoning

PM3 cannot be assessed. The case bundle contains no affected-proband observation of NM_017617.5:c.5885G>A with a second NOTCH1 pathogenic/likely pathogenic variant, no phase data, and no evidence for a relevant recessive NOTCH1 disease mechanism.

Evidence considered

• No NOTCH1 CSPEC/VCEP framework was available in the case bundle, so generic ACMG/AMP criteria were considered.
• The ClinVar query for NM_017617.5:c.5885G>A returned no result and supplied no affected-proband, second-allele, or phase information.
• No PMIDs were identified in the literature packet, and the bundle contains no publication documenting this variant in trans with a pathogenic/likely pathogenic NOTCH1 allele in an affected individual.

Still missing

• Affected-proband genotype data identifying any second NOTCH1 variant.
• Parental or read-based phasing to establish whether the two variants are in trans.
• Evidence that the relevant NOTCH1 disorder is inherited recessively and that the second allele is pathogenic or likely pathogenic.

Sources

Reasoning

PM5 requires a different missense change at the same residue (p.Arg1962X, X not His) with an established pathogenic classification. The dedicated pm5_candidates.json artifact for this case explicitly returns found=false with zero same-residue candidates identified (candidate_summary.same_residue_candidates = 0) and states 'Unable to confirm classic same-residue PM5 semantics safely; skipping automatic candidate harvesting,' recommending level=not_applicable pending manual review. The variant itself is also absent from ClinVar, so no automated cross-check of alternate-residue changes was possible. No comparator pathogenic variant at codon 1962 is documented in the available evidence.

Evidence considered

• pm5_candidates.json: found=false, eligible_for_classic_pm5_search=false, same_residue_candidates=0, recommendation.level='not_applicable', reason='Unable to confirm classic same-residue PM5 semantics safely; skipping automatic candidate harvesting'
• case_summary.json clinvar.evidence_sentence: 'This variant is absent from ClinVar.'

Still missing

• pm5_candidates.json harvesting was skipped by the pipeline (eligible_for_classic_pm5_search = false); no automated same-residue comparator search was completed
• No ClinVar record exists for any alternate substitution at Arg1962
• Manual literature/ClinVar review of other p.Arg1962X changes recommended per pm5_candidates.json hermes_instruction

Sources

Reasoning

PM6 cannot be assessed: no reported de novo occurrence, parental testing result, or proband phenotype is available in the case-evidence bundle.

Rule

Generic ACMG/AMP PM6 requires an assumed de novo occurrence without confirmed parentage; no de novo case observation is present.

Evidence considered

• No VCEP/gene-specific specification was retrieved; generic ACMG/AMP evidence interpretation applies.
• The consolidated bundle contains no case-level report of a de novo occurrence or parental testing.

Still missing

• Obtain the proband's phenotype and diagnostic context.
• Obtain parental testing results; if a de novo event is reported without parentage confirmation, document the testing method and result.

Reasoning

PP1 cannot be assessed because no family pedigree, relative genotypes, affected/unaffected status, or informative meioses are provided.

Rule

Generic ACMG/AMP PP1 requires cosegregation of the variant with disease in multiple affected family members; no segregation data are present.

Evidence considered

• No VCEP/gene-specific specification was retrieved; generic ACMG/AMP evidence interpretation applies.
• The consolidated bundle contains no pedigree, family testing, phenotype segregation, or meiosis data.

Still missing

• Obtain a pedigree with affected and unaffected relatives.
• Obtain variant testing and phenotype status for informative relatives and count informative cosegregations/meioses.

Reasoning

PP2 requires the gene to have a low rate of benign missense variation and missense to be a common mechanism of disease, as established by gene-specific curation. No ClinGen VCEP specification is available for NOTCH1 in this case (cspec.found=false), and no missense constraint metric (e.g., gnomAD missense Z-score, ExAC/gnomAD constraint), missense-variant benign/pathogenic ratio, or NOTCH1-specific mechanism statement is present anywhere in the evidence bundle (prefetch, evidence.json, case_summary.json). NOTCH1-related disease mechanisms are heterogeneous across phenotypes (e.g., haploinsufficiency/loss-of-function in some cardiovascular and developmental phenotypes versus gain-of-function/dominant-negative missense in T-ALL and some structural disease), and the case does not specify an indication/phenotype context, so applying PP2 without gene-specific curation data would be speculative.

Evidence considered

• case_summary.json cspec.found = false (no ClinGen NOTCH1 VCEP specification in this case)
• vcep_materials.json note: 'no cspec url available'; final_classification_framework.source = 'generic_acmg_fallback'

Still missing

• No ClinGen VCEP or gene-specific missense-constraint curation available for NOTCH1
• No gnomAD missense Z-score or missense benign/pathogenic ratio retrieved in this case
• Indication/phenotype context not specified to determine which NOTCH1 disease mechanism (LOF vs GOF) applies

Reasoning

No phenotype for the tested individual and no phenotype-specific case evidence were provided, so phenotype specificity cannot be evaluated.

Rule

Generic ACMG/AMP PP4 requires a patient phenotype or family history highly specific for a disease with a single genetic etiology.

Evidence considered

• The consolidated bundle contains no proband phenotype or variant-specific clinical case report.

Still missing

• Detailed patient phenotype, indication for testing, and exclusion of alternative etiologies are needed.

Reasoning

No source provides phenotype-ascertained healthy adult carriers. Population datasets report no homozygotes, but NOTCH1-associated disease is not established here as a fully penetrant recessive early-onset condition for which absence of homozygotes would adjudicate BS2; population carrier counts alone cannot support BS2.

Rule

Generic ACMG/AMP BS2: requires observation in healthy adult individual(s) for a disorder expected to be fully penetrant at an early age; this evidence is unavailable.

Evidence considered

• gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada each report zero homozygotes for this allele, but do not provide the phenotype evidence required for BS2.

Still missing

• Phenotype- and age-documented unaffected adult carrier data are required to assess BS2.

Sources

Reasoning

Same evidentiary gap as PS3: no VCEP framework for NOTCH1, no PMIDs identified (literature_pass.json total_pmids=0), no full-text papers (fulltext_map.json empty), and OncoKB found no variant-specific functional data. No basis to assess whether functional studies show no damaging effect (BS3).

Evidence considered

• literature_pass.json: total_pmids=0, note='No PMIDs found across evidence sources.'
• vcep_materials.json: note='no cspec url available'; final_classification_framework.found=false, source='generic_acmg_fallback'.
• fulltext_map.json is empty ({}), confirming no full-text papers were retrieved for this case.
• OncoKB evidence_sentence: 'OncoKB did not identify variant-specific reviewed functional evidence for this variant; gene-level curated context is available for reviewer follow-up.'

Still missing

• No published functional/validated assay data showing normal/wild-type-like activity for p.Arg1962His identified.
• No VCEP-specific BS3 assay calibration available for NOTCH1.
• Literature search returned zero PMIDs; a targeted search would be needed to assess this criterion.

Reasoning

BS4 cannot be assessed because no tested affected relative lacking the variant, nor any other informative non-segregation observation, is provided.

Rule

Generic ACMG/AMP BS4 requires lack of segregation with disease; no informative non-segregation data are present.

Evidence considered

• No VCEP/gene-specific specification was retrieved; generic ACMG/AMP evidence interpretation applies.
• The consolidated bundle contains no family genotype-phenotype data supporting or refuting segregation.

Still missing

• Obtain pedigree, phenotype, and genotype data for informative affected and unaffected relatives.
• Document any confirmed non-segregation of NM_017617.5:c.5885G>A with the relevant phenotype.

Reasoning

BP1 requires the gene to be one in which disease is caused primarily by truncating/loss-of-function variants, making a missense variant less likely pathogenic. No gene-specific curation establishing NOTCH1's primary disease mechanism as truncating-only is available in this case (no cspec, no VCEP materials, no missense-constraint data). NOTCH1 pathogenic mechanisms reported in the literature span both loss-of-function (truncating/haploinsufficiency) and missense gain-of-function/dominant-negative effects depending on phenotype, so applying BP1 (which requires a truncating-predominant mechanism) cannot be supported without gene-specific data absent from this bundle.

Evidence considered

• case_summary.json cspec.found = false; vcep_materials.json note: 'no cspec url available'
• final_classification_framework.json (via vcep_materials.json embedded summary): source = 'generic_acmg_fallback', framework_mode = null, indicating no gene-specific mechanism framework was resolved for NOTCH1 in this case

Still missing

• No gene-specific curation confirming NOTCH1 disease is driven primarily by truncating variants
• No missense-constraint or variant-spectrum data retrieved to evaluate BP1 applicability
• No VCEP-specific BP1 rule available for NOTCH1

Reasoning

BP2 cannot be assessed. The case bundle contains no observation of NM_017617.5:c.5885G>A in cis with a pathogenic/likely pathogenic NOTCH1 variant, nor an observation in trans with such a variant in a context supporting a benign interpretation.

Evidence considered

• No NOTCH1 CSPEC/VCEP framework was available in the case bundle, so generic ACMG/AMP criteria were considered.
• The ClinVar query for NM_017617.5:c.5885G>A returned no result and supplied no co-occurrence or phase information.
• No PMIDs were identified in the literature packet, and the bundle contains no publication documenting a relevant cis or trans co-occurrence for this variant.

Still missing

• A documented co-occurring pathogenic/likely pathogenic NOTCH1 variant.
• Validated cis/trans phase information for the co-occurring variants.
• Phenotype, penetrance, and inheritance information required to interpret any trans observation for the relevant NOTCH1 disorder.

Sources

Reasoning

No affected individual, phenotype, or independently established alternative molecular diagnosis was provided; BP5 cannot be evaluated.

Rule

Generic ACMG/AMP BP5 requires a variant found in a case with an alternative molecular basis for disease.

Evidence considered

• The consolidated bundle contains no individual-level diagnostic findings or co-occurring molecular diagnosis.

Still missing

• An affected individual's phenotype and a documented alternative molecular diagnosis that fully explains it are needed.