Five-step scale from benign to pathogenic. This variant is uncertain. Select a step to read what it means.
This variantScenario
BenignPathogenic
Uncertain.The recorded classification is uncertain significance.
In plain language
Gene-level background from the recorded interpretation.
Words with a dotted underline have a short definition. Hover or tap them.
About POLE
POLE encodes the catalytic subunit of DNA polymerase epsilon, the enzyme that replicates the leading strand of DNA during cell division and participates in DNA repair. It contains a proofreading domain that corrects replication errors, keeping the accumulation of mutations in check. Germline mutations in POLE cause polyposis and predispose to colorectal cancer, and are also linked to a rare syndrome of facial dysmorphism, immunodeficiency, livedo, and short stature. Somatic mutations, particularly in the proofreading domain, occur in colorectal and endometrial cancers, where they drive an ultra-mutated tumor phenotype and are associated with better responses to immune checkpoint inhibitors.
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 León-Castillo et al. 2020 custom POLE framework. 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.
No pathogenic criterion marked as applied in the recorded assessment.
Benign criteria applied · 1
BP4Supporting
BP4 · supporting
Met (supporting): SpliceAI max delta 0.075 is below the <0.1 threshold, indicating no predicted splice impact.
Recorded classification: Uncertain significance
Only BP4 (supporting) is met: SpliceAI max delta 0.075 < 0.1; no pathogenic criterion is met, and this single supporting benign criterion satisfies neither the Likely Benign nor the Benign rules, yielding VUS.
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
119 / 1,602,392 alleles
gnomAD v4.1: 119 / 1,602,392 alleles. Recorded AF 7.4264e-05. Recorded homozygote count 0.
gnomAD v2.1
This variant is present in gnomAD v2.1 (AF= 0.000233588; MAF= 0.02336%, 65/278268 alleles, homozygotes = 0) and has highest observed frequency in the East Asian population (AF= 0.00297229; MAF= 0.29723%, 59/19850 alleles, homozygotes = 0); grpmax FAF= 0.00227514.
gnomAD-Canada
This variant is present in gnomAD-Canada v1.0 (AF= 0.0007057546145494029, 13/18420 alleles, homozygotes = 0).
ACMG/AMP criteria under León-Castillo et al. 2020 custom POLE framework, with the data and sources behind each one.
28 criteria with recorded statuses: 1 met · 2 not met · 17 not assessed · 8 not applicable. Open any row for the full reasoning, the rule and the evidence behind it.
No official ClinGen CSPEC/VCEP specification exists for POLE (cspec found=false); the governing framework is the local custom POLE framework (Leon-Castillo et al. 2020 package, framework_mode custom_gene_framework), which customizes only PM1, PS4, PP3 and BP4 - all restricted to exact missense variants. It contains no PVS1, PM4 or BP3 customization, so generic ACMG/AMP 2015 rules (PMID 25741868) with the ClinGen SVI PVS1 guidance (Abou Tayoun et al. 2018, PMID 30192042/PMC6185798) govern this consequence/LoF group.
NM_006231.4:c.6658-19C>G is an intronic single-nucleotide substitution (19 bp upstream of an exon boundary; VariantValidator intronic context '47i') with predicted protein consequence NP_006222.2:p.? for all transcripts. It is not a nonsense, frameshift, initiation-codon, or canonical +/-1,2 splice variant, and SpliceAI predicts no splice impact (max delta 0.075). With no RNA/minigene evidence of aberrant splicing, there is no predicted or demonstrated null (LoF) transcript: PVS1 is not met.
Because the variant is an intronic SNV rather than an in-frame indel or stop-loss change, neither PM4 (protein length change in a non-repeat region) nor BP3 (in-frame indel in a repeat region without known function) can apply by consequence class; both are recorded not_applicable, and neither contributes evidence in either direction.
The pipeline's variant-level PVS1 scaffold (pvs1_variant_assessment) independently reaches the same conclusion (bucket 'other', canonical_splice_consensus=false, apply_generic_pvs1_framework=false); the gene-context gate (pvs1_gene_context) is not relied upon for the PVS1 call because its supporting literature does not robustly establish germline POLE LoF causality.
ClinVar Benign/Likely benign submissions, gnomAD population frequencies, and the pvs1 gene-context scaffold were not used to infer any PVS1/PM4/BP3 call; determinations rest on the variant's sequence consequence, transcript annotation, splice prediction, and the governing generic rule definitions.
NM_006231.4:c.6658-19C>G is an intronic SNV 19 bp upstream of an exon boundary (VariantValidator exon context 47i) with predicted protein consequence NP_006222.2:p.? for all transcripts; it produces no amino acid substitution.
All five codon/residue/domain-group criteria (PS1, PM1, PM5, PP2, BP1) are premised on an amino-acid-level (missense) change: PS1/PM5 compare the changed residue to established pathogenic substitutions, PP2/BP1 require the variant to be a missense variant, and PM1 locates an altered residue within a hotspot or critical functional domain.
For PM1 specifically, vcep_materials.json domain_tables is empty ({}) for POLE (no authoritative VCEP domain entry list), and the governing local custom framework (León-Castillo et al. 2020-derived) limits PM1 to a defined set of exact exonuclease-domain missense substitutions, none of which is an intronic change at c.6658-19.
Accordingly, PS1, PM1, PM5, PP2, and BP1 are each recorded as not_applicable for this variant rather than met or not_met; no criterion in this group contributes evidence toward the classification.
This group covers the de novo and segregation criteria PS2, PM6, PP1, and BS4 for POLE NM_006231.4:c.6658-19C>G.
Governing framework check: no ClinGen POLE CSPEC is available (cspec.found=false); the local custom framework built from León-Castillo et al. 2020 customizes only PM1/PS4/PP3/BP4 and explicitly falls back to generic ACMG/AMP 2015 (PMID:25741868) for every other criterion, including all four in this group.
The case bundle contains no proband-specific or family-based data at all: no proband phenotype/genotype, no parental testing, no de novo assertion, no segregation or non-segregation meioses, and no variant-positive family reports in the literature.
Consequently PS2, PM6, PP1, and BS4 are each recorded as not_assessed because the requisite observations are absent; each criterion lists its specific missing data as gaps rather than defaulting to not_met, since absence of evidence in this variant-centric bundle cannot establish that the criteria fail.
No criterion in this group was inferred from any source's final classification (e.g., the ClinVar Likely benign/Benign aggregate was not converted into segregation or de novo evidence).
Paper-level data were taken from paper_extracts.json (PMID:25741868); no other PMID has an entry, so no other full text was treated as selectively read for this group.
Functional group covers PS3 and BS3 only. Governing framework check: no official ClinGen CSPEC for POLE; the local custom POLE framework (León-Castillo et al. 2020 package) supplies customized rules for PM1/PS4/PP3/BP4 only, so PS3/BS3 fall back to generic ACMG/AMP 2015 definitions (PMID:25741868).
Variant NM_006231.4:c.6658-19C>G is intronic (19 bp upstream of an exon; predicted protein p.?), so any PS3/BS3-grade functional evidence would need to come from a well-established RNA/splicing assay of this exact variant.
The evidence bundle contains no functional assay data at all for this variant: no RNA/cDNA or minigene splicing studies, no protein/enzyme assays; COSMIC has no somatic record; literature_pass and paper_extracts surfaced only the ACMG/AMP guideline (PMID:25741868) and Sherloc (PMID:28492532), neither of which mentions the variant or reports functional experiments.
SpliceAI (max delta 0.07) and ClinVar Likely benign/Benign labels were reviewed and explicitly excluded from PS3/BS3 because in-silico prediction and laboratory classifications are not well-established functional studies; they are handled under in-silico and population/other criteria groups.
Net contribution of the functional group: PS3 not_assessed and BS3 not_assessed; the group neither supports nor opposes pathogenicity and contributes no strength points to the final classification until functional (RNA/splicing) evidence for c.6658-19C>G is supplied.
The allelic group adjudicates PM3 (moderate pathogenic) and BP2 (supporting benign). Both are phase- and observation-dependent criteria: PM3 needs the variant in trans with a pathogenic variant in an affected proband in a recessive context; BP2 needs in-trans co-occurrence with a pathogenic variant in dominant disease or in-cis co-occurrence in any mode.
No such observation exists anywhere in this case: ClinVar VCV000246255 holds only per-submitter benign classifications without phase or co-occurrence detail, gnomAD reports heterozygous carriers with zero homozygotes (population-level only), and the only fetched literature (PMID 25741868, PMID 28492532) never mentions NM_006231.4:c.6658-19C>G.
Both PM3 and BP2 are therefore returned not_assessed: the variant contributes no allelic (phase/co-occurrence) evidence, pathogenic or benign, to the final ACMG/AMP combination. The determinant gap is the total absence of proband/family genotype data, phase testing, and any report of a second POLE allele in cis or trans.
Framework note: no ClinGen CSPEC and no gene-specific PM3/BP2 specification exist for POLE (the local custom framework covers only PM1, PS4, PP3, BP4), so both criteria were evaluated against generic ACMG/AMP 2015 (PMID 25741868).
No ClinGen CSPEC/VCEP rule specification exists for POLE (cspec found=false in prefetch); the governing framework is the local custom POLE framework derived from Leon-Castillo et al. 2020 (vcep_materials.json, framework_mode custom_gene_framework). Its PP3/BP4 rules apply only to exact missense variants present in Supplementary Tables S2/S3; NM_006231.4:c.6658-19C>G is deep intronic (c.6658-19, p.?) and absent from those tables, so per the framework's own fallback instruction the generic ACMG/AMP in-silico calibration (generic_acmg_combination_rules) governs PP3/BP4 for this variant.
This variant is an intronic substitution, not a missense substitution, so exactly one computational mechanism is evaluated per the generic calibration: the SpliceAI splice path. The REVEL missense-path branch does not apply, and BayesDel is never used for PP3/BP4 because no verified published calibration threshold/PMID is available to this pipeline.
SpliceAI max delta score = 0.075 (DS_AG=0.0, DS_AL=0.075, DS_DG=0.009, DS_DL=0.001), below the 0.1 BP4 supporting threshold and far below the 0.2 PP3 supporting threshold: BP4 is met at supporting strength and PP3 is not met. No prediction is double-counted across PP3/BP4/BP7.
BP7 is not applicable: generic ACMG/AMP BP7 covers synonymous variants only, this variant is intronic, and the local POLE custom framework defines no BP7 clause for intronic variants; the negative splice-prediction evidence is carried by BP4.
ClinVar submissions (5x Likely benign, 1x Benign, single-submitter) and the GeneDx submission's in-silico sentence were not used to infer any PP3/BP4/BP7 call; only primary computational evidence (SpliceAI) and the governing rule documents were applied.
No criteria match.
Met · 1
BP4SupportingMet (supporting): SpliceAI max delta 0.075 is below the <0.1 threshold, indicating no predicted splice impact.
Details ▾Hide ▴
Reasoning
Applying the generic ACMG/AMP splice-path threshold (the local custom POLE framework's PP3/BP4 rules are restricted to exact missense variants listed in its Supplementary Tables S2/S3, and this deep intronic variant is absent from them, triggering the framework's fallback to the generic in-silico workflow): SpliceAI max delta score = 0.075, below the BP4 supporting threshold of <0.1 for intronic/synonymous/non-canonical-splice-position variants, indicating no predicted splice impact. BP4 is met at supporting strength. REVEL and BayesDel do not contribute: the variant is not missense (REVEL not found, score null) and BayesDel has no verified published calibration threshold in this pipeline, so it is never used for PP3/BP4.
Rule
generic_acmg_pp3_bp4_splice_path
Evidence considered
SpliceAI Lookup result for NM_006231.4:c.6658-19C>G: DS_AG=0.0 (acceptor gain), DS_AL=0.075 (acceptor loss), DS_DG=0.009 (donor gain), DS_DL=0.001 (donor loss), max delta score = 0.075; tool evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.07).' All four delta scores are far below any published/operating threshold for predicted splice alteration (source: spliceai).
Generic ACMG/AMP in-silico calibration (output/generic_acmg_classification_rules.md, registered as generic_acmg_combination_rules, source basis PMID:25741868; SVI-recommended SpliceAI thresholds as reflected in the document, per Walker et al. 2023 referenced therein): SpliceAI max delta <0.1 -> BP4 supporting for intronic/synonymous/non-canonical-splice-position variants; max delta 0.075 satisfies this.
POLE local custom framework BP4 rule (REVEL class 'Likely benign' with >=4 benign in silico results for exact missense variants in Supplementary Tables S2/S3) does not apply: c.6658-19C>G is intronic and absent from the missense-only tables, so the framework's generic fallback governs (see vcep_materials.json final_classification_framework and index_guidance).
REVEL score null/not found and BayesDel not found in prefetch.json; BayesDel is treated as not_available for PP3/BP4 under this pipeline's policy regardless of score because no verified published calibration threshold/PMID is available.
paper_extracts PMID:25741868: generic ACMG/AMP guidance that in silico splice predictors have ~90-100% sensitivity but ~60-80% specificity, that multiple in silico programs count as a single piece of evidence, and that confirmation of splice impact requires functional RNA/protein analysis; used as context for treating the single SpliceAI prediction as one line of computational evidence.
PVS1Not met: the variant is not a nonsense, frameshift, or canonical splice-site change, and SpliceAI max delta 0.075 predicts no splicing effect.
Details ▾Hide ▴
Reasoning
PVS1 (generic ACMG/AMP, per the ClinGen SVI PVS1 recommendations) applies to predicted null variants: nonsense, frameshift, canonical +/-1,2 splice-site, initiation-codon, or exon-level loss-of-function changes in a gene where LoF is an established disease mechanism. NM_006231.4:c.6658-19C>G is a single-nucleotide substitution located 19 bp upstream of an exon boundary (VariantValidator intronic context '47i'; predicted protein NP_006222.2:p.?), so it is not a nonsense, frameshift, or initiation-codon change and it is NOT at a canonical +/-1,2 splice position (variant-level PVS1 scaffold records canonical_splice_consensus=false, variant_bucket='other', apply_generic_pvs1_framework=false). Reaching PVS1 for a non-canonical deep-intronic change would require functional demonstration of aberrant splicing producing a transcript predicted to undergo NMD; here SpliceAI predicts no splice impact (max delta score 0.075: DS_AG=0.0, DS_AL=0.075, DS_DG=0.009, DS_DL=0.001) and no RNA/minigene functional evidence exists in the case. The governing local custom POLE framework (Leon-Castillo et al. 2020 package) does not customize PVS1 (customized criteria are only PM1/PS4/PP3/BP4), so generic PVS1 rules apply. No predicted or demonstrated loss-of-function effect: PVS1 not met. Note the gene-context gate (pvs1_gene_context) flags LoF-mechanism eligibility but its supporting sources do not robustly concern POLE germline LoF; the call above rests on the variant-level absence of any null-variant consequence, independent of that gate.
Rule
generic_acmg_pvs1_null_variant_rule_svi
Evidence considered
VariantValidator normalization (prefetch.json): NM_006231.4:c.6658-19C>G is an intronic SNV 19 bp 5' of an exon boundary (intronic exon context '47i'), predicted protein consequence NP_006222.2:p.? for all transcripts - not a nonsense, frameshift, initiation-codon, or canonical splice change.
Variant-level PVS1 scaffold (pvs1_variant_assessment): consequence_class 'unknown', canonical_splice_consensus=false, variant_bucket='other', apply_generic_pvs1_framework=false, no suggested default strength - the variant is not in any default generic PVS1 null-variant bucket (nonsense, frameshift, canonical +/-1,2 splice).
SpliceAI Lookup (spliceai) for NM_006231.4:c.6658-19C>G: DS_AG=0.0, DS_AL=0.075, DS_DG=0.009, DS_DL=0.001, max delta score 0.075 ('SpliceAI predicts no significant splice impact'); per the SpliceAI method paper (Jaganathan et al. 2019, PMID 30661751) delta scores >=0.2 denote predicted splice alteration, so 0.075 provides no support for native-acceptor loss that could generate a null transcript.
ClinGen SVI PVS1 recommendations (pvs1_generic_framework; Abou Tayoun et al. 2018 Hum Mutat, PMID 30192042, PMC6185798): PVS1 is reserved for null variants (nonsense/frameshift/canonical splice-site/exon-level, with NMD or truncation expectations and LoF-mechanism checks); non-canonical splice-region variants require demonstrated aberrant splicing/NMD, which is absent for this -19 intronic variant.
ACMG/AMP 2015 (PMID 25741868) defines PVS1 as a null variant (nonsense, frameshift, canonical +/-1/2 splice, initiation codon, single/multi-exon deletion) where loss of function is a known mechanism; the definitional source was consulted to confirm this deep intronic SNV is outside PVS1 scope without functional splice data.
pvs1_gene_context claims LoF-mechanism-supported/eligible for generic PVS1, but its listed supporting sources do not concern POLE germline LoF (PMID:41530594 is a MAN2B1 alpha-mannosidosis report; PMID:41551532 reports POLE variants only as VUS in Hungarian FAP patients; PMID:35395208 is FBXW7; PMID:37071997 is SRSF1; PMID:41673532 is CDKN2A-negative melanoma with MBD4/PMEL variants). Not relied upon for the PVS1 determination.
Local custom POLE framework (vcep_materials.json final_classification_framework) lists customized criteria PM1, PS4, PP3, BP4 only and provides no gene-specific PVS1 clause, so generic ACMG/AMP PVS1 with SVI guidance governs.
Still missing
No RNA/cDNA or minigene functional assay evidence to exclude a splicing effect of this deep intronic (-19) change; PVS1 via aberrant-splicing/NMD cannot be further evaluated without such data.
Gene-level PVS1 eligibility in pvs1_gene_context is supported by sources that do not robustly establish germline POLE loss-of-function causality (POLE germline disease is characterized predominantly by exonuclease-domain missense variants; the closest source reports POLE variants only as VUS in FAP patients).
PP3Not met: SpliceAI max delta 0.075 is below the >0.2 supporting threshold, indicating no predicted splice impact.
Details ▾Hide ▴
Reasoning
NM_006231.4:c.6658-19C>G is a deep intronic substitution (19 bp upstream of the exon boundary, variant_exonic_positions exon 47i, protein consequence NP_006222.2:p.?) - not a missense substitution and not a canonical +/-1,2 splice-site change. The local custom POLE framework (Leon-Castillo et al. 2020 package) PP3 rule applies only when the exact missense variant appears in Supplementary Tables S2/S3 with REVEL class 'likely disease causing' and <=1 benign in silico result; this intronic variant is absent from those missense-only tables, so per the framework's own fallback instruction the generic ACMG/AMP in-silico calibration governs. Under that calibration only the SpliceAI splice-path branch applies to intronic variants: SpliceAI max delta score is 0.075 (DS_AL), well below the PP3 supporting threshold of >0.2, so PP3 is not met.
Rule
generic_acmg_pp3_bp4_splice_path
Evidence considered
SpliceAI Lookup result for NM_006231.4:c.6658-19C>G: DS_AG=0.0, DS_AL=0.075, DS_DG=0.009, DS_DL=0.001, max delta score = 0.075; tool evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.07).' (source: spliceai).
Generic ACMG/AMP in-silico calibration (output/generic_acmg_classification_rules.md, registered as generic_acmg_combination_rules, source basis PMID:25741868) specifies for intronic/synonymous/non-canonical-splice-position variants: evaluate SpliceAI only - max delta >0.2 -> PP3 supporting, <0.1 -> BP4 supporting, [0.1,0.2] gray zone; the REVEL missense-path rule does not apply to this intronic variant and BayesDel is never used because no verified published threshold/PMID is available.
POLE local custom framework (vcep_materials/final_classification_framework, Leon-Castillo et al. 2020-derived) customizes PP3/BP4 only for exact missense variants present in Supplementary Tables S2 (PATH-250-323-s003.xlsx) and S3 (PATH-250-323-s004.xlsx); those tables list in silico results for POLE missense mutations, and c.6658-19C>G (intronic, p.?) is not among them, so the custom PP3/BP4 rule cannot be applied and the framework's fallback to the generic in-silico workflow applies.
REVEL score not found (null) and BayesDel score not found in prefetch/evidence for this variant; both are inapplicable here because the variant has no missense consequence, and BayesDel additionally has no verified published calibration threshold available to this pipeline.
PS2Not assessed: no proband description, parental genotypes, or de novo observations were available.
Details ▾Hide ▴
Reasoning
No proband-specific data exist in this case bundle: there is no affected proband description, no phenotype-consistency statement, and no parental testing result (no maternal/paternal genotypes, no maternity/paternity confirmation). The ClinVar record for the exact variant (ClinVar ID 246255; 6 submissions, all germline origin, 5 Likely benign + 1 Benign, review status 'criteria provided, single submitter' / one 'no assertion criteria provided') contains no de novo assertion field and no family/genotype observations. Prefetch sources did not include any de novo database (steps: normalize, hci_prior, revel, cspec, bayesdel, clinvar, pmid_metadata), and the only full-text paper in the registry (PMID:25741868) does not mention the variant. PS2 (de novo with confirmed parentage) therefore cannot be evaluated. The ClinVar Benign/Likely benign aggregate classification was not used to infer anything about de novo status, per the rule against deriving a criterion from another source's final classification.
Rule
PS2 (generic ACMG/AMP 2015, Richards et al., PMID:25741868): de novo variant (maternity and paternity confirmed); applied at reduced strength when parentage is not confirmed. No POLE-specific PS2 specification exists in the governing custom framework.
Evidence considered
ClinVar variation 246255: 6 submissions for NM_006231.4(POLE):c.6658-19C>G, all with germline origin (5 Likely benign, 1 Benign; one submission origin unknown); no submission carries a de novo assertion or parental-testing observation.
POLE has no official ClinGen CSPEC in this case (cspec.found=false); the governing local custom framework (León-Castillo et al. 2020-derived) customizes only PM1/PS4/PP3/BP4 and explicitly falls back to generic ACMG/AMP 2015 for all other criteria, so PS2 is defined per Richards et al. 2015 (PMID:25741868): de novo with maternity and paternity confirmed.
paper_extracts for PMID:25741868 confirms the exact variant c.6658-19C>G is not mentioned anywhere in that text; it provides no de novo or case-level observations.
No case reports, cohort reports, or functional studies in this case's source registry describe the variant as de novo in a phenotype-consistent proband.
Still missing
No proband genotype/phenotype data supplied for this case (variant-centric review only).
No parental (maternal/paternal) testing results are available to confirm or refute a de novo occurrence.
No de novo-specific database (e.g., DenovoDB-style source) was prefetched for this variant.
PS3Not assessed: no functional study of this exact variant - RNA analysis, minigene assay, or other - was available.
Details ▾Hide ▴
Reasoning
No ClinGen CSPEC exists for POLE and the governing local custom POLE framework (León-Castillo et al. 2020 package) customizes only PM1/PS4/PP3/BP4, so the generic ACMG/AMP 2015 definition applies: PS3 requires well-established functional studies demonstrating a damaging effect on gene/protein function. The variant is intronic (NM_006231.4:c.6658-19C>G, intron 47, 19 bp upstream of an exon, predicted p.?), so PS3-relevant functional evidence would have to come from a well-established RNA/splicing assay of this exact variant. The evidence bundle contains no such study: no patient RNA/cDNA analysis, minigene splice assay, or any POLE functional assay for this variant was fetched or cited. The only variant-level signal with any functional-adjacent relevance is SpliceAI (in silico, max delta score 0.07), which is explicitly not functional evidence under ACMG/AMP 2015 (in-silico splice predictions require RNA/protein confirmation of actual impact). ClinVar submissions (5x Likely benign, 1x Benign) cite only the ACMG/AMP guideline and Sherloc papers and report no functional assay; no PS3 evidence signal appears in any submission. PS3 is therefore not assessable from available evidence.
Evidence considered
Governing framework: ClinGen CSPEC for POLE not found (cspec found=false); the local custom POLE framework package (vcep_path_250_323, León-Castillo et al. 2020) specifies customized rules only for PM1/PS4/PP3/BP4 and contains no PS3/BS3 functional-evidence specification, so generic ACMG/AMP 2015 PS3/BS3 definitions apply (PMID:25741868).
Variant is intronic (c.6658-19C>G, 19 bp upstream of an exon per VariantValidator intronic context; predicted protein p.?); PS3-relevant functional evidence would require a well-established RNA/splicing assay performed on this exact variant.
Full-text extraction of PMID:25741868 (ACMG/AMP standards and guidelines) confirms it contains no POLE-specific or variant-specific functional data and that the exact variant c.6658-19C>G is not mentioned; the paper's only relevant guidance is that splice prediction tools are in silico and that confirmation of actual splice impact requires functional RNA or protein analysis.
SpliceAI (in silico splice predictor) gives max delta score 0.07 for this variant, predicting no significant splice impact; an in-silico prediction is not a well-established functional study and cannot satisfy PS3 under ACMG/AMP 2015 (PMID:25741868).
ClinVar variation 246255 (5x Likely benign, 1x Benign submissions, single-submitter review status): reviewed submissions show no functional assay reported and no PS3/BS3 evidence signals; the only cited PMIDs are the ACMG/AMP guideline (25741868) and Sherloc (28492532), both criteria-guidance papers with no variant-level functional data.
Still missing
No functional assay evidence for c.6658-19C>G was identified: no RNA-based splicing analysis of the exact variant (patient RNA/cDNA or minigene splice assay) and no other well-established POLE functional study. PS3 remains not_assessed until such a validated assay of this exact intronic variant is provided.
PS4Not assessed: insufficient evidence was available.
Details ▾Hide ▴
Reasoning
PS4 could not be assessed because its agent did not produce valid output.
Still missing
attempt 1: Hermes timed out after 200s and left no usable output_file
attempt 2: Hermes timed out after 200s and left no usable output_file
attempt 3: Hermes timed out after 200s and left no usable output_file
PM2Not assessed: insufficient evidence was available.
Details ▾Hide ▴
Reasoning
PM2 could not be assessed because its agent did not produce valid output.
Still missing
attempt 1: Hermes timed out after 200s and left no usable output_file
attempt 2: Hermes timed out after 200s and left no usable output_file
attempt 3: Hermes timed out after 200s and left no usable output_file
PM3Not assessed: no affected-proband or allele-phase data was available, and this POLE disorder is autosomal dominant rather than recessive.
Details ▾Hide ▴
Reasoning
PM3 (moderate, pathogenic) requires, for a recessive disorder, observation of the variant in trans with a pathogenic variant in an affected proband. No ClinGen CSPEC exists for POLE, and the local custom POLE framework (León-Castillo et al. 2020-derived, custom_internal) specifies rules only for PM1, PS4, PP3 and BP4, so PM3 falls back to generic ACMG/AMP 2015 (PMID 25741868, Table 3). No source in the case bundle reports any affected-proband observation of NM_006231.4:c.6658-19C>G, alone or with a second POLE allele: ClinVar VCV000246255 (6 germline submissions: 5 Likely benign, 1 Benign; no expert-panel submissions) contains per-submitter classifications only, with no allele-phase or co-occurrence content in any criterion-lead submission; the only fetched literature (PMID 25741868, PMID 28492532) is method guidance that never mentions this variant (paper_extracts confirms variant_mentioned=false for PMID 25741868); COSMIC is negative (somatic context) and is not a germline allelic source. gnomAD v2.1/v4.1/Canada show heterozygous population carriers only (0 homozygotes in 65/278,268, 119/1,602,392 and 13/18,420 alleles respectively), which is population frequency data, not an affected-proband in-trans observation. Additionally, the ClinVar condition context for this variant is 'Colorectal cancer, susceptibility to, 12' (autosomal dominant POLE disease), and PM3 is designed for recessive disorders, so rule applicability itself is uncertain here. Evidence is therefore insufficient in both directions: PM3 is not met, but the absence of any proband/family/phase data means the variant cannot be declared to have failed PM3 either. Returned not_assessed with the gaps listed.
Rule
Generic ACMG/AMP 2015 (Richards et al., PMID 25741868, Table 3) PM3: for recessive disorders, detected in trans with a pathogenic variant (Pathogenic Moderate). No gene-specific POLE PM3 specification exists (local POLE framework customizes only PM1/PS4/PP3/BP4; no ClinGen CSPEC retrieved).
Evidence considered
Rule basis: no ClinGen CSPEC for POLE and the local custom POLE framework (León-Castillo et al. 2020-derived, v1) specifies customized rules only for PM1, PS4, PP3 and BP4; PM3 therefore falls back to generic ACMG/AMP 2015 (Richards et al., PMID 25741868, Table 3): 'PM3 - For recessive disorders, detected in trans with a pathogenic variant' at moderate strength.
ClinVar VCV000246255 (6 germline submissions: 5 Likely benign, 1 Benign; review status 'criteria provided'; no expert-panel submissions) was reviewed: none of the four criterion-lead submissions (SCV004017061, SCV005872673, SCV000806845, SCV001730159) or their validated PMID trails (25741868, 28492532) provide allele-phase, co-occurrence, or affected-proband observations of c.6658-19C>G; only per-submitter benign classifications are recorded.
gnomAD v4.1 (119/1,602,392 alleles, AF 0.00743%, 0 homozygotes; EAS AF 0.1428%) and gnomAD v2.1 (65/278,268 alleles, AF 0.0234%, 0 homozygotes; EAS AF 0.2972%) document heterozygous population carriers only; population cohorts cannot supply affected-proband in-trans observations, and the zero homozygote counts provide no biallelic observation of this variant.
The ClinVar condition context for this variant is 'Colorectal cancer, susceptibility to, 12', an autosomal dominant POLE tumor-susceptibility phenotype; PM3 is intended for recessive disorders, so applicability of the rule would in any case require a recessive disease context not present in this case's evidence.
Literature coverage for this case is limited to two method papers (PMID 25741868 ACMG/AMP 2015 and PMID 28492532 Sherloc) retrieved via the ClinVar submission trail; paper_extracts for PMID 25741868 confirms the exact variant c.6658-19C>G is not mentioned, and no case-level in-trans observation of this variant appears in any fetched source.
Still missing
No affected-proband observation of c.6658-19C>G in any zygosity: the case bundle contains no proband phenotype/genotype or family data for this variant
No phase information: the variant has never been reported in trans with a pathogenic POLE variant in an affected individual (no such case report exists in ClinVar submissions or the fetched literature)
No second (likely) pathogenic POLE allele documented in any proband carrying c.6658-19C>G
POLE germline disease in this variant's ClinVar context (colorectal cancer susceptibility 12) is autosomal dominant, so the recessive-oriented PM3 rule would require a recessive disease context that is not established here
No VCEP/CSPEC or gene-specific PM3 specification exists for POLE; the local custom framework customizes only PM1/PS4/PP3/BP4, leaving only generic ACMG/AMP 2015 PM3
PM6Not assessed: no proband genotype, family testing, or assumed-de-novo observation was available.
Details ▾Hide ▴
Reasoning
PM6 requires an assumed de novo occurrence in a proband without confirmation of maternity/paternity. The case bundle contains no proband genotype, no triofamily testing, and no statement of an assumed de novo event for c.6658-19C>G. ClinVar submissions for this exact variant report only germline carrier observations with Likely benign/Benign assertions and no de novo or parental data; literature records in the registry do not mention the variant. With zero meioses or parental-genotype information available, PM6 cannot be assessed.
Rule
PM6 (generic ACMG/AMP 2015, Richards et al., PMID:25741868): assumed de novo occurrence, without confirmation of maternity and paternity. No POLE-specific PM6 specification exists in the governing custom framework.
Evidence considered
ClinVar variation 246255 submissions were reviewed for any assumed-de-novo or parental-testing annotation; none of the 6 submissions (SCV004017061, SCV005872673, SCV000293730, SCV000806845, SCV000489617, SCV001730159) provides such data, and none carries an affected-proband observation.
The governing framework for this case is the local custom POLE framework, which does not customize de novo criteria; PM6 is therefore defined by generic ACMG/AMP 2015 (PMID:25741868): assumed de novo, without confirmation of maternity and paternity.
PMID:25741868 full-text extraction shows the exact variant c.6658-19C>G is not mentioned; no proband-based de novo evidence can be drawn from the reviewed literature.
Still missing
No proband-parent trio genotype data (an assumed de novo event cannot be documented).
No maternal/paternal testing and no confirmation status for any parental genotypes.
No literature case report of this variant in a proband that would support an assumed de novo occurrence.
PP1Not assessed: no segregation data - affected-family genotypes or informative meioses - was available.
Details ▾Hide ▴
Reasoning
PP1 requires co-segregation of the variant with disease in affected family members in a gene definitively known to cause the disease. No segregation observations exist for c.6658-19C>G: there are no affected-family-member genotypes, no informative meioses, no pedigree data, and no segregation statement in any registry source. The ClinVar record lists only population-style laboratory submissions (conditions: colorectal cancer susceptibility 12 / not provided / not specified) with no segregation counts. The literature triage note references a 'segregation' theme, but no qualifying paper was retained (only PMID:25741868 and PMID:28492532, both methodology papers that do not mention the variant). PP1 cannot be assessed without segregation data.
Rule
PP1 (generic ACMG/AMP 2015, Richards et al., PMID:25741868): co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease. No POLE-specific PP1 specification exists in the governing custom framework.
Evidence considered
ClinVar variation 246255 contains no co-segregation data: submissions do not report affected-relative genotypes or segregation counts for c.6658-19C>G.
literature_pass.json retained 2 papers (PMID:25741868, PMID:28492532), both methodological; paper_extracts confirms the variant is absent from PMID:25741868, and neither provides familial segregation observations.
PP1 definition applied per generic ACMG/AMP 2015 (PMID:25741868): co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease; the local POLE custom framework does not modify PP1.
The ClinVar Likely benign/Benign assertions were not used to infer segregation status (final classifications from other groups are not evidence for this criterion).
Still missing
No affected family members genotyped for this variant; zero segregation observations/meioses available.
No pedigree or familial case report describing c.6658-19C>G co-segregating with disease.
No POLE-specific PP1 rule or segregation threshold is defined in the governing custom framework (falls back to generic ACMG/AMP 2015).
PP4Not assessed: insufficient evidence was available.
Details ▾Hide ▴
Reasoning
PP4 could not be assessed because its agent did not produce valid output.
Still missing
attempt 1: Hermes timed out after 200s and left no usable output_file
attempt 2: Hermes timed out after 200s and left no usable output_file
attempt 3: Hermes timed out after 200s and left no usable output_file
PP5Not assessed: insufficient evidence was available.
Details ▾Hide ▴
Reasoning
PP5 could not be assessed because its agent did not produce valid output.
Still missing
attempt 1: Hermes timed out after 200s and left no usable output_file
attempt 2: Hermes timed out after 200s and left no usable output_file
attempt 3: Hermes timed out after 200s and left no usable output_file
BA1Not assessed: insufficient evidence was available.
Details ▾Hide ▴
Reasoning
BA1 could not be assessed because its agent did not produce valid output.
Still missing
attempt 1: Hermes timed out after 200s and left no usable output_file
attempt 2: Hermes timed out after 200s and left no usable output_file
attempt 3: Hermes timed out after 200s and left no usable output_file
BS1Not assessed: insufficient evidence was available.
Details ▾Hide ▴
Reasoning
BS1 could not be assessed because its agent did not produce valid output.
Still missing
attempt 1: Hermes timed out after 200s and left no usable output_file
attempt 2: Hermes timed out after 200s and left no usable output_file
attempt 3: Hermes timed out after 200s and left no usable output_file
BS2Not assessed: insufficient evidence was available.
Details ▾Hide ▴
Reasoning
BS2 could not be assessed because its agent did not produce valid output.
Still missing
attempt 1: Hermes timed out after 200s and left no usable output_file
attempt 2: Hermes timed out after 200s and left no usable output_file
attempt 3: Hermes timed out after 200s and left no usable output_file
BS3Not assessed: no well-established functional study of this exact variant was available to show an absence of damaging effect.
Details ▾Hide ▴
Reasoning
As for PS3, no applicable gene-specific PS3/BS3 specification exists (no ClinGen CSPEC; the local custom POLE framework customizes only PM1/PS4/PP3/BP4), so the generic ACMG/AMP 2015 definition applies: BS3 requires well-established functional studies showing no damaging effect on protein/gene function. For this intronic variant (c.6658-19C>G, 19 bp upstream of an exon), that would require a validated RNA/splicing assay demonstrating no effect on splicing of the exact variant. No such functional study exists in the evidence bundle. The aggregate ClinVar Likely benign/Benign labels (5+1 submissions) and the low SpliceAI max delta score (0.07) are not well-established functional studies and must not be used as BS3 evidence; population frequency and in-silico predictions are evaluated under their own criteria (BA1/BS1/BP4-type) and cannot substitute for the functional study BS3 demands. BS3 is therefore not assessable from available evidence.
Evidence considered
Governing framework: no POLE-specific PS3/BS3 rule exists in the local custom POLE framework (vcep_path_250_323) or in any ClinGen CSPEC (cspec found=false); generic ACMG/AMP 2015 BS3 definition (well-established functional studies show no damaging effect) applies (PMID:25741868).
Full-text extraction of PMID:25741868 confirms no functional study of this variant is reported in the ACMG/AMP guideline itself and that in-silico splice predictions are not evidence of splicing impact without RNA/protein functional confirmation.
SpliceAI max delta score 0.07 predicts no significant splice impact in silico; this is computational prediction, not a well-established functional study, and is insufficient for BS3 per ACMG/AMP 2015 (PMID:25741868).
ClinVar variation 246255 submissions (Likely benign x5, Benign x1) cite only ACMG/AMP (25741868) and Sherloc (28492532) guidance papers; none reports a functional assay of c.6658-19C>G, and no BS3 evidence signal was found in any submission - the ClinVar labels themselves are not functional evidence.
Still missing
No well-established functional study showing no damaging effect of c.6658-19C>G was identified (no RNA/splicing assay of the exact variant, e.g., patient RNA/cDNA or minigene analysis). BS3 remains not_assessed until such validated functional data exist; ClinVar benign/likely-benign labels and SpliceAI are not acceptable substitutes for BS3.
BS4Not assessed: no affected-family genotypes or non-segregation observations were available.
Details ▾Hide ▴
Reasoning
BS4 requires documented non-segregation of the variant with disease in affected family members (e.g., an affected relative who does not carry the variant). The bundle contains no affected-family-member genotypes and no non-segregation observations for c.6658-19C>G. Population frequency data (gnomAD v2.1 total AF 0.023%, v4.1 total AF 0.0074%, East Asian AF up to ~0.30%) address population-based benignity arguments under other criteria (e.g., BS1/BA1, outside this group) and are not non-segregation evidence. With no meioses showing an affected non-carrier, BS4 cannot be assessed.
Rule
BS4 (generic ACMG/AMP 2015, Richards et al., PMID:25741868): non-segregation of the variant with disease in affected family members. No POLE-specific BS4 specification exists in the governing custom framework.
Evidence considered
ClinVar variation 246255 submissions contain no familial non-segregation observations for c.6658-19C>G (only germline carrier submissions with Likely benign/Benign assertions).
BS4 definition applied per generic ACMG/AMP 2015 (PMID:25741868): non-segregation of the variant with disease in affected family members; the local POLE custom framework does not modify BS4.
gnomAD v2.1/v4.1 and gnomAD-Canada carrier frequencies were considered and excluded from BS4, because population allele frequency is not equivalent to familial non-segregation evidence (it belongs to population-frequency criteria outside this group).
paper_extracts for PMID:25741868 confirms no variant-level or family-level data for c.6658-19C>G in the reviewed literature.
Still missing
No affected relative tested as a non-carrier of c.6658-19C>G; zero informative meioses for non-segregation.
No pedigree or familial case-control data in any registry source.
No gene-specific BS4 specification exists; generic ACMG/AMP 2015 governs, but no qualifying observations are available to trigger it.
BP2Not assessed: no co-occurrence or allele-phase observation with a pathogenic POLE allele was available.
Details ▾Hide ▴
Reasoning
BP2 (supporting, benign) requires observation of the variant in trans with a pathogenic variant for a fully dominant disease/mechanism, or in cis with a pathogenic variant in any inheritance pattern. No ClinGen CSPEC exists for POLE and the local custom POLE framework specifies rules only for PM1, PS4, PP3 and BP4, so BP2 falls back to generic ACMG/AMP 2015 (PMID 25741868, Table 4). No source in the case bundle contains any co-occurrence or phase observation for NM_006231.4:c.6658-19C>G: ClinVar VCV000246255 submissions record only per-submitter benign classifications (no family/phase/co-occurrence detail), the only fetched papers (PMID 25741868, PMID 28492532) are method documents that do not mention the variant, and no proband-level report of this variant together with any second POLE allele exists anywhere in the bundle. gnomAD shows zero homozygotes, but absence of homozygotes in population cohorts is not an observation of in-trans (or in-cis) co-occurrence with a pathogenic allele in affected or carrier individuals, which is what BP2 requires. Evidence is insufficient in both directions, so BP2 is returned not_assessed rather than met or not_met.
Rule
Generic ACMG/AMP 2015 (Richards et al., PMID 25741868, Table 4) BP2: observed in trans with a pathogenic variant for a fully dominant disease/mechanism; or observed in cis with a pathogenic variant in any inheritance pattern (Benign Supporting). No gene-specific POLE BP2 specification exists (local POLE framework customizes only PM1/PS4/PP3/BP4; no ClinGen CSPEC retrieved).
Evidence considered
Rule basis: no ClinGen CSPEC for POLE and the local custom POLE framework (León-Castillo et al. 2020-derived, v1) specifies customized rules only for PM1, PS4, PP3 and BP4; BP2 therefore falls back to generic ACMG/AMP 2015 (Richards et al., PMID 25741868, Table 4): 'BP2 - Observed in trans with a pathogenic variant for a fully dominant disease/mechanism; or observed in cis with a pathogenic variant in any inheritance pattern' at supporting (benign) strength.
ClinVar VCV000246255 (6 germline submissions, 5 Likely benign and 1 Benign; no expert-panel submissions) was reviewed for co-occurrence/phase content: none of the submissions or their validated PMID citations (25741868, 28492532) report c.6658-19C>G together with any second POLE variant, in cis or in trans.
gnomAD v4.1 (119/1,602,392 alleles, 0 homozygotes) and gnomAD v2.1 (65/278,268 alleles, 0 homozygotes) were checked for biallelic carriers: none exist, but population-carrier absence does not substitute for the patient-level in-trans/in-cis co-occurrence observation that BP2 requires.
The case bundle contains no proband, family, or phase data for this variant at all (variant-only review; no phenotyping or segregation records), and the only fetched literature (PMID 25741868, PMID 28492532) is method guidance that never mentions the variant.
Still missing
No observation of c.6658-19C>G in trans with a pathogenic POLE variant in a patient, as required for BP2 in the dominant-disease context (colorectal cancer susceptibility 12)
No observation of c.6658-19C>G in cis with a pathogenic POLE variant in any proband or carrier
No familial segregation, phase, or parental-genotype testing data for this variant in the case bundle or in any ClinVar submission detail
No gene-specific POLE BP2 specification exists (local custom framework customizes only PM1/PS4/PP3/BP4; no ClinGen CSPEC retrieved)
BP5Not assessed: insufficient evidence was available.
Details ▾Hide ▴
Reasoning
BP5 could not be assessed because its agent did not produce valid output.
Still missing
attempt 1: Hermes timed out after 200s and left no usable output_file
attempt 2: Hermes timed out after 200s and left no usable output_file
attempt 3: Hermes timed out after 200s and left no usable output_file
BP6Not assessed: insufficient evidence was available.
Details ▾Hide ▴
Reasoning
BP6 could not be assessed because its agent did not produce valid output.
Still missing
attempt 1: Hermes timed out after 200s and left no usable output_file
attempt 2: Hermes timed out after 200s and left no usable output_file
attempt 3: Hermes timed out after 200s and left no usable output_file
Not applicable · 8· recorded as not applicableShow reasons ▾Hide reasons ▴PS1N/ANot applicable: the intronic variant produces no amino acid change (p.?) to compare against an established pathogenic variant.
Details ▾Hide ▴
Reasoning
PS1 requires that the queried variant produce the same amino-acid change as a previously established pathogenic variant regardless of the nucleotide change. NM_006231.4:c.6658-19C>G is an intronic SNV located 19 bp upstream of an exon boundary (VariantValidator intronic exon context '47i'); the predicted protein consequence is NP_006222.2:p.? for all transcripts, so no amino acid substitution occurs. There is no amino acid change to compare against an established pathogenic variant, making PS1 inapplicable to this variant by definition.
Evidence considered
VariantValidator annotation: NM_006231.4:c.6658-19C>G is intronic (genome context NC_000012.12(NM_006231.4):c.6658-19C>G, exon context 47i) with predicted protein consequence NP_006222.2:p.? for all transcripts - no amino acid substitution is produced.
ACMG/AMP 2015 (PMID 25741868) defines PS1 as the same amino acid change as a previously established pathogenic variant; because this intronic variant has no amino acid change (p.?), the criterion cannot apply.
PM1N/ANot applicable: no amino acid residue is altered (p.?), so none can fall within a hotspot or critical domain.
Details ▾Hide ▴
Reasoning
PM1 (mutational hotspot / critical well-established functional domain) is a protein-level criterion: it requires the altered amino-acid residue to fall within a hotspot or critical domain. This variant is intronic (c.6658-19C>G) with no residue change (p.?), so no amino-acid residue is located in any domain. vcep_materials.json domain_tables is empty ({}) for POLE, providing no authoritative VCEP domain entry against which the variant's residue could be checked. The governing local POLE framework (León-Castillo et al. 2020-derived) restricts PM1 to exact exonuclease-domain missense substitutions (strong: p.P286R, p.V411L, p.S297F, p.A456P, p.S459F; moderate: p.F367S, p.L424I, p.M295R, p.P436R, p.M444K, p.D368Y; supporting: p.A465V, p.L424V, p.T278M, p.A428T); c.6658-19C>G is not a missense variant and is not among these exact substitutions, and its intronic position maps far downstream of the exonuclease-domain residues covered by the framework. PM1 is therefore not applicable.
Evidence considered
vcep_materials.json domain_tables is empty ({}) for POLE, so there is no authoritative VCEP domain entry list in which to place the variant's residue; PM1 could not be supported by a domain-table hit.
Local POLE custom framework derived from León-Castillo et al. 2020 (vcep_path_250_323) applies PM1 only to exact exonuclease-domain missense substitutions (strong tier p.P286R/p.V411L/p.S297F/p.A456P/p.S459F; moderate tier p.F367S/p.L424I/p.M295R/p.P436R/p.M444K/p.D368Y; supporting tier p.A465V/p.L424V/p.T278M/p.A428T) and explicitly not to other substitutions at the same codon; the queried intronic variant is not a missense change and is not among these variants.
VariantValidator reports the variant as intronic with predicted protein consequence NP_006222.2:p.?, so no amino-acid residue is altered and no residue can be assessed for location within a critical functional domain or hotspot.
ACMG/AMP 2015 (PMID 25741868) defines PM1 as location in a mutational hotspot and/or critical well-established functional domain; the definition operates at the protein-residue level, which is not reached by an intronic p.? variant.
PM4N/ANot applicable: a single-nucleotide intronic substitution, not an in-frame indel or stop-loss change, so no protein length change is produced.
Details ▾Hide ▴
Reasoning
PM4 (protein length change) applies to in-frame deletions/insertions in a non-repeat region or stop-loss variants that alter protein length. NM_006231.4:c.6658-19C>G is a single-nucleotide substitution in an intron (19 bp upstream of an exon boundary) with predicted protein consequence NP_006222.2:p.? for all transcripts: it is not an in-frame indel, not a stop-loss variant, and no protein length change is produced or predicted. The only route to a PM4-type protein length change would be splice-mediated in-frame exon skipping, which is unsupported here (SpliceAI max delta 0.075 predicts no splice impact) and unproven by any RNA study. The governing local custom POLE framework contains no PM4 customization. PM4 is therefore not applicable to this variant by consequence class.
Evidence considered
VariantValidator normalization (prefetch.json): NM_006231.4:c.6658-19C>G is an intronic SNV with predicted protein consequence NP_006222.2:p.? - no amino-acid insertion/deletion and no stop-loss, so no protein length change at the primary-transcript level.
ACMG/AMP 2015 (PMID 25741868) defines PM4 as protein length changes from in-frame deletions/insertions in a non-repeat region or stop-loss variants; an intronic SNV with no coding consequence cannot satisfy the in-frame indel/stop-loss precondition.
SpliceAI (spliceai) max delta 0.075 provides no prediction of splice alteration that could produce an in-frame exon-skip protein-length change; no RNA/cDNA or minigene evidence of an alternative in-frame transcript is present in the case.
Local custom POLE framework (vcep_materials.json) customizes only PM1/PS4/PP3/BP4 and contains no PM4 rule for POLE; generic PM4 definition applies unchanged.
PM5N/ANot applicable: no missense change occurs (p.?), so there is no residue to compare with an established pathogenic substitution.
Details ▾Hide ▴
Reasoning
PM5 requires a novel missense change at an amino-acid residue where a different missense change has previously been established as pathogenic. NM_006231.4:c.6658-19C>G is intronic and produces no amino acid change (NP_006222.2:p.?), so there is no missense change and no affected residue to compare with an established pathogenic substitution at that residue. PM5 is inapplicable by definition; no comparator search is meaningful for a non-coding change.
Evidence considered
ACMG/AMP 2015 (PMID 25741868) defines PM5 as a novel missense change at an amino acid residue where a different missense change was determined to be pathogenic; an intronic variant with no protein consequence (p.?) cannot satisfy the missense-change precondition.
VariantValidator annotation: NM_006231.4:c.6658-19C>G is intronic with predicted protein consequence NP_006222.2:p.? for all transcripts - no amino acid residue is changed, so no same-residue comparison applies.
PP2N/ANot applicable: the variant is intronic with no amino acid change, so it is not a missense variant.
Details ▾Hide ▴
Reasoning
PP2 is defined as a missense variant in a gene that has a low rate of benign missense variation and in which missense variants are a common mechanism of disease. The criterion is predicated on the queried variant itself being a missense variant. c.6658-19C>G is intronic with no amino acid change (p.?) and is therefore not a missense variant, so PP2 cannot be applied regardless of POLE's missense burden or mechanism.
Evidence considered
ACMG/AMP 2015 (PMID 25741868) defines PP2 as applying to a missense variant in a gene with a low rate of benign missense variation and where missense is a common disease mechanism; the criterion explicitly requires the variant under evaluation to be missense.
VariantValidator annotation: NM_006231.4:c.6658-19C>G is intronic with predicted protein consequence NP_006222.2:p.?, i.e. not a missense variant, so the PP2 precondition on variant type is not met.
BP1N/ANot applicable: not a missense variant (p.?), and POLE disease is driven by missense rather than truncating variants.
Details ▾Hide ▴
Reasoning
BP1 is defined as a missense variant in a gene for which primarily truncating variants are known to cause disease. The criterion is predicated on the queried variant being a missense variant. c.6658-19C>G is intronic with no amino acid change (p.?), so it is not a missense variant and BP1 cannot be applied; additionally, the framework materials describe POLE-associated disease as driven by exonuclease-domain missense variants, which would not satisfy BP1's truncation-mechanism premise in any case.
Evidence considered
ACMG/AMP 2015 (PMID 25741868) defines BP1 as a missense variant in a gene for which primarily truncating variants are known to cause disease; the criterion explicitly requires the variant under evaluation to be missense.
VariantValidator annotation: NM_006231.4:c.6658-19C>G is intronic with predicted protein consequence NP_006222.2:p.?, i.e. not a missense variant, so the BP1 precondition on variant type is not met.
Local POLE framework materials (vcep_path_250_323, León-Castillo et al. 2020) characterize pathogenic POLE variation as exonuclease-domain missense hotspots rather than primarily truncating changes, further inconsistent with BP1's premise.
BP3N/ANot applicable: the variant is a single-nucleotide substitution, not an in-frame indel in a repeat region.
Details ▾Hide ▴
Reasoning
BP3 (benign supporting) applies to in-frame deletions/insertions located in a repeat region without known function. NM_006231.4:c.6658-19C>G is a single-nucleotide substitution (not an indel) in an intronic position (c.6658-19, 19 bp upstream of an exon boundary), so the in-frame-indel precondition of BP3 is not met regardless of repeat-region context. Repeat-region context was nonetheless checked: gnomAD-Canada region flags for this locus report lcr=false and segdup=false, and no evidence in the case places the variant in a repetitive/low-complexity region with unknown function. The governing local custom POLE framework contains no BP3 customization. BP3 is therefore not applicable to this variant by consequence class.
Evidence considered
ACMG/AMP 2015 (PMID 25741868) defines BP3 as in-frame deletions/insertions in a repeat region without known function; the queried variant is an SNV, not an in-frame indel, so the criterion's precondition fails by variant type.
gnomAD-Canada v1.0 raw record (gnomad_canada) for 12-132625013-G-C reports region_flag lcr=false and segdup=false and variant_class SNV ('multi-snv'), providing no indication that the variant sits in a low-complexity/segmental-duplication repeat context relevant to BP3.
VariantValidator normalization (prefetch.json): NM_006231.4:c.6658-19C>G is an intronic SNV (intronic context '47i') with predicted protein consequence p.?, consistent with a non-indel, non-repeat-region change.
Local custom POLE framework (vcep_materials.json) customizes only PM1/PS4/PP3/BP4 and defines no BP3 clause for POLE; generic BP3 definition applies unchanged.
BP7N/ANot applicable: the variant is deep intronic rather than a synonymous coding change; the no-splice-impact evidence is captured under BP4.
Details ▾Hide ▴
Reasoning
BP7 as defined by generic ACMG/AMP (Richards et al. 2015, PMID:25741868) applies to synonymous (silent) coding variants for which splicing prediction algorithms predict no impact to the splice consensus sequence nor creation of a new splice site. NM_006231.4:c.6658-19C>G is a deep intronic substitution (19 bp upstream of the exon, protein consequence NP_006222.2:p.?), not a synonymous coding change. The local custom POLE framework customizes only PM1/PS4/PP3/BP4 and contains no gene-specific BP7 clause extending it to intronic variants. BP7 is therefore not applicable; the no-predicted-splice-impact evidence for this intronic variant is already captured under BP4 (met, supporting).
Evidence considered
Variant normalization (prefetch.json/variant_validator) shows NM_006231.4:c.6658-19C>G is an intronic substitution at exon 47 intronic position with predicted protein consequence NP_006222.2:p.? (not a synonymous/silent coding substitution), so BP7's synonymous-variant scope is not met.
Generic ACMG/AMP BP7 definition (Richards et al. 2015, PMID:25741868) is limited to synonymous variants; the local custom POLE framework (vcep_materials.json) lists customized criteria PM1, PS4, PP3, BP4 only and provides no BP7 rule for intronic variants, so generic BP7 scope applies unchanged.
AF 7.4e-05 · 119 / 1,602,392 alleles · 0 hom · FAF 0.0011
East Asian64 / 44,812
0.0014
Remaining individuals43 / 62,104
0.00069
Admixed American7 / 59,918
0.00012
African/African American2 / 74,814
2.7e-05
South Asian1 / 90,750
1.1e-05
European (non-Finnish)2 / 1,170,258
1.7e-06
0 alternate alleles recorded in 4 other groups: European (Finnish), Amish, Middle Eastern, Ashkenazi Jewish.
gnomAD v2.1
AF 0.00023 · 65 / 278,268 alleles · 0 hom · FAF 0.0023
East Asian59 / 19,850
0.003
Remaining individuals1 / 7,110
0.00014
Admixed American3 / 35,166
8.5e-05
African/African American1 / 24,638
4.1e-05
South Asian1 / 30,364
3.3e-05
0 alternate alleles recorded in 3 other groups: Ashkenazi Jewish, European (Finnish), European (non-Finnish).
gnomAD-Canada
AF not recorded · ? / ? alleles
Statements
gnomAD v4.1
This variant is present in gnomAD v4.1 (AF= 7.4264e-05; MAF= 0.00743%, 119/1602392 alleles, homozygotes = 0) and has highest observed frequency in the East Asian population (AF= 0.00142819; MAF= 0.14282%, 64/44812 alleles, homozygotes = 0); grpmax FAF= 0.0011474.
gnomAD v2.1
This variant is present in gnomAD v2.1 (AF= 0.000233588; MAF= 0.02336%, 65/278268 alleles, homozygotes = 0) and has highest observed frequency in the East Asian population (AF= 0.00297229; MAF= 0.29723%, 59/19850 alleles, homozygotes = 0); grpmax FAF= 0.00227514.
gnomAD-Canada
This variant is present in gnomAD-Canada v1.0 (AF= 0.0007057546145494029, 13/18420 alleles, homozygotes = 0).
BP4SupportingMet (supporting): SpliceAI max delta 0.075 is below the <0.1 threshold, indicating no predicted splice impact.
Details ▾Hide ▴
Reasoning
Applying the generic ACMG/AMP splice-path threshold (the local custom POLE framework's PP3/BP4 rules are restricted to exact missense variants listed in its Supplementary Tables S2/S3, and this deep intronic variant is absent from them, triggering the framework's fallback to the generic in-silico workflow): SpliceAI max delta score = 0.075, below the BP4 supporting threshold of <0.1 for intronic/synonymous/non-canonical-splice-position variants, indicating no predicted splice impact. BP4 is met at supporting strength. REVEL and BayesDel do not contribute: the variant is not missense (REVEL not found, score null) and BayesDel has no verified published calibration threshold in this pipeline, so it is never used for PP3/BP4.
Rule
generic_acmg_pp3_bp4_splice_path
Evidence considered
SpliceAI Lookup result for NM_006231.4:c.6658-19C>G: DS_AG=0.0 (acceptor gain), DS_AL=0.075 (acceptor loss), DS_DG=0.009 (donor gain), DS_DL=0.001 (donor loss), max delta score = 0.075; tool evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.07).' All four delta scores are far below any published/operating threshold for predicted splice alteration (source: spliceai).
Generic ACMG/AMP in-silico calibration (output/generic_acmg_classification_rules.md, registered as generic_acmg_combination_rules, source basis PMID:25741868; SVI-recommended SpliceAI thresholds as reflected in the document, per Walker et al. 2023 referenced therein): SpliceAI max delta <0.1 -> BP4 supporting for intronic/synonymous/non-canonical-splice-position variants; max delta 0.075 satisfies this.
POLE local custom framework BP4 rule (REVEL class 'Likely benign' with >=4 benign in silico results for exact missense variants in Supplementary Tables S2/S3) does not apply: c.6658-19C>G is intronic and absent from the missense-only tables, so the framework's generic fallback governs (see vcep_materials.json final_classification_framework and index_guidance).
REVEL score null/not found and BayesDel not found in prefetch.json; BayesDel is treated as not_available for PP3/BP4 under this pipeline's policy regardless of score because no verified published calibration threshold/PMID is available.
paper_extracts PMID:25741868: generic ACMG/AMP guidance that in silico splice predictors have ~90-100% sensitivity but ~60-80% specificity, that multiple in silico programs count as a single piece of evidence, and that confirmation of splice impact requires functional RNA/protein analysis; used as context for treating the single SpliceAI prediction as one line of computational evidence.
PP3Not met: SpliceAI max delta 0.075 is below the >0.2 supporting threshold, indicating no predicted splice impact.
Details ▾Hide ▴
Reasoning
NM_006231.4:c.6658-19C>G is a deep intronic substitution (19 bp upstream of the exon boundary, variant_exonic_positions exon 47i, protein consequence NP_006222.2:p.?) - not a missense substitution and not a canonical +/-1,2 splice-site change. The local custom POLE framework (Leon-Castillo et al. 2020 package) PP3 rule applies only when the exact missense variant appears in Supplementary Tables S2/S3 with REVEL class 'likely disease causing' and <=1 benign in silico result; this intronic variant is absent from those missense-only tables, so per the framework's own fallback instruction the generic ACMG/AMP in-silico calibration governs. Under that calibration only the SpliceAI splice-path branch applies to intronic variants: SpliceAI max delta score is 0.075 (DS_AL), well below the PP3 supporting threshold of >0.2, so PP3 is not met.
Rule
generic_acmg_pp3_bp4_splice_path
Evidence considered
SpliceAI Lookup result for NM_006231.4:c.6658-19C>G: DS_AG=0.0, DS_AL=0.075, DS_DG=0.009, DS_DL=0.001, max delta score = 0.075; tool evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.07).' (source: spliceai).
Generic ACMG/AMP in-silico calibration (output/generic_acmg_classification_rules.md, registered as generic_acmg_combination_rules, source basis PMID:25741868) specifies for intronic/synonymous/non-canonical-splice-position variants: evaluate SpliceAI only - max delta >0.2 -> PP3 supporting, <0.1 -> BP4 supporting, [0.1,0.2] gray zone; the REVEL missense-path rule does not apply to this intronic variant and BayesDel is never used because no verified published threshold/PMID is available.
POLE local custom framework (vcep_materials/final_classification_framework, Leon-Castillo et al. 2020-derived) customizes PP3/BP4 only for exact missense variants present in Supplementary Tables S2 (PATH-250-323-s003.xlsx) and S3 (PATH-250-323-s004.xlsx); those tables list in silico results for POLE missense mutations, and c.6658-19C>G (intronic, p.?) is not among them, so the custom PP3/BP4 rule cannot be applied and the framework's fallback to the generic in-silico workflow applies.
REVEL score not found (null) and BayesDel score not found in prefetch/evidence for this variant; both are inapplicable here because the variant has no missense consequence, and BayesDel additionally has no verified published calibration threshold available to this pipeline.
BP7N/ANot applicable: the variant is deep intronic rather than a synonymous coding change; the no-splice-impact evidence is captured under BP4.
Details ▾Hide ▴
Reasoning
BP7 as defined by generic ACMG/AMP (Richards et al. 2015, PMID:25741868) applies to synonymous (silent) coding variants for which splicing prediction algorithms predict no impact to the splice consensus sequence nor creation of a new splice site. NM_006231.4:c.6658-19C>G is a deep intronic substitution (19 bp upstream of the exon, protein consequence NP_006222.2:p.?), not a synonymous coding change. The local custom POLE framework customizes only PM1/PS4/PP3/BP4 and contains no gene-specific BP7 clause extending it to intronic variants. BP7 is therefore not applicable; the no-predicted-splice-impact evidence for this intronic variant is already captured under BP4 (met, supporting).
Evidence considered
Variant normalization (prefetch.json/variant_validator) shows NM_006231.4:c.6658-19C>G is an intronic substitution at exon 47 intronic position with predicted protein consequence NP_006222.2:p.? (not a synonymous/silent coding substitution), so BP7's synonymous-variant scope is not met.
Generic ACMG/AMP BP7 definition (Richards et al. 2015, PMID:25741868) is limited to synonymous variants; the local custom POLE framework (vcep_materials.json) lists customized criteria PM1, PS4, PP3, BP4 only and provides no BP7 rule for intronic variants, so generic BP7 scope applies unchanged.
SpliceAI predicts no significant splice impact for this variant (max delta score = 0.07).
2 paper records listed. 1 cited in the classification. None of the listed paper records explicitly indicates that the variant is named.
Cited in the classification
Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology.
Richards S et al. · 2015 May
This paper (Richards et al. 2015) is the ACMG/AMP standards and guidelines for sequence variant interpretation that defines the PVS1, PM4 and BP3 criteria used by this group's generic fallback: PVS1 as a null variant (nonsense, frameshift, canonical +/-1/2 splice-site, initiation codon, single/multi-exon deletion) in a gene where loss of function is a known mechanism; PM4 as protein length changes from in-frame deletions/insertions in a non-repeat region or stop-loss variants; and BP3 as in-frame deletions/insertions in a repeat region without known function. It also gives generic guidance that splicing prediction tools have roughly 90-100% sensitivity but only 60-80% specificity and that confirmation of actual splice impact requires functional RNA or protein analysis. It does not describe the POLE gene or any POLE-specific findings, and the exact variant c.6658-19C>G is not mentioned anywhere in the text.
For example, rule PM3 could be upgraded to strong if there were multiple observations of detection of the variant in trans (on opposite chromosomes) with other pathogenic variants (see section PM3 for further guidance).
Location Full-text paragraph on criteria combination and Table 5 scoring rules (flexible-weight discussion), character offset ~24,906 of the extracted text file publications/25741868.txt; the fetched extract covers the introduction through the odds-ratio guidance and does not include the Table 3/Table 4 criterion-definition rows themselves
Context ACMG/AMP consensus standards document (joint consensus recommendation of ACMG and AMP); not a variant-level experimental study.
Other recorded publications
Sherloc: a comprehensive refinement of the ACMG-AMP variant classification criteria.
The evidence statements recorded by the interpretation pipeline. They have not been reinterpreted by the page.
gnomAD v4.1
This variant is present in gnomAD v4.1 (AF= 7.4264e-05; MAF= 0.00743%, 119/1602392 alleles, homozygotes = 0) and has highest observed frequency in the East Asian population (AF= 0.00142819; MAF= 0.14282%, 64/44812 alleles, homozygotes = 0); grpmax FAF= 0.0011474.
gnomAD v2.1
This variant is present in gnomAD v2.1 (AF= 0.000233588; MAF= 0.02336%, 65/278268 alleles, homozygotes = 0) and has highest observed frequency in the East Asian population (AF= 0.00297229; MAF= 0.29723%, 59/19850 alleles, homozygotes = 0); grpmax FAF= 0.00227514.
gnomAD-Canada
This variant is present in gnomAD-Canada v1.0 (AF= 0.0007057546145494029, 13/18420 alleles, homozygotes = 0).
ClinVar
This variant has been reported in ClinVar as Likely benign (5 clinical laboratories) and as Benign (1 clinical laboratory). (ClinVarID = 246255)
In silico
SpliceAI predicts no significant splice impact for this variant (max delta score = 0.07).
Somatic
This variant has not previously been reported in somatic cancers (COSMIC).
This variant was interpreted by pipeline 7.2.0.
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.