POLE c.2964G>T, p.(Ser988=)

Reasoning

PM2 is met: the variant is absent from gnomAD v2.1 (exome), gnomAD v4.1 (exome), and gnomAD-Canada v1.0 (HostSeq genomes) - i.e., absent in large control population databases, satisfying the generic ACMG/AMP 2015 PM2 definition 'Absent in controls (or at extremely low frequency if recessive)' (Richards et al., PMID 25741868). Strength is Supporting per the ClinGen SVI Bayesian recalibration (Tavtigian et al., PMID 32720330), consistent with current practice (original 2015 paper listed Moderate; SVI downgraded to Supporting). Caveats: (1) the variant is synonymous (p.Ser988=) with SpliceAI max delta 0.01 (no predicted splice impact), so PM2's weight for this variant should be interpreted in the context of the functional/splicing group; (2) no per-ancestry allele counts are available in the bundle, so ancestry-specific frequency could not be confirmed, and a sequencing/annotation artifact cannot be fully excluded without raw counts; screenshots of the gnomAD queries are archived for audit.

Rule

Generic ACMG/AMP 2015 PM2 (Richards et al., PMID 25741868), strength Supporting per ClinGen SVI (Tavtigian et al., PMID 32720330); not customized by the POLE local framework.

Evidence considered

• gnomAD v2.1 (exome): variant absent (search_status=absent).
• gnomAD v4.1 (exome): variant absent (search_status=absent).
• gnomAD-Canada v1.0 (HostSeq genomes): variant absent (search_status=absent).
• PM2 generic definition: 'Absent in controls (or at extremely low frequency if recessive)' (Richards et al., PMID 25741868); absence in all three queried population databases satisfies the criterion.
• Strength Supporting per ClinGen SVI Bayesian framework recalibration of PM2 (Tavtigian et al., PMID 32720330).

Still missing

• No numeric allele counts, coverage/quality metrics, or per-ancestry breakdown in the bundle (gnomAD source_data empty); absence is documented by evidence sentences and screenshots only.
• Variant is synonymous; PM2 application assumes the variant is evaluable for pathogenicity (splice impact minimal per SpliceAI delta 0.01), which is outside this group's remit.

Sources

Reasoning

BP7 is met at Supporting strength under the generic ACMG/AMP 2015 rule (the custom POLE framework does not customize BP7). The variant is synonymous: Mutalyzer and VariantValidator both predict no amino-acid change (NP_006222.2:p.(Ser988=) / p.(=)), it lies in exon 25 (c.2964), and it is not at a canonical splice consensus position (pvs1_variant_assessment: canonical_splice_consensus=false). SpliceAI predicts no splice impact: max delta score 0.01 (DS_AG 0.01, DS_AL 0.0, DS_DG 0.0, DS_DL 0.0), far below the 0.2 high-impact delta threshold (Jaganathan et al. 2019, PMID 30661751), indicating no effect on acceptor/donor gain or loss and no creation of a cryptic splice site. This satisfies BP7: a synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site.

Rule

Generic ACMG/AMP 2015 (Richards et al., PMID 25741868) BP7: a synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site. Strength: Supporting.

Evidence considered

• spliceai: SpliceAI max delta score 0.01 (DS_AG 0.01, DS_AL 0.0, DS_DG 0.0, DS_DL 0.0) for NM_006231.4:c.2964G>T (SpliceAI Lookup, Broad Institute, distance=500, basic mode); below the 0.2 high-impact delta threshold (Jaganathan et al. 2019, PMID 30661751), predicting no splice impact and no new splice-site creation.
• pvs1_variant_assessment: consequence_class='synonymous', protein NP_006222.2:p.(Ser988=), canonical_splice_consensus=false.
• generic_acmg_combination_rules: BP7 criterion from Richards et al. 2015 (PMID 25741868) - synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site.

Sources

Reasoning

The governing framework is the local custom POLE framework derived from Leon-Castillo et al. 2020 (no official ClinGen CSPEC exists for POLE; cspec found=false). Its PS4 rule applies PS4_Supporting only to an exact missense variant recurrent in both the COSMIC and TCGA endometrial carcinoma cohorts in Supplementary Table S1 with combined EC count >=10 and belonging to the established pathogenic set (p.P286R, p.V411L, p.A456P, p.S297F). The queried variant c.2964G>T is synonymous (p.Ser988=), is absent from Supplementary Table S1 (verified by direct cell scan of PATH-250-323-s002.xlsx: 101 recurrent variants listed, none at residue 988), and has no COSMIC record, so the gene-specific PS4 rule is determinatively not triggered. Generic ACMG/AMP PS4 (case-control enrichment) is also unsupported: no case-control or affected-vs-control prevalence study exists for this variant, and the only literature source (PMID:25394175) is an ACMG/NSGC referral-indications practice guideline that does not mention the variant anywhere in its full text.

Rule

Custom POLE framework PS4 (Leon-Castillo et al. 2020, J Pathol 250:323-335): 'Apply PS4_Supporting only when the exact POLE missense variant is recurrent in both the COSMIC and TCGA endometrial carcinoma cohorts in Supplementary Table S1, has a combined EC count >=10, and belongs to the established/pathogenic set highlighted by Leon-Castillo et al. 2020 (p.P286R, p.V411L, p.A456P, p.S297F). Do not upgrade above Supporting.' Fallback: ACMG/AMP 2015 PS4 (PMID 25741868) - prevalence in affected individuals statistically increased over controls.

Evidence considered

• Custom POLE framework rule payload (final_classification_framework, Leon-Castillo et al. 2020 custom framework v1): PS4_Supporting requires exact missense variant recurrent in both COSMIC and TCGA EC cohorts (Supplementary Table S1), combined EC count >=10, in established set p.P286R/p.V411L/p.A456P/p.S297F; capped at Supporting.
• PATH-250-323-s002.xlsx Supplementary Table S1 directly parsed: 102 rows incl. header; 101 recurrent POLE variants with COSMIC/TCGA counts; no entry contains c.2964G>T, 2964, S988, or Ser988.
• COSMIC search (evidence.json): 'This variant has not previously been reported in somatic cancers (COSMIC)' - no result for POLE c.2964G>T.
• ClinVar VCV001798204: single submission (Ambry Genetics, clinical testing, germline), 'Likely benign', 1-star; no case-control or prevalence data attached.
• PMID:25394175 full text (publications/25394175.txt, 117 KB) searched for c.2964G>T / p.Ser988= / S988 / g.133237651C>A: zero hits; guideline document provides no enrichment evidence.

Still missing

• No germline case-control or affected-vs-control prevalence data exists for c.2964G>T; only somatic-recurrence resources (COSMIC; TCGA via framework Supplementary Table S1) were available and the variant is absent from both.

Sources

Reasoning

Variant NM_006231.4:c.2964G>T is synonymous (p.Ser988=) and was evaluated under the generic ACMG/AMP 2015 in-silico workflow because the governing custom POLE framework (León-Castillo et al. 2020-derived, 'leon-castillo-2020-custom-framework-v1') restricts its PP3 rule to exact missense variants listed in Supplementary Tables S2/S3 with REVEL class 'likely disease causing' and <=1 benign in silico result; this variant is synonymous and is absent from both tables (verified by direct inspection of PATH-250-323-s003.xlsx and PATH-250-323-s004.xlsx), so the custom rule is not triggered. Under the generic workflow, PP3 requires multiple lines of computational evidence supporting a deleterious effect. The only computational predictor available for this variant is SpliceAI, which predicts no splice impact (max delta score 0.01), i.e., evidence in the opposite direction. REVEL returned no score (missense-only predictor; no entry for this synonymous variant), BayesDel returned no score and additionally has no verified published PP3/BP4 threshold available to this pipeline (treated as insufficiently calibrated), and aGVGD is not available in this case bundle. No line of computational evidence supports a deleterious effect, so PP3 is not met.

Rule

Generic ACMG/AMP 2015 (Richards et al., PMID 25741868) PP3: multiple lines of computational evidence support a deleterious effect on the gene or gene product. Custom POLE framework PP3 rule inapplicable (variant is synonymous and absent from Supplementary Tables S2/S3).

Evidence considered

• spliceai: SpliceAI max delta score 0.01 (DS_AG 0.01, DS_AL 0.0, DS_DG 0.0, DS_DL 0.0) for NM_006231.4:c.2964G>T (12-133237651-C-A); far below the 0.2 high-impact delta threshold (Jaganathan et al. 2019, PMID 30661751), predicting no splice impact - contradicts a deleterious computational prediction.
• revel: REVEL v1.3 lookup for 12-132661065-C-A returned no score (found=false); missense-only predictor, variant is synonymous - not available for PP3.
• bayesdel: BayesDel noAF chr12 lookup returned no score (found=false); additionally no verified published threshold is available to this pipeline, so BayesDel is treated as insufficiently calibrated for PP3/BP4 - not used.
• vcep_path_250_323_s003 / vcep_path_250_323_s004: c.2964G>T / p.Ser988= is absent from Supplementary Tables S2 and S3 (missense-only variant lists); the custom POLE PP3 rule (REVEL 'likely disease causing' + <=1 benign result) is not triggered.
• final_classification_framework: custom POLE framework customizes PP3/BP4 only for exact missense variants in the supplementary tables; inapplicable to this synonymous variant, so the generic in-silico workflow (Richards et al. 2015, PMID 25741868) applies.
• pvs1_variant_assessment: canonical_splice_consensus=false and consequence_class='synonymous' for this variant, consistent with a non-splice-site change (no PP3-supporting splice context).

Still missing

• No REVEL score available for c.2964G>T (REVEL is a missense-only predictor and the variant is synonymous; lookup in revel-v1.3 chr12 data returned no entry).
• No BayesDel score available; BayesDel also has no verified published PP3/BP4 calibration threshold named for this pipeline, so it is treated as not_available regardless of score.
• aGVGD results are not present in this case bundle.

Sources

Reasoning

PP5 applies only for an exact-variant ClinVar expert-panel classification of Pathogenic or Likely pathogenic. ClinVar VCV001798204 has exactly one submission (SCV002749143, Ambry Genetics, clinical testing, germline), review status 'criteria provided, single submitter' (1 star), classified 'Likely benign'; expert_panel_submissions = 0 and no validated PMIDs exist. There is no expert-panel (3-star) pathogenic assertion, and the only classification present is a non-expert clinical laboratory submission in the benign direction in any case. Per adjudication policy, ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions must never trigger PP5 (or BP6). PP5 is therefore not met.

Rule

PP5 adjudication policy: apply PP5 only for an exact-variant ClinVar expert-panel (3-star) classification of Pathogenic/Likely pathogenic; ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions never trigger PP5.

Evidence considered

• ClinVar submission audit (prefetch.json): submissions_total=1, usable_for_classification=1, usable_for_criterion_leads=0, expert_panel_submissions=0, validated_pmids=0, criterion_lead_submissions=0.
• ClinVar record VCV001798204: classification 'Likely benign', review_status 'criteria provided, single submitter', review_stars=1, submitter Ambry Genetics (org 61756), date_last_evaluated 2020-05-29.
• PMID:25394175 (tagged candidate PP5) full text reviewed: guideline paper; does not classify this variant.

Still missing

• No ClinVar expert-panel (or 3-star) classification exists for this variant; only a single 1-star clinical laboratory submission (Ambry Genetics, SCV002749143).

Sources

Reasoning

Variant NM_006231.4:c.2964G>T (POLE, p.Ser988=) is absent from gnomAD v2.1 (exome), gnomAD v4.1 (exome), and gnomAD-Canada v1.0 (genomes) (all report search_status=absent; no allele counts present in source_data). BA1 requires a population allele frequency too high for the disorder (generic ACMG/AMP 2015, Richards et al., PMID 25741868). A frequency of zero cannot meet BA1 under any published threshold. No ClinGen CSPEC exists for POLE; the local custom framework (Leon-Castillo et al. 2020) customizes only PM1/PS4/PP3/BP4, so the generic ACMG/AMP rule applies.

Rule

Generic ACMG/AMP 2015 BA1 (Richards et al., PMID 25741868); not customized by the POLE local framework (which customizes PM1/PS4/PP3/BP4 only).

Evidence considered

• gnomAD v2.1 (exome): variant absent (search_status=absent).
• gnomAD v4.1 (exome): variant absent (search_status=absent).
• gnomAD-Canada v1.0 (HostSeq genomes): variant absent (search_status=absent).
• Generic ACMG/AMP 2015 BA1: allele frequency too high for disorder (Richards et al., PMID 25741868) - frequency of 0% cannot meet the criterion.

Still missing

• Quantitative allele counts and per-ancestry frequencies are not present in the bundle (gnomAD source_data empty), so absolute absence rests on the evidence sentences/screenshots rather than numeric counts.

Sources

Reasoning

BS1 requires the allele frequency to be greater than expected for the disorder (generic ACMG/AMP 2015, PMID 25741868). The variant is absent (frequency 0) from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0, which is the opposite of the condition required; it cannot be more common than expected if it is not observed at all. No disease-prevalence- or penetrance-calibrated expected-frequency estimate was needed because the observed frequency is zero.

Rule

Generic ACMG/AMP 2015 BS1 (Richards et al., PMID 25741868); not customized by the POLE local framework.

Evidence considered

• gnomAD v2.1 (exome): variant absent (search_status=absent).
• gnomAD v4.1 (exome): variant absent (search_status=absent).
• gnomAD-Canada v1.0 (HostSeq genomes): variant absent (search_status=absent).
• Generic ACMG/AMP 2015 BS1: allele frequency greater than expected for disorder (Richards et al., PMID 25741868) - frequency of 0% is not greater than expected.

Still missing

• No disease prevalence/penetrance-calibrated BS1 expected-frequency threshold (e.g., Whiffin-type calculation) is available in the bundle; immaterial here because the observed frequency is 0, but would be needed if the variant were observed at low frequency.

Sources

Reasoning

BS2 requires the variant to be observed in a healthy adult individual (homozygous for a recessive disorder, heterozygous for a dominant disorder) with full penetrance expected at an early age (generic ACMG/AMP 2015, PMID 25741868). No such observations exist in the evidence bundle: the variant is absent from all three population databases (gnomAD v2.1, v4.1, gnomAD-Canada), meaning no population-based healthy carriers or homozygotes were identified, and no clinical healthy-adult observation is reported. ClinVar (1-star, single submitter, Ambry Genetics 'Likely benign', criteria not provided) contributes no usable BS2 evidence and was not used as a criterion source.

Rule

Generic ACMG/AMP 2015 BS2 (Richards et al., PMID 25741868); not customized by the POLE local framework.

Evidence considered

• gnomAD v2.1 (exome): variant absent - no healthy carriers/homozygotes observed.
• gnomAD v4.1 (exome): variant absent - no healthy carriers/homozygotes observed.
• gnomAD-Canada v1.0 (HostSeq genomes): variant absent - no healthy carriers/homozygotes observed.
• Generic ACMG/AMP 2015 BS2: observed in healthy adult with full penetrance expected at early age (Richards et al., PMID 25741868) - no such observation documented.
• ClinVar entry (VCV001798204, Ambry Genetics, 'Likely benign', criteria provided single submitter, 1 star) reviewed but not used as criterion evidence (no criteria provided; final classifications of other sources must not be used to infer criteria).

Still missing

• No homozygote or healthy-adult carrier observation data for this variant anywhere in the bundle; population absence itself argues against BS2 (BS2 requires presence in healthy individuals).

Sources

Reasoning

BP4 (multiple lines of computational evidence suggest no impact on the gene or gene product) was evaluated under the generic ACMG/AMP 2015 workflow; the custom POLE BP4 rule requires an exact missense variant in Supplementary Tables S2/S3 with REVEL class 'Likely benign'/'Likely-benign' and >=4 benign in silico results, which does not apply to this synonymous variant (absent from both tables, verified directly). The only computational evidence of 'no impact' available for this variant is the SpliceAI prediction (max delta score 0.01). Per the group rule that a single prediction must not be double-counted across PP3/BP4/BP7, that splice prediction is allocated to BP7, the specific ACMG/AMP criterion for synonymous variants with no predicted splice impact, and is not additionally counted toward BP4. No independent benign computational lines are available (REVEL no score, BayesDel no score and uncalibrated, aGVGD absent). With no independent evidence line, BP4 is not met.

Rule

Generic ACMG/AMP 2015 (Richards et al., PMID 25741868) BP4: multiple lines of computational evidence suggest no impact on gene or gene product. Custom POLE framework BP4 rule inapplicable (variant is synonymous and absent from Supplementary Tables S2/S3).

Evidence considered

• spliceai: SpliceAI max delta score 0.01 for NM_006231.4:c.2964G>T, below the 0.2 high-impact threshold (Jaganathan et al. 2019, PMID 30661751) - the sole no-impact prediction; allocated to BP7, not double-counted for BP4.
• revel: REVEL v1.3 lookup returned no score for this synonymous variant - no independent benign-predictor line available.
• bayesdel: BayesDel noAF lookup returned no score and has no verified published PP3/BP4 threshold - treated as insufficiently calibrated, not used.
• vcep_path_250_323_s003 / vcep_path_250_323_s004: variant absent from Supplementary Tables S2 and S3; custom POLE BP4 rule (REVEL 'Likely benign' + >=4 benign results) not triggered.
• final_classification_framework: custom POLE framework BP4 customization inapplicable to this variant; generic in-silico workflow applies.

Still missing

• No independent (non-splice) computational line is available to support BP4: REVEL no score (missense-only predictor), BayesDel no score / no verified published threshold, aGVGD absent from the case bundle.
• The single available no-impact prediction (SpliceAI) is consumed by BP7 for this synonymous variant to avoid double-counting.

Sources

Reasoning

BP6 applies only for an exact-variant ClinVar expert-panel classification of Benign or Likely benign. Although ClinVar VCV001798204 carries a 'Likely benign' label, that label comes from a single ordinary clinical laboratory submission (Ambry Genetics, SCV002749143, review status 'criteria provided, single submitter', 1 star) with expert_panel_submissions = 0. Per adjudication policy, ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions must never trigger BP6. The 'Likely benign' label is recorded as context only and is not usable as BP6 evidence. BP6 is therefore not met.

Rule

BP6 adjudication policy: apply BP6 only for an exact-variant ClinVar expert-panel (3-star) classification of Benign/Likely benign; ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions never trigger BP6.

Evidence considered

• ClinVar submission audit (prefetch.json): submissions_total=1, expert_panel_submissions=0, usable_for_criterion_leads=0; only submission is Ambry Genetics (clinical laboratory, clinical testing, germline).
• ClinVar record VCV001798204: classification 'Likely benign', review_stars=1, review_status 'criteria provided, single submitter', date_last_updated 2026-04-13.
• Adjudication policy: BP6 requires an exact-variant ClinVar expert-panel (3-star) Benign/Likely benign classification; 1-star laboratory submissions are ineligible.

Still missing

• No ClinVar expert-panel benign classification exists for this variant; the sole 'Likely benign' label is a 1-star single-laboratory submission that policy excludes from BP6.

Sources

Reasoning

PS2 (de novo pathogenic, strong) requires a documented de novo occurrence of the variant in a proband with both parents tested and maternity/paternity confirmed. No de novo observation for NM_006231.4:c.2964G>T (p.Ser988=) is documented in any source in this case record: the ClinVar submission audit (VCV001798204; Ambry Genetics SCV002749143, germline clinical testing, Likely benign, 'criteria provided, single submitter') extracted zero evidence signals (criteria_found empty, usable_for_criterion_leads false, no criterion leads) and no de novo assertion; the literature packet contains no case report or proband study mentioning this variant (total 1 PMID, an ACMG/NSGC referral-indications guideline); and a grep of the only full-text paper (PMID:25394175) for c.2964G>T, p.Ser988=, segregation, de novo, and proband terms returned zero matches. No proband/parental testing data exist to even begin PS2 evaluation. Criterion falls back to generic ACMG/AMP 2015 (Richards et al., PMID 25741868) because the governing local custom POLE framework (Leon-Castillo et al. 2020, PATH-250-323 package) customizes only PM1/PS4/PP3/BP4 and contains no segregation/de novo specification; no ClinGen CSPEC for POLE was found.

Rule

Generic ACMG/AMP 2015 PS2: de novo variant (confirmed maternity and paternity) in a patient with the disease and no family history — Richards et al. 2015, PMID 25741868 (no gene-specific PS2 rule in the governing POLE framework)

Evidence considered

• ClinVar VCV001798204 submission audit: single germline clinical-testing submission (Ambry SCV002749143), Likely benign, single-submitter review status; evidence_signals/criteria_found empty and criterion_leads empty — no PS2/de novo assertion
• Full text of PMID:25394175 screened (grep for c.2964G>T, p.Ser988=, S988, segregation, de novo, proband): zero matches; paper is an ACMG/NSGC referral-indications practice guideline, not a case report
• Governing framework check: final_classification_framework.json (leon-castillo-2020-custom-framework-v1) customizes PM1/PS4/PP3/BP4 only; PS2 falls back to generic ACMG/AMP 2015 (PMID 25741868)

Still missing

• No proband case report with parental testing results is available; de novo status of c.2964G>T has never been assessed or reported
• Maternity/paternity confirmation data (or lack thereof) not available for any carrier
• No literature case series or segregation study mentioning this exact variant was retrieved

Sources

Reasoning

Governing framework: no ClinGen CSPEC/VCEP for POLE exists; the local custom framework (León-Castillo et al. 2020, vcep_path_250_323) customizes only PM1, PS4, PP3 and BP4 and does not specify PS3, so generic ACMG/AMP 2015 (PMID 25741868) applies: PS3 requires well-established functional studies demonstrating a deleterious effect. No functional (assay-based) study of NM_006231.4:c.2964G>T (p.Ser988=) exists in this case's evidence. The variant is synonymous; the only plausible functional effect is aberrant splicing, and the only related datum is an in silico SpliceAI prediction (max delta score 0.01), which is not a functional study and cannot satisfy PS3. The sole retrieved publication (PMID:25394175) is an ACMG/NSGC referral-indications guideline with no POLE content (full-text search returned 0 matches for POLE/variant terms), and the León-Castillo package contains no experimental assays for this variant (its 'functional status' content is limited to in silico prediction tools applied to non-synonymous somatic variants). Functional evidence is therefore insufficient, and PS3 cannot be assessed.

Evidence considered

• Generic ACMG/AMP 2015 PS3 definition (PMID 25741868): 'well-established functional studies show a deleterious effect' — no gene-specific PS3 specification exists in the local POLE framework
• Full-text review of PMID:25394175 — referral-indications guideline; no POLE or variant content; no functional data
• Review of León-Castillo et al. 2020 package (vcep_path_250_323) — functional-status content limited to in silico predictors for non-synonymous variants; no assay involving c.2964G>T
• SpliceAI Lookup: max delta score 0.01 (in silico splice prediction; not a functional assay)

Still missing

• No well-established functional study (e.g., splicing reporter/minigene assay, RNA-seq, or protein expression assay) for c.2964G>T (p.Ser988=)
• No published functional assay of any kind for this exact variant

Sources

Reasoning

PM3 requires, for a recessive disorder, that the variant under evaluation be detected in trans with a pathogenic variant, with phase confirmed (ACMG/AMP 2015, PMID 25741868). No such observation exists in this case. The case bundle contains no proband or family genotype data, no phase determination, and no second POLE variant: ClinVar variation 1798204 holds a single 'Likely benign' submission (Ambry Genetics, criteria provided/single submitter, 1 star) with no criterion leads and no validated PMIDs; the pm5_candidates artifact contains 0 candidates; no segregation or affected-proband observations are reported anywhere in the bundle. The governing framework is the local León-Castillo et al. 2020 custom POLE framework (customizes only PM1/PS4/PP3/BP4; official ClinGen CSPEC absent), so PM3 falls back to generic ACMG/AMP 2015, but even under the generic rule there is no trans/phase evidence to evaluate. A synonymous variant (p.Ser988=) with no documented biallelic context cannot be scored. The criterion is therefore not assessed, and the gap is the complete absence of genotype/phase data, not a judgment that the criterion is unmet.

Rule

PM3: For recessive disorders, detected in trans with a pathogenic variant (ACMG/AMP 2015; Richards et al., PMID 25741868, Table 5). Applicable via generic fallback because the local custom POLE framework does not define an allelic-criteria specification.

Evidence considered

• PM3 rule text: 'For recessive disorders, detected in trans with a pathogenic variant' (Richards et al. 2015, ACMG/AMP standards, PMID 25741868, Table 5) - governs PM3 because the custom POLE framework does not customize allelic criteria
• clinvar: variation 1798204, single submission SCV002749143 (Ambry Genetics), 'Likely benign', criteria provided/single submitter; submission_audit reports usable_for_criterion_leads=0, validated_pmids=0, no expert panel - provides no trans/cis or phase information and no second variant
• pm5_candidates: total_candidates_seen=0, same_residue_candidates=0 - no comparator variant evidence that could inform allelic configuration
• final_classification_framework: León-Castillo et al. 2020 custom POLE framework customizes only PM1, PS4, PP3, BP4; PM3/BP2 therefore fall back to generic ACMG/AMP 2015 combination rules (generic_acmg_combination_rules, PMID 25741868)
• PMID:25394175: full text reviewed; paper is an ACMG/NSGC cancer predisposition referral guideline and does not mention POLE c.2964G>T, p.Ser988=, or any phase/trans/cis observation

Still missing

• No proband/family genotype data: phase (trans/cis) of c.2964G>T relative to any other POLE allele was never determined
• No pathogenic variant in trans identified: no second POLE variant reported in ClinVar submission SCV002749143, and pm5_candidates contains 0 candidate comparator variants
• No affected-proband observations with biallelic genotype or segregation data supporting a recessive POLE disorder (e.g., FILS/IMAGEI) phenotype
• No germline inheritance-mode context (dominant PPAP vs recessive FILS/IMAGEI) is documented for the proband in this case

Sources

Reasoning

PM6 (assumed de novo, moderate) applies to a de novo occurrence without confirmation of maternity/paternity. No de novo occurrence of NM_006231.4:c.2964G>T is reported anywhere in this case record, with or without parentage confirmation: the ClinVar submission (VCV001798204, Ambry SCV002749143) carries no criteria signals and no de novo origin assertion, no proband report exists in the literature packet, and the only available full-text paper (PMID:25394175) contains no variant-specific data (grep-confirmed). Because no de novo event is documented at all, PM6 cannot be applied. Adjudicated under generic ACMG/AMP 2015 (PMID 25741868); the governing POLE custom framework provides no PM6 specification.

Rule

Generic ACMG/AMP 2015 PM6: assumed de novo, without confirmation of paternity and maternity — Richards et al. 2015, PMID 25741868 (no gene-specific PM6 rule in the governing POLE framework)

Evidence considered

• ClinVar VCV001798204 submission audit: no de novo or PM6 evidence signal in the single submission (evidence_signals/criteria_found empty)
• Full text of PMID:25394175 screened: no mention of c.2964G>T, p.Ser988=, or de novo occurrence
• Governing framework check: final_classification_framework.json customizes PM1/PS4/PP3/BP4 only; PM6 falls back to generic ACMG/AMP 2015 (PMID 25741868)

Still missing

• No de novo occurrence documented for this variant in any proband (confirmed or assumed)
• Parental testing results unavailable for any observed carrier
• No case-level literature (e.g., exome or cancer-predisposition cohort reports) mentioning this exact variant was retrieved

Sources

Reasoning

PP1 (co-segregation with disease in affected family members, supporting) requires at least one informative meiosis demonstrating that affected relatives carry the variant. No segregation study, pedigree, or family testing data for NM_006231.4:c.2964G>T exist in this case record: no affected relatives are reported as tested in the ClinVar submission (VCV001798204, Ambry SCV002749143 — zero criteria signals), no segregation paper was retrieved by the literature pass (criteria_index lists only PP5/PS4 candidates; no segregation-themed paper exists), and the single full-text paper (PMID:25394175) is a referral guideline with no variant content. With zero informative meioses documented, PP1 is not assessable. Adjudicated under generic ACMG/AMP 2015 (PMID 25741868); the governing POLE custom framework has no PP1 specification.

Rule

Generic ACMG/AMP 2015 PP1: co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease — Richards et al. 2015, PMID 25741868 (no gene-specific PP1 rule in the governing POLE framework)

Evidence considered

• ClinVar VCV001798204 submission audit: no segregation or family-testing assertion in the single submission
• literature_pass.json: total 1 PMID, none segregation-themed; criteria_index maps only PP5 and PS4 candidates; no PP1 candidate papers
• Full text of PMID:25394175 screened: no segregation data relevant to this variant
• Governing framework check: final_classification_framework.json customizes PM1/PS4/PP3/BP4 only; PP1 falls back to generic ACMG/AMP 2015 (PMID 25741868)

Still missing

• No family segregation testing data (affected relatives not tested or not reported) for this variant
• Zero informative meioses available to support or refute co-segregation
• No segregation-focused publication mentioning c.2964G>T was identified in the literature pass

Sources

Reasoning

PP4 requires the variant to be found in a patient (or family) with a phenotype or family history highly specific for the gene's disease. No proband phenotype, clinical history, or family-history data for any carrier of c.2964G>T exists anywhere in the evidence bundle: there are no case reports, the only literature source (PMID:25394175) is a referral guideline that does not mention the variant, and the ClinVar record carries no phenotype assertions beyond the generic condition label 'Hereditary cancer-predisposing syndrome'. The variant is also synonymous (p.Ser988=) with no predicted splice impact (SpliceAI max delta 0.01), so there is no phenotype-specificity signal to weigh. Evidence is insufficient; phenotype information would have to come from the clinical case file or a new reported case.

Rule

ACMG/AMP 2015 PP4 (PMID 25741868): phenotype or family history highly specific for a disease with a single genetic etiology.

Evidence considered

• Searched all bundle sources (case_summary, literature_pass, prefetch clinvar audit, evidence.json) for carrier phenotype or family-history data for c.2964G>T: none present.
• PMID:25394175 full text reviewed: no mention of the variant; no phenotype data tied to it.
• ClinVar VCV001798204: no phenotype descriptions in the single submission; condition label only 'Hereditary cancer-predisposing syndrome'.

Still missing

• No proband phenotype or family-history information is available for any carrier of c.2964G>T; without a reported case, phenotype specificity (PP4) cannot be evaluated.
• No case-level clinical data (personal cancer history, polyposis/endometrial phenotype, family history) is present in the case bundle.

Reasoning

BS3 also falls back to generic ACMG/AMP 2015 (PMID 25741868): 'well-established functional studies show no deleterious effect.' No functional studies of c.2964G>T (p.Ser988=) are available in this case. All related observations are non-functional: SpliceAI max delta 0.01 (in silico splice prediction), absence from gnomAD v2.1/v4.1/gnomAD-Canada (population frequency data), absence from COSMIC (somatic data), OncoKB 'Unknown Oncogenic Effect' (not variant-specific, no functional information), and a 1-star ClinVar 'Likely benign' submission (Ambry Genetics, criteria provided but no criteria/PMIDs disclosed). Per policy, ClinVar's final classification is not used to infer any criterion, and in silico/population/somatic observations cannot substitute for the functional studies BS3 requires. Such benign-leaning observations would support other criteria (e.g., BP7-type reasoning for a synonymous variant with no predicted splice impact), not BS3. Functional evidence is therefore insufficient, and BS3 cannot be assessed.

Evidence considered

• Generic ACMG/AMP 2015 BS3 definition (PMID 25741868): 'well-established functional studies show no deleterious effect' — no gene-specific BS3 specification in the local POLE framework
• SpliceAI Lookup: max delta score 0.01 — in silico splice prediction only; not a functional study
• ClinVar variation 1798204: 1-star Likely benign, single submitter (Ambry Genetics), no criteria provided — final classification not used to infer criteria
• OncoKB: 'Unknown Oncogenic Effect', not variant-specific — no functional information
• Full-text review of PMID:25394175 — no POLE or variant content; no functional data
• Review of León-Castillo et al. 2020 package (vcep_path_250_323) — no functional assays for this variant

Still missing

• No functional assay demonstrating absence of effect (e.g., minigene/splicing reporter assay or RNA-based confirmation) for c.2964G>T
• No RNA- or protein-level experimental data for this synonymous variant

Sources

Reasoning

BS4 (non-segregation with disease in affected family members, benign supporting) requires an affected family member who does NOT carry the variant, establishing absence of segregation. No family non-segregation data exist for NM_006231.4:c.2964G>T: no affected relatives have been tested and reported as variant-negative in the ClinVar submission (VCV001798204, Ambry SCV002749143 — zero criteria signals, no criterion leads), and no segregation study was retrieved. The ClinVar 'Likely benign' classification of this synonymous variant was deliberately NOT used to infer BS4 (no evidence signals or criteria accompany the submission, and a criterion cannot be inferred from another source's final classification). Population absence (gnomAD v2.1, v4.1, and gnomAD-Canada all report zero carriers) is population-frequency evidence relevant to PM2/BA1/BS1 and cannot substitute for family non-segregation, so it was not applied to BS4. Adjudicated under generic ACMG/AMP 2015 (PMID 25741868); the governing POLE custom framework has no BS4 specification.

Rule

Generic ACMG/AMP 2015 BS4: non-segregation with disease in affected family members — Richards et al. 2015, PMID 25741868 (no gene-specific BS4 rule in the governing POLE framework)

Evidence considered

• ClinVar VCV001798204 submission audit: no BS4/non-segregation assertion in the single submission; 'Likely benign' classification noted but not used as BS4 evidence because no criteria signals accompany it
• Full text of PMID:25394175 screened: no family non-segregation data relevant to this variant
• gnomAD v2.1, v4.1, and gnomAD-Canada reviewed for context only: variant absent (zero carriers) — population-frequency evidence, explicitly NOT applied to BS4
• Governing framework check: final_classification_framework.json customizes PM1/PS4/PP3/BP4 only; BS4 falls back to generic ACMG/AMP 2015 (PMID 25741868)

Still missing

• No affected family members tested and reported as non-carriers for this variant
• No family segregation/non-segregation study available for c.2964G>T
• Cannot distinguish BS4 (family non-segregation) from population-frequency or ClinVar-submitter arguments without pedigree data

Sources

Reasoning

BP2 requires that the variant be observed in trans with a pathogenic variant for a fully dominant disease (or threshold effect) or in cis with a pathogenic variant in any inheritance pattern (ACMG/AMP 2015, PMID 25741868). No such observation is available. The case bundle contains no genotype data establishing either allele configuration: there is no documented second POLE variant (ClinVar holds a single 'Likely benign' submission with no criterion leads and no validated PMIDs; pm5_candidates has 0 candidates), no family studies, no phase determination, and no proband phenotype/inheritance context (POLE germline disease spans autosomal dominant PPAP and autosomal recessive FILS/IMAGEI, but no proband phenotype is recorded here to anchor either mode). The governing framework is the local León-Castillo et al. 2020 custom POLE framework, which does not customize BP2, so the generic ACMG/AMP 2015 definition applies; however, the trans/cis observation required by the rule is entirely absent. BP2 is therefore not assessed, with the gap being the absence of any allelic-configuration observation rather than evidence that the variant is benign in cis/trans.

Rule

BP2: Observed in trans with a pathogenic variant for a fully dominant disease/threshold or in cis with a pathogenic variant in any inheritance pattern (ACMG/AMP 2015; Richards et al., PMID 25741868, Table 6). Applicable via generic fallback because the local custom POLE framework does not define an allelic-criteria specification.

Evidence considered

• BP2 rule text: 'Observed in trans with a pathogenic variant for a fully dominant disease/threshold or in cis with a pathogenic variant in any inheritance pattern' (Richards et al. 2015, ACMG/AMP standards, PMID 25741868, Table 6) - governs BP2 via generic fallback since the custom POLE framework does not customize allelic criteria
• clinvar: variation 1798204, single submission SCV002749143 (Ambry Genetics), 'Likely benign', criteria provided/single submitter; no criterion leads, no validated PMIDs, no expert panel - contains no trans/cis observation and no accompanying second variant
• pm5_candidates: 0 candidates total - no allelic-comparator data available
• final_classification_framework: León-Castillo et al. 2020 custom POLE framework customizes only PM1, PS4, PP3, BP4; BP2 falls back to generic ACMG/AMP 2015 (generic_acmg_combination_rules, PMID 25741868)
• PMID:25394175: full text reviewed; ACMG/NSGC referral guideline with no mention of the variant or any allele-configuration observation

Still missing

• No observation of the variant in trans with a pathogenic variant in a fully dominant POLE disorder (PPAP) - no phase data available
• No observation of the variant in cis with a pathogenic variant in any inheritance pattern - no phase data available
• No family segregation/genotyping data and no documented second POLE allele in the proband (ClinVar submission SCV002749143 reports no accompanying variant; pm5_candidates = 0)
• Proband inheritance mode and phenotype not documented, so a dominant-vs-recessive context for any hypothetical allelic observation cannot be established

Sources

Reasoning

BP5 requires the variant to be found in a case with an alternate molecular basis for the disease. No case-level reports of c.2964G>T exist in the bundle: no carrier is described, no alternate pathogenic variant/etiology is documented, and the only literature source does not mention the variant. There is therefore no information to establish or exclude an alternate molecular diagnosis. Evidence is insufficient; this criterion would require case-level data (e.g., a reported carrier with a separate causal variant).

Rule

ACMG/AMP 2015 BP5 (PMID 25741868): variant found in a case with an alternate molecular basis for disease.

Evidence considered

• Searched all bundle sources (case_summary, literature_pass, prefetch clinvar audit, evidence.json) for any reported carrier of c.2964G>T with an alternate molecular diagnosis: none found.
• PMID:25394175 full text reviewed: no variant mention, no case-level alternate-etiology data.

Still missing

• No case-level data exists for c.2964G>T; no reported carrier with a documented alternate molecular basis for disease (e.g., pathogenic variant in another gene explaining the phenotype).