APC c.8511T>C, p.(Ser2837=)

Reasoning

PM2_supporting is met under the APC-specific instruction to use the gnomAD v2.1.1 non-cancer dataset: this variant is absent from gnomAD v2.1. The v4.1 observation is extremely rare (5/1,612,964 alleles; total AF 3.09988e-06; grpmax FAF 7.9e-07; no homozygotes) and does not meet BA1 or BS1. Its raw total AF lies marginally above the APC PM2 AC>1 numerical cutoff of 3e-06, so v4.1 should not be substituted for the VCEP-designated v2.1 dataset to increase PM2 strength.

Rule

APC InSiGHT VCEP PM2_supporting: use total gnomAD v2.1.1 non-cancer data; AF <=0.000003 when AC >1, or AF <0.00001 when AC <=1.

Evidence considered

• The APC InSiGHT specification directs PM2 assessment using the total population from the gnomAD v2.1.1 non-cancer dataset and assigns PM2 at supporting strength only. Its rarity rule is AF <=0.000003 when allele count is >1, or AF <0.00001 when allele count is <=1.
• gnomAD v2.1 reports the variant absent.
• gnomAD v4.1 reports 5/1,612,964 alleles (total AF 3.09988e-06), joint grpmax FAF 7.9e-07, highest raw subpopulation AF 3.38249e-05 (1/29,564 Ashkenazi Jewish alleles), and zero homozygotes.

Still missing

• A contemporaneous APC VCEP update explicitly specifying how to apply the PM2 AC-dependent threshold to gnomAD v4 joint data would remove the cross-version boundary ambiguity.

Sources

Reasoning

The APC VCEP explicitly states BP4 is not applicable for missense variants (PP3/BP4 in APC use a splicing-predictor-only path, not conservation/missense predictors), but is applicable to synonymous/intronic variants where multiple in silico splicing predictors suggest no impact on gene or gene product. This variant is synonymous (p.(Ser2837=)). SpliceAI, the VCEP-recommended primary tool, shows max delta score = 0.00, indicating no predicted splicing impact, consistent with the BP4 rule threshold (scores well under the 0.2 moderate-impact cutoff). This supports BP4 at the VCEP-defined supporting strength. BayesDel/REVEL are not applicable as this is a synonymous, not missense, variant, and in any case REVEL lookup returned no score for this variant (found: false) and no verified BayesDel threshold is available to this pipeline.

Rule

APC CSPEC v2.1 BP4 (synonymous/intronic sub-rule), Supporting

Evidence considered

• APC CSPEC v2.1 BP4 rule: 'Missense variants: BP4 is not applicable. Synonymous (silent) or intronic variants: Multiple in silico splicing predictors suggest no impact on gene or gene product.'
• SpliceAI evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.00).'
• REVEL lookup for this variant returned found: false / revel_score: null, so no REVEL-based missense evidence exists (and is inapplicable given synonymous consequence).

Still missing

• Only a single splicing predictor (SpliceAI) is documented in the case bundle; the VCEP rule text says 'multiple in silico splicing predictors' but only one predictor's result was captured here. Confirmatory MaxEntScan/VarSeak scores were not available in the case evidence.

Sources

Reasoning

APC CSPEC v2.1 BP7 (Benign Supporting, applicable to APC) is met for 'a synonymous (silent) or intronic variant at or beyond +7/-21 for which multiple splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site.' NM_001127510.3:c.8511T>C is a synonymous coding variant (p.(Ser2837=)) located within exon 17, well beyond any +7/-21 intronic boundary (it is an exonic, not intronic, position). SpliceAI predicts no significant splice impact (max delta score = 0.00), consistent with no predicted disruption of the splice consensus sequence and no new cryptic splice site. The VCEP explicitly permits BP7 to be combined with BP4 for the same synonymous/splicing-negative evidence base, so both are applied here without double counting distinct predictions.

Rule

APC CSPEC v2.1 BP7, Supporting

Evidence considered

• APC CSPEC v2.1 BP7 strength rule: 'A synonymous (silent) or intronic variant at or beyond +7/-21 for which multiple splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site.'
• APC CSPEC v2.1 BP7 instructionsToUse: 'The use of BP7 with BP4 is allowed.'
• SpliceAI evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.00).'
• Variant is exonic (exon 17, protein NP_001120982.1:p.(Ser2837=)), consistent with the synonymous variant class targeted by this rule.

Still missing

• Case bundle documents only SpliceAI results; the VCEP recommends corroboration from MaxEntScan and/or VarSeak ('multiple' predictors) which were not present in the evidence bundle.

Sources

Reasoning

NM_001127510.3:c.8511T>C is a synonymous variant (NP_001120982.1:p.(Ser2837=)) in APC exon 17, not a missense variant, so the APC VCEP explicitly excludes computational conservation-based PP3 for this variant type. Per the APC CSPEC v2.1, PP3 for non-canonical variants requires that multiple in silico splicing predictors support a deleterious effect. SpliceAI (the VCEP-recommended primary splice predictor) predicts no splice impact (max delta score = 0.00, well below the 0.2 minimum threshold for even a moderate cryptic-splice-gain call per the VCEP's own instructions). No other splicing predictor result (MaxEntScan, VarSeak) is present in the case bundle to independently corroborate a deleterious effect. With the sole available predictor showing no impact, the 'multiple predictors support deleterious effect' threshold for PP3 is not met.

Rule

APC CSPEC v2.1 PP3 (non-canonical splicing variant sub-rule)

Evidence considered

• APC CSPEC v2.1 PP3 rule: 'In silico splicing predictors should be used for presumed missense variants... Non-canonical splicing variants: Multiple in silico splicing predictors support a deleterious effect.'
• SpliceAI evidence_sentence: 'SpliceAI predicts no significant splice impact for this variant (max delta score = 0.00).'
• VCEP instructionsToUse for splice predictors: SpliceAI score 0.8-1 = loss of native site; 0.2-0.8 = moderate cryptic gain. Observed max delta score of 0.00 is far below any impact threshold.

Still missing

• MaxEntScan and VarSeak scores (VCEP-recommended predictors) are not present in the case bundle; only SpliceAI was scraped.

Sources

Reasoning

APC-specific BA1 requires a gnomAD popmax filtering allele frequency (FAF) of at least 0.1% (0.001). gnomAD v4.1 reports a joint grpmax FAF of 7.9e-07 (0.000079%), far below this threshold; gnomAD v2.1 is absent.

Rule

APC InSiGHT VCEP BA1: gnomAD popmax filtering AF >=0.1% (0.001).

Evidence considered

• The APC InSiGHT specification directs use of the gnomAD non-cancer dataset and defines BA1 as popmax FAF >=0.001.
• gnomAD v4.1: 5/1,612,964 alleles (total AF 3.09988e-06), highest raw subpopulation AF 3.38249e-05 in Ashkenazi Jewish individuals, and joint grpmax FAF 7.9e-07; no homozygotes were observed.
• gnomAD v2.1 reports the variant absent.

Sources

Reasoning

APC-specific BS1 requires gnomAD popmax FAF of at least 0.001% (0.00001). The observed gnomAD v4.1 joint grpmax FAF is 7.9e-07 (0.000079%), below the threshold. The single Ashkenazi Jewish allele produces a raw AF of 3.38249e-05, but the VCEP rule explicitly uses filtering allele frequency, so the sparse-subpopulation raw AF does not satisfy BS1.

Rule

APC InSiGHT VCEP BS1: gnomAD popmax filtering AF >=0.001% (0.00001).

Evidence considered

• The APC InSiGHT specification defines BS1 as gnomAD popmax filtering AF >=0.001% (0.00001).
• gnomAD v4.1 joint grpmax FAF is 7.9e-07; the highest raw subpopulation observation is 1/29,564 alleles in Ashkenazi Jewish individuals (AF 3.38249e-05), with no homozygotes.
• gnomAD v2.1 reports the variant absent.

Sources

Reasoning

No case-level evidence documents NM_001127510.3:c.8511T>C as de novo in an affected proband with confirmed maternity and paternity. No parental genotypes, trio testing, relationship confirmation, phenotype-linked proband observation, or de novo score is available.

Rule

APC VCEP PS2: use the APC de novo score. At least 4 points supports Very Strong, 2-3.5 points Strong, and 1-1.5 points Moderate. A confirmed-parentage proband with phenotype score >=1 contributes 2 points; with phenotype score 0.5 contributes 1 point.

Evidence considered

• The APC VCEP framework requires PS2 de novo scoring and specifies that confirmed parental relationships with an APC phenotype score of at least 1 contribute 2 points per proband; no such case observation is present in the consolidated evidence bundle.
• The source registry and literature pass contain no variant-specific publication identified for de novo, parental, or family evidence.

Still missing

• Affected proband status and APC phenotype point category are not documented.
• Maternal and paternal genotypes for the variant are not documented.
• Parental relationships have not been confirmed.
• Validated trio or parental testing, including assessment of possible parental mosaicism, is not available.

Sources

Reasoning

The APC VCEP (InSiGHT Hereditary Colorectal Cancer/Polyposis EP, cspec v2.1) requires validated RNA assay data (premature stop codon, exon 13/14 skipping, or other splice aberration with quantified transcript levels) or, for beta-catenin-domain missense variants, protein assay data (increased beta-catenin transcriptional activity or decreased beta-catenin binding) to apply PS3. No RNA-based splicing/expression assay or protein functional assay of NM_001127510.3:c.8511T>C (p.Ser2837=) was found anywhere in the case bundle (evidence.json, literature_pass.json, source_registry.json). The only splicing-relevant datum present is an in-silico SpliceAI prediction (max delta score = 0.00, no significant splice impact), which is a computational prediction, not a validated functional assay, and cannot be used to satisfy PS3 per this VCEP's assay-based rule structure. The PMIDs retrieved by literature triage (25741868, 20301519, 21368914, 23788249, 27854360, 35802134, 28492532) are ACMG/AMP methodology and secondary-findings policy papers, none of which report a functional assay of this specific variant.

Still missing

• No RNA assay (e.g., RT-PCR/RNA-seq on patient blood/lymphoblastoid RNA) has been performed or reported for c.8511T>C to assess mRNA splicing/stability.
• No protein functional assay (beta-catenin transcriptional activity or SPR binding) is applicable/available since this is a synonymous variant, not a beta-catenin-domain missense change.
• SpliceAI in-silico prediction alone is insufficient to meet the VCEP's PS3 assay-based evidentiary requirement.

Reasoning

No exact-variant affected-proband observations, phenotype-point assignments, or case-control enrichment data are available. APC-specific PS4 requires phenotype points: Supporting for 1-1.5, Moderate for 2-3.5, Strong for 4-15.5, and Very Strong for at least 16.

Rule

APC VCEP PS4: phenotype-point thresholds are Supporting 1-1.5, Moderate 2-3.5, Strong 4-15.5, and Very Strong at least 16.

Evidence considered

• The APC VCEP specifies PS4 strength by cumulative phenotype points (1-1.5 Supporting; 2-3.5 Moderate; 4-15.5 Strong; at least 16 Very Strong).
• The available literature packet contains general APC/FAP resources and no variant-specific affected-proband or case-control evidence for c.8511T>C (p.Ser2837=).

Still missing

• Exact-variant observations in unrelated affected individuals with phenotype details sufficient to assign APC VCEP phenotype points.
• A case-control analysis or sufficiently documented affected-proband series for NM_001127510.3:c.8511T>C.

Sources

Reasoning

No evidence documents a presumed de novo occurrence of NM_001127510.3:c.8511T>C in an affected individual when parental relationships are unconfirmed. No parental genotype data, phenotype-linked proband, or de novo score is available.

Rule

APC VCEP PM6: use the APC de novo score. Scores of 2-3.5 support Strong, 1-1.5 Moderate, and 0.5 Supporting. With unconfirmed parental relationships, a proband with phenotype score >=1 contributes 1 point and one with phenotype score 0.5 contributes 0.5 points.

Evidence considered

• The APC VCEP framework applies PM6 using the APC de novo score when parental relationships are unconfirmed; no presumed de novo observation is present in the consolidated evidence bundle.
• The source registry and literature pass contain no variant-specific publication identified for de novo, parental, or family evidence.

Still missing

• Affected proband status and APC phenotype point category are not documented.
• Parental genotypes and the basis for presuming de novo status are not documented.
• Whether maternity and paternity were confirmed is not documented.
• Independent presumed de novo observations are not documented.

Sources

Reasoning

No family segregation evidence is available for NM_001127510.3:c.8511T>C. The bundle contains no pedigree, affected relative genotypes, phenotype concordance data, or informative meioses.

Rule

APC VCEP PP1: Supporting requires segregation in 3-4 meioses in >=1 family; Moderate requires 5-6 meioses in >=1 family; Strong requires >=7 meioses in >=2 families.

Evidence considered

• The APC VCEP framework requires co-segregation in 3-4 meioses for Supporting, 5-6 meioses for Moderate, or at least 7 meioses in at least 2 families for Strong; no informative meiosis is present in the consolidated evidence bundle.
• The source registry and literature pass contain no variant-specific publication identified for segregation or pedigree evidence.

Still missing

• Pedigree structure and number of informative meioses are not documented.
• Genotypes and APC phenotype details of affected relatives are not documented.
• Variant-positive affected relatives across one or more families are not documented.
• Age, penetrance, and phenotype information needed to judge segregation informativeness are not documented.

Sources

Reasoning

No qualifying healthy-individual observations or homozygotes are available. gnomAD v4.1 reports zero homozygotes, and the record lacks age, colonoscopy, phenotype, and testing-indication data needed to assign healthy-individual points. The APC specification does not permit use of gnomAD non-cancer data for heterozygous healthy individuals.

Rule

APC InSiGHT VCEP BS2: >=3 healthy-individual points (supporting), >=10 points or >=2 homozygous observations (strong).

Evidence considered

• APC InSiGHT permits BS2 at supporting strength for >=3 healthy-individual points and at strong strength for >=10 points or >=2 homozygous observations; gnomAD non-cancer data may be searched for homozygotes but not used as heterozygous healthy-individual evidence.
• gnomAD v4.1 reports 0 homozygotes among 1,612,964 alleles.

Still missing

• Qualifying healthy-individual evidence meeting the APC VCEP definition, totaling at least 3 points for BS2_supporting or 10 points for BS2_strong, or at least two independent homozygous observations.

Sources

Reasoning

Per the APC VCEP specification, BS3 for a synonymous variant such as c.8511T>C (p.Ser2837=) requires an RNA assay in a constitutional patient sample demonstrating no mRNA aberration (Strong tier requires additional confirmation of biallelic expression or NMD inhibition use; Supporting tier applies without that confirmation). No such RNA assay data for this variant was found in evidence.json, literature_pass.json, prefetch.json, or vcep_materials.json. The protein-assay pathway to BS3 is restricted by the VCEP to missense variants within the beta-catenin binding domain (codons 959-2129) and does not apply to this synonymous change. The only relevant computational output present, SpliceAI (max delta score = 0.00, predicting no splice impact), is an in-silico splicing prediction, not an RNA-based functional assay, and per the VCEP's assay-specific wording cannot substitute for the required wet-lab mRNA study to invoke BS3.

Still missing

• No RNA/mRNA assay (RT-PCR, minigene, or patient-derived transcript analysis) demonstrating absence of splicing aberration is available for c.8511T>C.
• Biallelic expression / NMD-inhibition status is unknown because no RNA study exists, precluding assessment even at the Supporting tier without further wet-lab data.
• SpliceAI computational prediction (delta=0.00) is suggestive but not an accepted substitute for the VCEP's required functional RNA assay for BS3.

Reasoning

No informative non-segregation evidence is available for NM_001127510.3:c.8511T>C. The bundle contains no affected relative who tested negative for the variant and no phenotype-point documentation needed for APC VCEP BS4.

Rule

APC VCEP BS4: Strong requires one affected variant-negative family member with >=1 phenotype point, or two affected variant-negative members each with >=0.5 phenotype points; Supporting requires one affected variant-negative member with >=0.5 phenotype points.

Evidence considered

• The APC VCEP framework requires either one affected variant-negative member scoring at least 1 phenotype point for Strong BS4, two such members each scoring at least 0.5 points for Strong BS4, or one affected variant-negative member scoring at least 0.5 points for Supporting BS4; no such observation is present in the consolidated evidence bundle.
• The source registry and literature pass contain no variant-specific publication identified for non-segregation or pedigree evidence.

Still missing

• Tested affected relatives who lack the variant are not documented.
• APC phenotype point assignments for variant-negative affected relatives are not documented.
• Pedigree, testing methodology, and genotype-quality information are not documented.
• No informative non-segregation event is available for review.

Sources

Reasoning

No case-level co-occurrence data identify a (Likely) Pathogenic APC variant, phase, or the number of independent observations for this variant; BP2 cannot be evaluated.

Rule

APC VCEP BP2_Supporting: observed in trans with a (Likely) Pathogenic APC variant, or at least three times in unknown phase with different (Likely) Pathogenic APC variants.

Evidence considered

• The APC VCEP BP2 rule permits supporting evidence for a variant observed in trans with a (Likely) Pathogenic APC variant, or at least three unknown-phase observations with different (Likely) Pathogenic APC variants.
• The consolidated case-evidence bundle contains no documented APC co-variant, phase determination, affected-proband co-occurrence, or qualifying count of unknown-phase co-occurrences for c.8511T>C.

Still missing

• Need documentation that c.8511T>C was observed in trans with a (Likely) Pathogenic APC variant, or at least three independent co-occurrences in unknown phase with different (Likely) Pathogenic APC variants.

Sources

Reasoning

No individual-level phenotype or molecular diagnosis establishes an alternate genetic basis for a colorectal polyposis phenotype.

Rule

APC VCEP BP5: only applicable for an alternate genetic basis of the colorectal polyposis phenotype.

Evidence considered

• APC VCEP limits BP5 to cases with an alternate genetic basis of the colorectal polyposis phenotype.
• No case-level phenotype or alternate-diagnosis evidence is present in the consolidated evidence bundle.

Still missing

• Patient phenotype and testing report documenting colorectal polyposis and an alternate molecular diagnosis that explains it.

Sources