MUTYH c.1186+17C>T, p.?

Reasoning

PS4 requires a statistically significant increase in the variant's frequency in affected individuals compared with controls, typically quantified as an odds ratio or equivalent enrichment statistic evaluated against a VCEP-defined threshold (met for an enrichment statistic requires value >= threshold). No case-control study, cohort, or case series reporting NM_001128425.2:c.1186+17C>T was identified. All seven triaged publications discuss MUTYH only at gene level and none names the variant: GeneReviews 'MUTYH Polyposis' (PMID:23035301), ACMG technical standards for inherited colorectal cancer testing (PMID:24310308), PDQ 'Genetics of Colorectal Cancer' (PMID:26389505), ACMG SF v3.0 (PMID:34012068) and v3.1 (PMID:35802134), the ACMG/AMP 2015 standards (PMID:25741868), and Sherloc (PMID:28492532); full-text extraction of the four retrieved papers (PMID:24310308, PMID:34012068, PMID:35802134, PMID:25741868) recorded variant_mentioned = false, and the ACMG SF policy documents list MUTYH only as a secondary-findings gene for MUTYH-associated polyposis with no variant-level data. The ClinVar record (variation 182700) contains eight laboratory submissions, none of which reports case enrichment and none of which is classified pathogenic or likely pathogenic. Because no enrichment value was reported anywhere, no numeric threshold comparison could be performed and the criterion cannot be satisfied. The governing InSiGHT Hereditary Colorectal Cancer/Polyposis MUTYH v1.0 specification (doc 1742141534) was located for this gene but its criterion-level PS4 rule text and any calibrated threshold were not available in this case, so the generic ACMG/AMP definition was applied.

Rule

Generic ACMG/AMP 2015 PS4: variant prevalence in affected individuals statistically significantly increased compared with controls; no gene-specific (InSiGHT MUTYH v1.0) PS4 threshold available in this case.

Evidence considered

• ClinVar variation 182700 (MUTYH c.1186+17C>T): 8 submissions total, aggregate classification 'Likely benign', review status 'criteria provided, single submitter'; submission audit counts expert_panel_submissions = 0. No submission reports case enrichment or a pathogenic/likely pathogenic classification.
• Literature triage (7 PMIDs) with variant-mention screening: every publication was screened for the exact variant and none mentions it (variant_mentioned = false for all 7 PMIDs). Themes were case_observation/functional for PMIDs 23035301, 24310308, 26389505, 34012068, 35802134, 25741868 and background_review for 28492532.
• PMID:24310308 (ACMG technical standards and guidelines for genetic testing for inherited colorectal cancer): MUTYH is referenced only as the gene underlying autosomal recessive MYH-associated polyposis; no variant-level MUTYH data, quote: 'in 2002, an autosomal recessive condition, MYH-associated polyposis (MAP), caused by mutations in the MUTYH gene, was described.'
• PMID:34012068 (ACMG SF v3.0) and PMID:35802134 (ACMG SF v3.1): policy/gene-list documents; MUTYH appears only as a listed secondary-findings gene for MAP (OMIM 608456, autosomal recessive, reporting only when 2 pathogenic/likely pathogenic variants are found). Quote from the ACMG SF v3.1 table: 'MAP\t1.0\t608456\tMUTYH\tAR\tP and LP (2 variants)'. No variant-level, case, or frequency data for c.1186+17C>T.
• PMID:25741868 (ACMG/AMP 2015): provides the generic definition of PS4 and the framework for combining criteria; it contains no MUTYH- or variant-specific data and does not name c.1186+17C>T.
• No odds ratio, relative risk, or case/control allele count for c.1186+17C>T exists in any retrieved source, so no numeric inequality against a PS4 threshold could be written or evaluated.
• gnomAD v2.1 shows the variant in population controls (total AF 0.000128372, 36/280434 alleles, 0 homozygotes), but presence in population databases is not case enrichment and cannot support PS4.

Still missing

• No case-control or affected-cohort data exist for c.1186+17C>T in the retrieved literature; an untargeted case-control dataset would be required to test PS4.
• The InSiGHT MUTYH v1.0 specification was located, but its criterion-level PS4 rule text (and any gene-specific enrichment threshold) was not retrieved, so generic ACMG/AMP wording was used.
• GeneReviews 'MUTYH Polyposis' (PMID:23035301) and PDQ 'Genetics of Colorectal Cancer' (PMID:26389505) had no full text available in this case; their abstracts/records are gene-level only and no variant-level case data was accessible.

Sources

Reasoning

PM1 requires the variant to be located in a mutational hotspot and/or a critical, well-established functional domain without benign variation. Two checks were performed. First, the authoritative domain source for this gene: vcep_materials.json contains no domain_tables entry for MUTYH (domain_tables = {}), so the VCEP supplied no approved critical-domain list against which a residue could be tested; the extracted MUTYH InSiGHT ruleset also carried no criterion-level payload (framework_complete false, rule_payload {}), and no hotspot resource returned data for this variant (hotspots.found = null, empty source_data). Second, the variant's own position: c.1186+17C>T lies 17 bp into the intron downstream of the exon ending at c.1186, i.e. it is non-coding and falls outside every coding region of MUTYH, so it cannot sit in a critical functional domain of the protein nor in a missense mutational hotspot, whatever those boundaries are. Because the variant cannot occupy a domain or hotspot, PM1 is not met; this conclusion depends only on the variant's non-coding location and is not affected by the missing VCEP domain table.

Evidence considered

• Governing framework: InSiGHT MUTYH VCEP specification v1.0 (key cspec) was consulted first; its domain table was checked and is absent from this case (vcep_materials domain_tables = {}).
• Variant position from the case evidence: NM_001128425.2:c.1186+17C>T is an intronic substitution 17 nucleotides from the exon boundary; no protein consequence is produced.
• Hotspot/somatic databases returned no data for this variant (hotspots.found = null; COSMIC found = false), so no hotspot membership could be claimed.
• Generic ACMG/AMP definition of PM1 (mutational hotspot and/or critical, well-established functional domain without benign variation) from the ACMG/AMP 2015 standards (key PMID:25741868), which contains no MUTYH-specific domain list.

Still missing

• vcep_materials.json domain_tables is empty for MUTYH, so no VCEP-approved critical functional domain list was available to test any residue against; this did not affect the verdict because the variant is intronic.
• No hotspot database result was returned for this variant (hotspots.found = null), and the case evidence contains no MUTYH domain-boundary annotation; irrelevant to an intronic +17 variant.

Sources

Reasoning

PM2 is not met under the generic supporting threshold of AF <= 0.0001 (ClinGen SVI 2020 recommendation to downgrade PM2 to supporting, attributed to PMID:25741868 in the calibration block). The variant is not absent from controls, and every available frequency exceeds the cutoff: gnomAD v2.1 non-cancer exomes 0.000119341 (28/234622 alleles, 1.19x the cutoff), gnomAD v2.1 all-comers 0.000128372 (36/280434, 1.28x), gnomAD v3.1 non-cancer genomes 0.000195999 (29/147960, 1.96x), gnomAD v4.1 all-comers 0.000231825 (374/1613286, 2.32x) and gnomAD-Canada v1.0 0.000543183 (10/18410, 5.4x). The grpmax FAF values (0.00017543 v2.1; 0.00026535 v4.1; 0.00020666 v3.1 non-cancer) and the European (non-Finnish) subgroup frequencies (0.000256997 in v2.1, 0.000291516 in v4.1, 0.000263866 in v2.1 non-cancer, 0.000324014 in v3.1 non-cancer) are likewise all above 0.0001, so no ancestry-specific or non-cancer source would rescue the criterion. Two caveats are recorded. First, the registered InSiGHT MUTYH specification has an empty criteria payload, so no VCEP-specific PM2 threshold could be read and the generic default was applied. Second, for an autosomal recessive condition such as MAP a VCEP-specific or recessive-calibrated PM2 cutoff can be less stringent than 0.0001, and the margin here is narrow - the lowest observed frequency exceeds the generic cutoff by only about 19% - so this criterion is explicitly flagged for human review. Homozygote data (0 in all datasets) do not alter the frequency-based verdict.

Rule

{'source': 'generic_clingen_svi_default', 'threshold': 0.0001, 'comparison': '<=', 'unit': 'allele frequency', 'population_selection': 'all-comers default (GNOMAD_V2_1/GNOMAD_V4_1) because the governing specification supplied no population-source instruction; non-cancer subsets and gnomAD-Canada also reported'}

Evidence considered

• gnomAD v2.1 (gnomad_v2) all-comers AF 0.000128372 (36/280434 alleles, 0 homozygotes); exome AF 0.000120463, genome AF 0.000191119; maximum ancestry European (non-Finnish) 0.000256997 (33/128406); grpmax FAF 0.00017543.
• gnomAD v4.1 (gnomad_v4) all-comers AF 0.000231825 (374/1613286 alleles, 0 homozygotes); exome AF 0.000236128, genome AF 0.000190524; maximum ancestry European (non-Finnish) 0.000291516 (344/1180038); grpmax FAF 0.00026535.
• Non-cancer subsets: gnomAD v2.1 non-cancer exomes AF 0.000119341 (28/234622 alleles, 0 homozygotes; NFE 0.000263866; grpmax FAF 0.00017776) and gnomAD v3.1 non-cancer genomes AF 0.000195999 (29/147960 alleles, 0 homozygotes; NFE 0.000324014; grpmax FAF 0.00020666).
• gnomAD-Canada v1.0 (gnomad_canada) AF 0.000543183 (10/18410 alleles, 0 homozygotes).
• Applied threshold: PM2 supporting AF <= 0.0001, the ClinGen SVI 2020 recommendation to downgrade PM2 to supporting, attributed to PMID:25741868 in the supplied calibration block. Every source listed above exceeds it, so PM2 is not satisfied on rarity grounds under the generic rule.
• The registered InSiGHT MUTYH specification supplies no retrievable PM2 rule, and the all-comers GNOMAD_V2_1/GNOMAD_V4_1 entries were used as the default source because no VCEP-specific population-source instruction was available; the non-cancer and gnomAD-Canada subsets are reported alongside for completeness and do not change the verdict.
• No publication reviewed for this criterion reports a frequency or carrier count for this variant, so PM2 rests entirely on the gnomAD frequency data above.

Still missing

• The governing InSiGHT MUTYH specification (cspec, doc 1742141534, v1.0) has an empty criteria payload in the available materials, so its PM2 threshold and any specified population source (e.g. an exome-only or non-cancer subset) could not be read; the generic 0.0001 supporting cutoff was applied instead.
• Because the lowest observed frequency (0.000119341) exceeds the generic cutoff by only ~19%, a VCEP-specific or autosomal-recessive-calibrated PM2 threshold less stringent than 0.0001 would change this verdict to met; the applicable gene-specific cutoff needs human confirmation.

Sources

Reasoning

PM3 requires that, for a gene causing a recessive disorder, the variant under assessment be detected in trans with a pathogenic variant. The governing gene-specific specification is the ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specification for MUTYH version 1.0 (cspec doc 1742141534), which is the applicable framework and confirms the relevant inheritance mode as autosomal recessive (MONDO:0012041 'familial adenomatous polyposis 2', preferred transcript NM_001128425.2). That specification is stored only as a metadata-level unstructured ruleset with an empty rule payload and no criterion-level PM3 rule text, so the generic ACMG/AMP 2015 PM3 definition was applied. This case is a variant-level assessment of the intronic change NM_001128425.2:c.1186+17C>T (NP_001121897.1:p.?, equivalent to NM_001048174.2:c.1102+17C>T), located 17 bp into an intron: there is no affected proband, no genotype, no family structure and no phase-resolved data of any kind (no segregation, allele-specific, long-read or trio phasing). No source examined in this case reports this variant in trans with a pathogenic or likely pathogenic MUTYH variant, and none reports it within a biallelic or compound-heterozygous MAP genotype. ClinVar variation 182700 (exact match) carries 8 submissions from clinical laboratories (6 Likely benign, 1 Likely Benign, 1 Benign), none expert-panel or VCEP; the submission audit recovered no criterion leads and no evidence signals beyond a generic in-silico justification, and no submission reports a cis/trans determination or a partner MUTYH allele. The case literature consists of guidance and policy documents (ACMG/AMP 2015 standards; ACMG secondary-findings lists v3.0 and v3.1; ACMG colorectal-cancer testing standards), none of which names this variant. The observation PM3 requires therefore does not exist in any available source, so PM3 is not met on the present evidence and contributes no pathogenic weight; the absence of a phased genotype rather than evidence of a benign configuration is the reason it cannot be scored either way.

Rule

PM3 (Richards et al. 2015, PMID:25741868): for recessive disorders, detected in trans with a pathogenic variant; same-criterion upgrade to strong is reserved for multiple in-trans observations. Applied generically because the governing MUTYH specification (cspec doc 1742141534) contains no machine-readable PM3 rule; MUTYH inheritance mode confirmed as autosomal recessive from that specification.

Evidence considered

• Governing framework: the applicable gene-specific specification is the ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MUTYH version 1.0 (cspec doc 1742141534). It is present in this case only as an unstructured, metadata-level ruleset (framework_mode='unstructured_ruleset', framework_complete=false, empty rule_payload, no criterion entries) whose content establishes gene/transcript/mode-of-inheritance only: MUTYH, MONDO:0012041 'familial adenomatous polyposis 2', preferred mode of inheritance 'Autosomal recessive inheritance', preferred transcript NM_001128425.2. Because no criterion-level PM3 rule is retrievable from the specification, the generic ACMG/AMP 2015 definition was applied.
• Criterion text applied (Richards et al. 2015, PMID:25741868): PM3 - 'For recessive disorders, detected in trans with a pathogenic variant.' The same paper's framework text adds that rule PM3 may be upgraded to strong 'if there were multiple observations of detection of the variant in trans (on opposite chromosomes) with other pathogenic variants', which underscores that an in-trans observation of the variant is the evidential core of PM3.
• The variant under assessment is intronic: NM_001128425.2:c.1186+17C>T (NP_001121897.1:p.?, equivalent to NM_001048174.2:c.1102+17C>T), 17 bp into the intron, i.e. not a canonical splice donor/acceptor position.
• No affected proband, genotype, family structure or phase-resolved data are present in this case: there is no segregation analysis, no allele-specific or long-read phasing, and no trio or pedigree data that could place this variant on a particular parental allele.
• ClinVar variation 182700 (exact match to NM_001048174.2(MUTYH):c.1102+17C>T): 8 clinical-laboratory submissions (6 'Likely benign', 1 'Likely Benign', 1 'Benign'), none from an expert panel or VCEP. The submission audit extracted zero criterion leads and zero evidence signals other than a generic in-silico/likely-benign justification from one laboratory; no submission reports an in-trans partner allele, a biallelic genotype, or any cis/trans phase determination, so the ClinVar record supplies no allelic evidence for PM3.
• No source examined in this case reports c.1186+17C>T in trans with a pathogenic or likely pathogenic MUTYH variant, and none reports it as part of a biallelic or compound-heterozygous MUTYH-associated polyposis genotype.
• PMID:35802134 (ACMG SF v3.1) documents MUTYH/MAP as autosomal recessive (OMIM 608456) and states that phenotypes inherited in an autosomal recessive fashion require 2 pathogenic and/or likely pathogenic variants to meet the reporting threshold, reinforcing that allelic evidence for a MUTYH variant requires a second pathogenic allele.
• PMID:24310308 (ACMG technical standards for inherited colorectal cancer testing) describes MYH-associated polyposis as an autosomal recessive condition caused by mutations in MUTYH; it reports no variant-level data and does not name c.1186+17C>T.
• No variant-specific clinical literature was retrievable: the case's gene-relevant sources are guidance/policy documents (PMID:25741868, PMID:34012068, PMID:35802134, PMID:24310308, PMID:28492532) that never name this variant, and the single case-observation-directed source (PMID:23035301, 'MUTYH Polyposis') has no retrieved full text or abstract and therefore could not be searched for an in-trans observation.

Still missing

• No proband-level or phase-resolved observation (family segregation, allele-specific or long-read phasing, or trio data) places c.1186+17C>T in trans with a pathogenic or likely pathogenic MUTYH allele; a rare in-trans configuration therefore cannot be formally excluded, only shown to be unreported.
• PMID:23035301 ('MUTYH Polyposis', the only case-observation-directed source for this gene in the case registry) has neither an abstract nor a fetched full text, so it could not be screened for a phase-resolved or in-trans observation.
• The governing InSiGHT MUTYH version 1.0 specification has no machine-readable criterion-level rule text in this case, so any VCEP-specific PM3 strength or point scheme could not be applied and the generic ACMG/AMP 2015 definition was used instead.

Sources

Reasoning

PP5 requires that the exact variant has been classified as pathogenic or likely pathogenic by a recognised ClinGen expert panel (a 3-star ClinVar assertion for that exact variant). ClinVar variation 182700 (NM_001048174.2(MUTYH):c.1102+17C>T, which is the same variant as NM_001128425.2:c.1186+17C>T) has an aggregate classification of 'Likely benign' with review status 'criteria provided, single submitter', i.e. no expert-panel review and no 3-star record. The submission audit shows 8 submissions, all from clinical laboratories (PreventionGenetics/Exact Sciences, Rigshospitalet Copenhagen, ARUP Laboratories, Labcorp Genetics/Invitae, GeneDx, Color Health, Ambry Genetics, Blueprint Genetics), each with is_expert_panel = false and review_status 'criteria provided, single submitter'; the audit's expert_panel_submissions count is 0. No submission carries a pathogenic or likely pathogenic classification. With no exact-variant expert-panel pathogenic assertion, the condition for PP5 is not satisfied.

Rule

Generic ACMG/AMP 2015 PP5: reputable source (ClinGen expert panel) reports the variant as pathogenic; applied only for an exact-variant expert-panel ClinVar assertion.

Evidence considered

• ClinVar variation 182700 aggregate: classification 'Likely benign', review_status 'criteria provided, single submitter'; variation_name NM_001048174.2(MUTYH):c.1102+17C>T with exact_match = true, matched_by transcript_hgvs.
• ClinVar submission audit: submissions_total 8, expert_panel_submissions 0; every submission has is_expert_panel = false, trust_tier 'single_submitter' (submitters: PreventionGenetics, Rigshospitalet Copenhagen, ARUP Laboratories, Labcorp Genetics/Invitae, GeneDx, Color Health, Ambry Genetics, Blueprint Genetics).
• No submission assigns pathogenic or likely pathogenic for this variant (6 Likely benign, 1 Likely Benign, 1 Benign), so neither an expert-panel nor a laboratory pathogenic assertion is present.
• The governing InSiGHT Hereditary Colorectal Cancer/Polyposis MUTYH v1.0 specification is registered for this gene, but no InSiGHT expert-panel classification of this exact variant appears in the ClinVar record.

Still missing

• No ClinGen expert-panel (InSiGHT) MUTYH classification of c.1186+17C>T exists in ClinVar at the time of review; if one is later published, PP5 must be re-evaluated.

Sources

Reasoning

BA1 is not met. The variant NM_001128425.2:c.1186+17C>T (NC_000001.10:g.45797316G>A; GRCh38 chr1-45331644-G-A) is present, not absent, in every population dataset queried, and no frequency comes close to the 0.05 stand-alone threshold (ACMG/AMP 2015, PMID:25741868). The all-comers frequencies are 0.000128372 (36/280434 alleles, 0 homozygotes) in gnomAD v2.1 and 0.000231825 (374/1613286 alleles, 0 homozygotes) in gnomAD v4.1. The highest ancestry is consistently European (non-Finnish), at 0.000256997 in v2.1 and 0.000291516 in v4.1, with grpmax FAF of 0.00017543 and 0.00026535 respectively. The non-cancer subsets are concordant with the all-comers totals (gnomAD v2.1 non-cancer exomes 0.000119341; gnomAD v3.1 non-cancer genomes 0.000195999), so inclusion of cancer cohorts is not inflating the estimate - an artefact worth checking in a cancer-predisposition gene such as MUTYH. The single highest value in any queried source, gnomAD-Canada v1.0 at 0.000543183 (10/18410 alleles), is still about 92-fold below 0.05, and the largest all-comers value is about 216-fold below it. The registered governing framework carries no retrievable BA1 rule, so the generic published default was applied; the verdict is insensitive to that choice because every candidate source sits one to two orders of magnitude away from the threshold.

Rule

{'source': 'generic_acmg_amp', 'threshold': 0.05, 'comparison': '>=', 'unit': 'allele frequency', 'population_selection': 'highest observed frequency across all-comers totals, non-cancer subsets, ancestry subgroups and gnomAD-Canada'}

Evidence considered

• gnomAD v2.1 (gnomad_v2; all-comers exomes+genomes, GRCh37 1-45797316-G-A): total AF 0.000128372 (36/280434 alleles, 0 homozygotes); exome AF 0.000120463 (30/249040), genome AF 0.000191119 (6/31394). Highest ancestry European (non-Finnish), AF 0.000256997 (33/128406, 0 homozygotes); grpmax FAF 0.00017543. Every ancestry subgroup shown (AFR, AMR, ASJ, EAS, FIN, NFE, SAS and their XX/XY splits) is at or below this; most non-European subgroups are 0.
• gnomAD v4.1 (gnomad_v4; all-comers, GRCh38 chr1-45331644-G-A): total AF 0.000231825 (374/1613286 alleles, 0 homozygotes); exome AF 0.000236128 (345/1461074), genome AF 0.000190524 (29/152212). Highest ancestry European (non-Finnish), AF 0.000291516 (344/1180038, 0 homozygotes); grpmax FAF 0.00026535.
• Non-cancer subsets reported alongside the all-comers totals: gnomAD v2.1 non-cancer exomes AF 0.000119341 (28/234622 alleles, 0 homozygotes; NFE 0.000263866; grpmax FAF 0.00017776) and gnomAD v3.1 non-cancer genomes AF 0.000195999 (29/147960 alleles, 0 homozygotes; NFE 0.000324014; grpmax FAF 0.00020666). These are essentially identical to the all-comers values, so the observed frequency is not a cancer-cohort artefact.
• gnomAD-Canada v1.0 (gnomad_canada; HostSeq genomes): AF 0.000543183 (10/18410 alleles, 0 homozygotes) - the single highest frequency in any cohort queried, and still ~92-fold below the 0.05 BA1 threshold.
• Applied threshold: BA1 stand-alone AF >= 0.05 (ACMG/AMP 2015, PMID:25741868), taken verbatim from the supplied published calibration block. The registered governing framework for this case (cspec; ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis MUTYH specification v1.0, doc 1742141534) carries an empty criteria payload, so no VCEP-specific BA1 threshold or population-source instruction could be read and the generic default with the all-comers GNOMAD_V2_1/GNOMAD_V4_1 entries was used.
• No source reviewed for this criterion reports a population frequency for this variant: PMID:25741868 (ACMG/AMP 2015, source of the BA1 definition) and PMID:34012068, PMID:35802134, PMID:24310308 contain no frequency data for MUTYH c.1186+17C>T. The ClinVar Likely benign assertion was not used as population evidence.

Sources

Reasoning

BS1 is not met. The generic strong threshold is AF >= 0.01 (ClinGen SVI default, with per-gene derivation preferred; Whiffin et al. 2017, PMID:28518168), and the highest frequency observed in any cohort, ancestry subgroup or non-cancer subset is 0.000543183 (gnomAD-Canada v1.0, 10/18410 alleles), roughly 18-fold below it. Within the all-comers datasets the maximum is 0.000291516 (gnomAD v4.1 European non-Finnish) and 0.000256997 (gnomAD v2.1 European non-Finnish); the non-cancer subsets are similar (v2.1 non-cancer exomes 0.000119341 overall and 0.000263866 in NFE; v3.1 non-cancer genomes 0.000195999 overall and 0.000324014 in NFE). No ancestry or sex-stratified subgroup reaches 0.01, and all reported homozygote counts are 0. Because MUTYH-associated polyposis is autosomal recessive, a gene-specific BS1 calibrated to disease prevalence and penetrance would be expected to sit at or below the generic default rather than above it; the observed maximum would only satisfy BS1 under a threshold below roughly 5e-4, which no available source supports. No publication reviewed for this criterion names the variant, and the registered InSiGHT MUTYH specification provides no retrievable BS1 rule, so the generic default was applied.

Rule

{'source': 'generic_svi_default', 'threshold': 0.01, 'comparison': '>=', 'unit': 'allele frequency', 'population_selection': 'highest observed frequency across all-comers totals, non-cancer subsets, ancestry subgroups and gnomAD-Canada; per-gene derivation preferred where available'}

Evidence considered

• gnomAD v2.1 (gnomad_v2) all-comers AF 0.000128372 (36/280434 alleles, 0 homozygotes), maximum ancestry European (non-Finnish) 0.000256997 (33/128406); grpmax FAF 0.00017543.
• gnomAD v4.1 (gnomad_v4) all-comers AF 0.000231825 (374/1613286 alleles, 0 homozygotes), maximum ancestry European (non-Finnish) 0.000291516 (344/1180038); grpmax FAF 0.00026535.
• Non-cancer subsets: gnomAD v2.1 non-cancer exomes AF 0.000119341 (28/234622 alleles, 0 homozygotes; NFE 0.000263866; grpmax FAF 0.00017776) and gnomAD v3.1 non-cancer genomes AF 0.000195999 (29/147960 alleles, 0 homozygotes; NFE 0.000324014; grpmax FAF 0.00020666) - both well below 0.01.
• gnomAD-Canada v1.0 (gnomad_canada) AF 0.000543183 (10/18410 alleles, 0 homozygotes) is the highest frequency in any queried source and is still ~18-fold below 0.01.
• Applied threshold: BS1 strong AF >= 0.01, the generic ClinGen/SVI default supplied in the calibration block and attributed to Whiffin et al. 2017 (PMID:28518168); per-gene derivation is preferred where calibrated population data exist, but no MUTYH-specific value was retrievable from the registered InSiGHT MUTYH specification (cspec, empty criteria payload).
• Ancestry and source selection do not change the verdict: the top frequency is European (non-Finnish) in every dataset, the non-cancer subsets track the all-comers totals, and no subgroup or cohort reaches 0.01. No paper reviewed for this criterion reports this variant.

Still missing

• No published MUTYH-specific BS1 threshold is available: the registered InSiGHT MUTYH specification has an empty criteria payload, and the GeneReviews MUTYH Polyposis chapter (PMID:23035301) was captured only as a screenshot with no extractable text, so a per-gene (prevalence/penetrance-calibrated) BS1 could not be derived or confirmed.

Sources

Reasoning

BS2 is not met. MUTYH-associated polyposis (familial adenomatous polyposis 2, MONDO:0012041) is autosomal recessive - the registered framework records 'Autosomal recessive inheritance' for MUTYH, and PMID:35802134 and PMID:24310308 both describe MAP as an autosomal recessive condition - so the applicable branch of BS2 requires the variant to have been observed in the homozygous state in a healthy individual; a heterozygous observation in a control population would not satisfy it. Homozygote counts are 0 in every dataset: gnomAD v2.1 all-comers (36 alleles), gnomAD v2.1 non-cancer exomes (28 exome alleles), gnomAD v3.1 non-cancer genomes (29 alleles), gnomAD v4.1 all-comers (374 alleles) and gnomAD-Canada v1.0 (10 alleles), and 0 in every ancestry and sex-stratified subgroup reported (including European non-Finnish, African, Admixed American, East Asian and Ashkenazi Jewish groups). The evidential requirement is therefore unmet and this criterion contributes nothing to classification. A power caveat is recorded: at the v4.1 all-comers frequency (~806,600 individuals) fewer than 0.05 homozygotes would be expected under Hardy-Weinberg, so the absence of homozygotes is uninformative rather than exculpatory and does not itself support benignity. The variant being intronic (c.1186+17C>T) and reported as Benign/Likely benign in ClinVar was not treated as BS2 evidence, and no source reports a phenotyped homozygote.

Rule

{'source': 'generic_acmg_amp', 'requirement': 'observation in a healthy individual of a homozygote (or biallelic variants) for an autosomal recessive disorder', 'population_selection': 'gnomAD homozygote counts, all datasets and ancestry subgroups'}

Evidence considered

• Homozygote counts are 0 in every source queried: gnomAD v2.1 all-comers (36 alleles, 0 hom), v2.1 non-cancer exomes (28 exome alleles, 0 hom), v3.1 non-cancer genomes (29 alleles, 0 hom), v4.1 all-comers (374 alleles, 0 hom) and gnomAD-Canada v1.0 (10 alleles, 0 hom); homozygote counts are also 0 in every ancestry and sex-stratified subgroup reported.
• Mode of inheritance: the registered framework for this case (cspec; InSiGHT MUTYH raw ruleset) records 'Autosomal recessive inheritance' for familial adenomatous polyposis 2 (MONDO:0012041), so BS2's applicable branch is the recessive (homozygous) one.
• PMID:35802134 (ACMG SF v3.1) lists MUTYH/MAP as autosomal recessive and states that pathogenic and likely pathogenic variants are reported only when 2 such variants are found; PMID:24310308 (ACMG technical standards for inherited colorectal cancer) states that MYH-associated polyposis is an autosomal recessive condition caused by MUTYH mutations. Both were consulted only to establish that the recessive (homozygote) branch applies; neither mentions c.1186+17C>T.
• Statistical power: with 1,613,286 v4.1 alleles (~806,600 individuals) at AF 0.000231825, fewer than 0.05 homozygotes would be expected under Hardy-Weinberg, so the observed zero homozygote count cannot be read as positive evidence of benignity; it simply means the BS2 requirement is not satisfied.
• The ClinVar Benign/Likely benign classification and submitter assertions were not used as population evidence for BS2, and no publication reviewed for this criterion reports a phenotyped homozygote.

Sources

Reasoning

BP2 is satisfied when the variant is observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder, or observed in cis with a pathogenic variant in any inheritance pattern. The governing specification is the ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specification for MUTYH version 1.0 (cspec doc 1742141534), which confirms MUTYH disease as autosomal recessive (MONDO:0012041 'familial adenomatous polyposis 2'); the specification is stored as an unstructured ruleset with no criterion-level BP2 rule, so the generic ACMG/AMP 2015 BP2 definition was applied. The dominant-disorder clause is structurally inapplicable here: no dominant MUTYH disorder is specified anywhere in the case, PMID:35802134 lists MAP (OMIM 608456) as 'AR' with reporting of two pathogenic/likely pathogenic variants, and PMID:24310308 likewise describes MYH-associated polyposis as an autosomal recessive condition. Only the cis clause can therefore apply to this gene. For that clause, no source provides any phase information for NM_001128425.2:c.1186+17C>T (NP_001121897.1:p.?) relative to a pathogenic MUTYH allele: the case contains no proband, no genotype and no phase-resolved data (no segregation, allele-specific or long-read phasing, no trio), and ClinVar variation 182700 contributes 8 clinical-laboratory submissions (6 Likely benign, 1 Likely Benign, 1 Benign; none expert-panel or VCEP) with no criterion leads and no evidence signals other than a generic in-silico justification, and no cis configuration reported. The case literature is guidance/policy material that never names this variant. A specific observation of the variant in cis with a pathogenic MUTYH allele - the only configuration by which BP2 could apply in this recessive gene - has thus not been made in any source, so BP2 is not met on the available evidence and contributes no benign weight.

Rule

BP2 (Richards et al. 2015, PMID:25741868): observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder, or observed in cis with a pathogenic variant in any inheritance pattern. Applied generically because the governing MUTYH specification (cspec doc 1742141534) contains no machine-readable BP2 rule; MUTYH inheritance mode confirmed as autosomal recessive from that specification, which makes the dominant-disorder clause unavailable.

Evidence considered

• Governing framework: ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MUTYH version 1.0 (cspec doc 1742141534) is the applicable gene-specific specification, but it is stored as an unstructured, metadata-level ruleset (framework_mode='unstructured_ruleset', framework_complete=false, empty rule_payload, no criterion entries) containing only gene/transcript/mode-of-inheritance metadata: MUTYH, MONDO:0012041 'familial adenomatous polyposis 2', preferred mode of inheritance 'Autosomal recessive inheritance'. No criterion-level BP2 rule was retrievable, so the generic ACMG/AMP 2015 definition was applied; the specification does establish the inheritance mode, which removes the dominant-disorder clause of BP2 from consideration.
• Criterion text applied (Richards et al. 2015, PMID:25741868): BP2 - 'Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.'
• PMID:35802134 (ACMG SF v3.1) lists MUTYH/MAP as autosomal recessive (OMIM 608456) with pathogenic and likely pathogenic variants reportable only when 2 such variants are found, and states that phenotypes inherited in an autosomal recessive fashion need 2 pathogenic and/or likely pathogenic variants to meet the reporting threshold; no dominant MUTYH disorder is listed.
• PMID:24310308 (ACMG technical standards for inherited colorectal cancer testing) describes MYH-associated polyposis as an autosomal recessive condition caused by mutations in the MUTYH gene, confirming that the only allelic configuration available to BP2 in this gene is cis with a pathogenic variant.
• No phase information of any kind exists in this case for NM_001128425.2:c.1186+17C>T (NP_001121897.1:p.?, equivalent to NM_001048174.2:c.1102+17C>T): there is no proband, no genotype, no segregation analysis, no allele-specific or long-read phasing and no trio data.
• ClinVar variation 182700 (exact match): 8 clinical-laboratory submissions (6 'Likely benign', 1 'Likely Benign', 1 'Benign'), none expert-panel or VCEP; the submission audit recovered no criterion leads and no evidence signals beyond a generic in-silico/likely-benign justification from one laboratory, and no submission reports the variant in cis (or trans) with a pathogenic MUTYH allele, so the record cannot support BP2.
• No source examined in this case reports c.1186+17C>T in cis with a pathogenic or likely pathogenic MUTYH allele, and none reports it in a biallelic or compound-heterozygous MAP genotype in which a pathogenic cis partner could be identified.
• The case literature is guidance and policy material (PMID:25741868, PMID:34012068, PMID:35802134, PMID:24310308, PMID:28492532); none names this variant, and the only case-observation-directed source (PMID:23035301, 'MUTYH Polyposis') has no retrieved full text or abstract and therefore could not be searched for a cis observation.

Still missing

• No cis/trans phasing exists for c.1186+17C>T against any pathogenic MUTYH allele in the sources examined, so the cis configuration BP2 requires is unobserved rather than formally refuted; haplotype-resolved or family-based phase data would be needed to close this.
• PMID:23035301 ('MUTYH Polyposis') is the only case-observation-directed source for this gene in the case registry and has no abstract or fetched full text, so it could not be screened for a cis co-occurrence with a pathogenic allele.
• The governing InSiGHT MUTYH version 1.0 specification provides no machine-readable criterion-level BP2 rule text in this case, so any VCEP-specific BP2 application or exclusion could not be verified against the specification and the generic ACMG/AMP 2015 definition was used.

Sources

Reasoning

BP6 requires that the exact variant has been classified as benign or likely benign by a recognised ClinGen expert panel (a 3-star ClinVar assertion for that exact variant). ClinVar variation 182700 (NM_001048174.2(MUTYH):c.1102+17C>T, exact match to NM_001128425.2:c.1186+17C>T) does carry an aggregate 'Likely benign' label, but that label's review status is 'criteria provided, single submitter' with no expert-panel review, and the submission audit shows 0 expert-panel submissions among 8 submissions from individual clinical laboratories (6 Likely benign, 1 Likely Benign, 1 Benign). The aggregate label is therefore a summation of independent laboratory submissions, not an expert-panel consensus assertion, and per the governing rule aggregate ClinVar labels and ordinary laboratory submissions must never trigger BP6. No exact-variant expert-panel benign assertion exists, so BP6 is not met despite the benign-leaning laboratory consensus.

Rule

Generic ACMG/AMP 2015 BP6: reputable source (ClinGen expert panel) reports the variant as benign; applied only for an exact-variant expert-panel ClinVar assertion.

Evidence considered

• ClinVar variation 182700 aggregate: classification 'Likely benign', review_status 'criteria provided, single submitter' (no expert-panel review); submission_summary 6 laboratories Likely benign, 1 Benign, 1 Likely Benign.
• ClinVar submission audit: expert_panel_submissions = 0; all 8 submissions have is_expert_panel = false and trust_tier 'single_submitter' (PreventionGenetics, Rigshospitalet Copenhagen, ARUP Laboratories, Labcorp Genetics/Invitae, GeneDx, Color Health, Ambry Genetics, Blueprint Genetics).
• GeneDx submission (SCV000211423, Benign) cites only PP3-type in-silico reasoning and no PMID; other submissions cite only generic ACMG references (PMID:25741868, PMID:28492532), none providing an expert-panel benign assertion for this variant.
• Aggregate ClinVar classification labels from non-expert submitters do not satisfy BP6 under the governing rule for this criterion.

Still missing

• No ClinGen expert-panel (InSiGHT) MUTYH classification of c.1186+17C>T exists in ClinVar at the time of review; if an expert panel later classifies this exact variant as benign/likely benign, BP6 would need re-evaluation.

Sources

Reasoning

PS2 requires a confirmed de novo occurrence (maternity and paternity both confirmed by testing) in a patient with the disease phenotype and no family history. No proband, trio, parental-testing, pedigree or phenotype data exist for this variant in any curated source: the ClinVar record for variation 182700 carries eight clinical-testing submissions with zero criterion leads and no segregation or de novo assertions, the governing InSiGHT MUTYH specification retrieval provides only gene-level metadata (no PS2 rule payload), and none of the screened publications report an affected proband or transmission information for c.1186+17C>T. Because no de novo occurrence is observed, PS2 cannot be met; because no parental testing was performed or reported, it also cannot be excluded. MUTYH is an autosomal recessive gene, so even a confirmed de novo event affecting one allele would need biallelic context before it could support pathogenicity. The criterion is therefore recorded as not assessed, with the missing proband/trio evidence identified as the gap.

Rule

ACMG/AMP 2015 PS2: de novo (both maternity and paternity confirmed) in a patient with the disease and no family history.

Evidence considered

• PS2 under ACMG/AMP 2015 requires a de novo variant confirmed in the proband with both maternity and paternity established by testing, in an individual with the phenotype and no family history; neither condition can be evaluated here because no proband-level data exist.
• The ClinVar record for variation 182700 (NM_001048174.2:c.1102+17C>T / NM_001128425.2:c.1186+17C>T) consists of eight germline clinical-testing submissions (six 'Likely benign', one 'Likely Benign', one 'Benign'); zero submissions were from an expert panel and the submission audit yielded zero criterion leads, so no PS2 (or any other) evidence lead is available from that record.
• The governing InSiGHT Hereditary Colorectal Cancer/Polyposis VCEP specification for MUTYH (CSPEC doc 1742141534, v1.0) was retrieved; it supplies gene-level metadata only (autosomal recessive, familial adenomatous polyposis 2, MONDO:0012041) and no machine-readable PS2 rule or threshold, so the generic ACMG/AMP 2015 definition was used.
• Screening of the already-fetched full texts (PMIDs 25741868, 34012068, 35802134, 24310308) found no report of this variant in any patient, no trio/parental testing, and no de novo observation; the variant is not mentioned in any of them.
• MUTYH disease is autosomal recessive (biallelic loss of function), so a single de novo allele would not by itself satisfy the gene-disease relationship required for PS2 evidence.

Still missing

• No affected proband with MUTYH-associated polyposis phenotype reported for this variant.
• No trio/parental testing (maternity and paternity confirmation) available to establish de novo status.
• No family history information available to satisfy the 'no family history' component of PS2.

Sources

Reasoning

Governing framework: the ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specification for MUTYH v1.0 (doc 1742141534) was consulted first, but its criterion rule payload is not populated in this case (no gene-specific criteria or strength calibrations extracted), so generic ACMG/AMP 2015 (PMID:25741868) governs. Under that framework PS3 requires well-established functional studies demonstrating a deleterious effect on gene or protein function. Variant context: NM_001128425.2:c.1186+17C>T is a deep intronic change 17 bp into intron 12, so a PS3 claim would have to rest on a validated splice/transcript assay (minigene, patient-derived RNA RT-PCR or RNA-seq showing aberrant splicing) or an equivalent protein-level assay; in-silico splice prediction, even when available, is PP3 evidence and not PS3. Evidence found: none. No RNA/splice or protein functional assay for this variant exists anywhere in the case evidence. The seven triaged PMIDs contain no MUTYH functional study; the four full texts that were fetched and extracted (PMID:25741868, PMID:24310308, PMID:34012068, PMID:35802134) do not mention the variant or report any assay, and the ClinVar submission audit recorded zero criterion leads usable for functional evidence. The SpliceAI lookup returned no scores at all (query timeout), so not even a calibrated in-silico splice signal is available, and the case annotation explicitly states that this missing result must not be read as evidence of absent splice impact. The only splice-prediction numbers present are Pangolin (splice gain 0.041, splice loss -0.002), which is an uncalibrated in-silico prediction with no ACMG-strength calibration and is not assay evidence. Because the requirement is a validated functional assay rather than a prediction, PS3 cannot be evaluated on the available evidence and is recorded as not assessed rather than not met. Note also that this variant's benign-leaning ClinVar submissions were not used to reason about PS3, since a source's final classification must never be substituted for functional assay data.

Rule

PS3 (generic ACMG/AMP 2015, PMID:25741868): Well-established functional studies show a deleterious effect. Applied at default strength, requiring a validated functional (or transcript/splice) assay; no MUTYH VCEP-specific PS3 rule was retrievable.

Evidence considered

• Consulted the ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specification for MUTYH v1.0 (cspec doc 1742141534) as the governing framework; its criterion rule payload is empty/not populated for this case, so no MUTYH-specific PS3 rule or strength calibration is available and the generic ACMG/AMP 2015 criteria (Richards et al., PMID:25741868) apply.
• Generic ACMG/AMP 2015 defines PS3 as 'well-established functional studies show a deleterious effect'; the guideline also states that for splice-region variants confirmation of splice impact requires functional analysis by RNA or protein methods, not prediction alone.
• No functional assay of any kind (minigene, RT-PCR/transcript analysis, RNA-seq, or protein-based assay) for MUTYH c.1186+17C>T was present in the case evidence, and a deep intronic +17 variant can only reach PS3 through such an assay.
• SpliceAI lookup for NM_001128425.2:c.1186+17C>T returned no scores (query timed out; max delta null), leaving no calibrated splice prediction; the case annotation states this is missing data and must not be used to support BP4 or to argue against PP3/PVS1.
• Pangolin output (splice gain 0.041, splice loss -0.002) is an uncalibrated in-silico prediction, not a functional assay, and has no ACMG-strength calibration for PS3.
• All seven triaged PMIDs were reviewed; the only paper triaged to the functional theme was the ACMG/AMP 2015 guideline itself, and none of the fetched full texts reports functional data on this variant. Variant-mention screening across the publication queue returned zero hits for c.1186+17C>T or NP_001121897.1:p.?, so no citation could be grounded to this variant.
• PMID:23035301 (MUTYH Polyposis, GeneReviews), PMID:26389505 (Genetics of Colorectal Cancer, PDQ) and PMID:28492532 (Sherloc) had no retrievable full text in this case (screenshots/abstract only) and were triaged to case-observation/background themes; nothing in their available content reports a functional assay for this variant.

Still missing

• No RNA/splice functional assay (minigene, patient-transcript RT-PCR or RNA-seq) was found or retrieved for MUTYH c.1186+17C>T; PS3 cannot be evaluated without one.
• SpliceAI lookup timed out and returned no scores, so no calibrated splice prediction is available for this intronic variant.
• The InSiGHT MUTYH VCEP criterion payload (doc 1742141534) is not populated locally, so no gene-specific PS3 strength calibration could be applied.

Sources

Reasoning

PM6 requires an assumed de novo occurrence without confirmation of maternity and paternity, again in a patient with the disease phenotype and no family history. No de novo observation of any kind is reported for c.1186+17C>T: the ClinVar record contains no de novo assertions or criterion leads, the curated sources contain no proband or pedigree, and none of the screened publications mention the variant at all. With neither a de novo observation nor an affected proband, the criterion cannot be met, and the absence of parental testing means de novo status is neither confirmed nor assumed. Recorded as not assessed, with the missing proband/family observation identified as the gap. MUTYH is autosomal recessive, so an assumed de novo single-allele event would also require biallelic context to carry pathogenicity evidence.

Rule

ACMG/AMP 2015 PM6: assumed de novo, but without confirmation of paternity and maternity.

Evidence considered

• PM6 covers an assumed de novo event when maternity/paternity have not been tested; it still requires the variant to be observed as apparently de novo in an individual with the phenotype and without a family history.
• The ClinVar record for variation 182700 lists eight germline clinical-testing submissions with no criterion leads and no de novo or inheritance assertions, so no assumed-de-novo observation can be drawn from it.
• The InSiGHT MUTYH VCEP specification available for this case (CSPEC doc 1742141534, v1.0) supplies no machine-readable PM6 rule or pedigree requirement, so the generic ACMG/AMP 2015 definition was applied.
• The variant is not mentioned in any of the fetched full texts (PMIDs 25741868, 34012068, 35802134, 24310308), and no case report or cohort linking it to an affected individual was identified, so no apparently de novo observation exists in the evidence for this case.
• MUTYH-associated polyposis is autosomal recessive, which limits the interpretive weight any single de novo allele could carry.

Still missing

• No observed de novo occurrence of the variant (confirmed or unconfirmed) in an affected individual.
• No proband or affected-individual data linking c.1186+17C>T to the MUTYH-associated polyposis phenotype.
• No family history information to satisfy the 'no family history' requirement.

Sources

Reasoning

PP1 requires co-segregation of the variant with disease in multiple affected family members of a gene definitively known to cause the disease, which in turn requires a pedigree with genotyped affected and unaffected relatives (informative meioses). The evidence assembled for this case contains no pedigree, no affected relatives, no genotyping of relatives and therefore zero informative meioses for c.1186+17C>T: the ClinVar record holds only laboratory clinical-testing submissions with no criterion leads or family data, the governing InSiGHT MUTYH specification retrieval produced no rule payload or segregation thresholds, and no screened publication reports this variant in any family. Co-segregation therefore cannot be demonstrated, and the criterion is recorded as not assessed with the absence of pedigree/segregation data (zero informative meioses) named as the gap.

Rule

ACMG/AMP 2015 PP1: co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.

Evidence considered

• PP1 requires co-segregation with disease in multiple affected family members; it is a pedigree-based criterion and cannot be evaluated without relatives whose genotype and phenotype status are known.
• The ClinVar record for variation 182700 contains eight germline clinical-testing submissions (six 'Likely benign', one 'Likely Benign', one 'Benign'), no expert-panel submission, and zero criterion leads; no submission reports family members, a pedigree, or segregation data.
• The InSiGHT MUTYH VCEP specification retrieved for this case (CSPEC doc 1742141534, v1.0) is an unapproved ruleset in preparation with an empty criteria payload, so no MUTYH-specific PP1 definition, meioses requirement, or strength modification could be applied; the generic ACMG/AMP 2015 definition was used.
• The variant is not mentioned in any fetched full text (PMIDs 25741868, 34012068, 35802134, 24310308); no publication was found that genotypes relatives of a carrier of c.1186+17C>T, so there are no informative meioses for this variant.
• Because the variant is intronic and reported in the general population (gnomAD v2.1 total AF 0.000128, 36/280434 alleles; gnomAD v4.1 total AF 0.000232, 374/1613286 alleles), a positive co-segregation signal would in any case require a family with multiple affected carriers - none exists in this case.

Still missing

• No pedigree or family structure available for any individual carrying c.1186+17C>T.
• No genotyping of affected and unaffected relatives, hence zero informative meioses with which to test co-segregation.
• No gene-specific PP1 co-segregation rule retrievable from the governing MUTYH specification.

Sources

Reasoning

SCOPE AND PATH SELECTION (settled before any score was read): the variant is NM_001128425.2:c.1186+17C>T (NC_000001.10:g.45797316G>A / NC_000001.11:g.45331644G>A), a single-nucleotide C>T substitution located 17 bp into intron 12 of the ClinGen-preferred MUTYH transcript NM_001128425.2, outside the canonical +/-1,2 donor dinucleotide. The variant-level consequence scaffold records consequence_class 'unknown' and canonical_splice_consensus false, and no amino-acid substitution is predicted (NP_001121897.1:p.?). PP3 is therefore in scope for this variant, but only through the SpliceAI path, because the variant is intronic rather than missense; the REVEL/missense path does not apply and its scores were not sought or used (the record also carries revel found=false and bayesdel found=false, score null in both cases). GOVERNING FRAMEWORK: the MUTYH InSiGHT Hereditary Colorectal Cancer/Polyposis VCEP specification v1.0 (CSPEC doc 1742141534) was located for this gene but exposes an empty criteria list and an empty rule payload (framework_complete=false, framework_mode 'unstructured_ruleset'), so it provides no gene-specific PP3 cutoff or decision tree; the generic ACMG/AMP in-silico rules therefore govern, and the published SpliceAI calibration is used verbatim (PP3 supporting >= 0.2, PP3 moderate >= 0.5; Jaganathan et al. 2019, Cell, PMID:30661751). EVIDENCE: the SpliceAI lookup for this variant returned NO scores (spliceai_available false, max_delta_score null, spliceai_error 'timeout: The read operation timed out'); an independent retry of the SpliceAI lookup API from this review session also failed with a network timeout, so the score could not be recovered. This is missing data, not a value of 0.00, and the source's own annotation states it must not be used to argue against PP3. The only splice-model output actually present is Pangolin (splice gain 0.041, splice loss -0.002); Pangolin has no ACMG-strength calibration in the supplied published thresholds and none in the (empty) MUTYH VCEP specification, so its values are not_available for PP3 and are not a substitute for SpliceAI. No threshold can therefore be tested: a SpliceAI max delta of >= 0.2 is required for even supporting PP3 and no such delta exists. CONCLUSION: PP3 is not_assessed. The criterion is in scope but the one predictor path calibrated for it produced no result, so this is an unresolved evidence gap rather than an evaluated-and-failed threshold; no strength tier is assigned. Human review is flagged so the SpliceAI lookup can be re-run. Not double-counted: no predictor value in this case was consumed by more than one of PP3/BP4/BP7.

Rule

Generic SpliceAI path for an intronic variant (Jaganathan et al. 2019, Cell, PMID:30661751): PP3 supporting at max delta >= 0.2, moderate at >= 0.5. No gene-specific MUTYH PP3 rule exists in the InSiGHT CSPEC specification v1.0 (empty payload). Criterion could not be evaluated because no SpliceAI max delta score was returned.

Evidence considered

• Variant identity and consequence: NM_001128425.2:c.1186+17C>T, a C>T substitution 17 bp into intron 12 of the ClinGen-preferred MUTYH transcript, with no predicted amino-acid change (NP_001121897.1:p.?); the variant-level scaffold records consequence_class 'unknown', variant_bucket 'other' and canonical_splice_consensus false. This places the variant on the intronic/splice-region path, so PP3 must be taken from SpliceAI and not from a missense predictor.
• No missense-path score exists or was used: the in-silico block records REVEL score null (found false) and BayesDel score null (found false), consistent with the variant not being a missense change; neither was consulted for a threshold decision.
• SpliceAI lookup returned no scores for this variant (spliceai_available false, max_delta_score null, spliceai_error 'timeout: The read operation timed out'). This is missing data rather than a delta of 0.00, and the source's own sentence states it must not be used to support BP4 or to argue against PP3. A direct retry of the SpliceAI lookup API from this review session also timed out, so the value could not be recovered.
• Published SpliceAI thresholds applied verbatim but untestable here: PP3 supporting requires max delta >= 0.2 and PP3 moderate >= 0.5 (Jaganathan et al. 2019, Cell, PMID:30661751). No max delta exists for this variant, so neither threshold can be met.
• Pangolin is the only splice predictor with a value for this variant (splice gain 0.041, splice loss -0.002). No ACMG-strength calibration for Pangolin is supplied in the published thresholds and none exists in the MUTYH VCEP specification, so these values are not_available for PP3 and cannot substitute for the missing SpliceAI result.
• The MUTYH InSiGHT Hereditary Colorectal Cancer/Polyposis VCEP/CSPEC specification v1.0 (doc 1742141534) was found for this gene but provides an empty criteria list and empty rule payload, so the generic ACMG/AMP in-silico framework and the published SpliceAI calibration were used instead of a gene-specific PP3 rule.
• The ACMG/AMP 2015 guideline (Richards et al.) treats in-silico splice predictors as a single, limited-specificity line of evidence and cautions that prediction alone does not establish a splice effect; it supplies no score for this variant.

Still missing

• SpliceAI returned no score for c.1186+17C>T (lookup timeout; max_delta_score null, spliceai_available false), and an independent retry of the SpliceAI lookup API from this session also timed out. Without a max delta score the generic PP3 cutoffs (supporting >= 0.2, moderate >= 0.5) cannot be applied; re-running SpliceAI is the single action needed to close this gap.
• Pangolin (splice gain 0.041, splice loss -0.002) is the only splice-model output available for this variant and has no ACMG-strength calibration in the supplied published thresholds or in the MUTYH InSiGHT VCEP specification v1.0 (empty payload), so it cannot be used to establish PP3.
• The MUTYH InSiGHT CSPEC record (doc 1742141534, v1.0) was retrieved but contains no criteria payload, so no gene-specific PP3 rule could be applied in place of the generic calibration.

Sources

Reasoning

PP4 requires that the proband's phenotype and/or family history be highly specific for a disease with a single-gene aetiology. No proband-level clinical information (adenomatous polyposis burden, colorectal cancer history, age at onset, family history, or indication for testing) accompanies this variant-level case, and the ClinVar submissions for variation 182700 carry condition labels of 'not provided'/'not specified' only, so there is no phenotype-specificity information to weigh against MUTYH-associated polyposis. The variant is intronic (NM_001128425.2:c.1186+17C>T, 17 bp into intron 12) and is present in population databases (gnomAD v2.1 total AF 0.000128372; 36/280434 alleles, 0 homozygotes), which is population data, not proband phenotype. Because the required proband phenotype/family history is absent, PP4 cannot be adjudicated as met or not met on the available record.

Rule

Generic ACMG/AMP 2015 PP4: patient's phenotype and/or family history is highly specific for a disease with a single genetic aetiology.

Evidence considered

• ClinVar variation 182700: 8 clinical-laboratory submissions with condition_labels limited to 'not provided'/'not specified'; no phenotype or family-history detail is recorded.
• No clinical case record, pedigree, or phenotype description accompanies this variant-level adjudication; nothing in the retrieved evidence describes an affected individual carrying c.1186+17C>T.
• The gene-disease context is MUTYH-associated polyposis / familial adenomatous polyposis 2 (MONDO:0012041, autosomal recessive, MUTYH), but no individual's presentation is available to test specificity against that phenotype.

Still missing

• No proband phenotype, age at onset, polyp count, colorectal cancer history, or family history is available in this case, so phenotype specificity for MUTYH-associated polyposis cannot be judged.
• ClinVar variation 182700 submissions carry only 'not provided'/'not specified' condition labels, giving no usable phenotype description.

Sources

Reasoning

Governing framework: the ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specification for MUTYH v1.0 (doc 1742141534) was consulted first; its criterion rule payload is not populated in this case, so no MUTYH-specific BS3 rule or strength calibration is available and generic ACMG/AMP 2015 (PMID:25741868) governs. Under that framework BS3 requires well-established functional studies showing no damaging effect on gene or protein function. Variant context: NM_001128425.2:c.1186+17C>T is a deep intronic variant 17 bp into intron 12, so a BS3 claim would require a validated assay demonstrating normal splicing or normal protein function for the variant allele. Evidence found: none. No functional assay for this variant exists in the case evidence; no paper in the seven-PMID triage set reports functional data on it, and none of the four extracted full texts (PMID:25741868, PMID:24310308, PMID:34012068, PMID:35802134) mentions the variant. Critically, the two reasons that might superficially look like BS3 support do not qualify: (1) the SpliceAI lookup returned no scores at all (query timeout), and the case annotation states explicitly that this missing result is not evidence of absent splice impact and must not be used benignly; (2) the ClinVar record (Benign/Likely benign, 2 stars, eight submissions) reflects submitter classifications, and a source's final classification must never be converted into BS3 - the ClinVar submission audit found zero criterion leads usable as evidence for this criterion. The only splice numbers available are Pangolin (splice gain 0.041, splice loss -0.002), which is an uncalibrated in-silico prediction, not a validated functional assay, and cannot satisfy BS3. BS3 is therefore recorded as not assessed: the criterion's evidentiary requirement (a validated functional assay) is absent rather than contradicted.

Rule

BS3 (generic ACMG/AMP 2015, PMID:25741868): Well-established functional studies show no damaging effect. Applied at default strength, requiring a validated functional/transcript assay; no MUTYH VCEP-specific BS3 rule was retrievable.

Evidence considered

• Consulted the ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specification for MUTYH v1.0 (cspec doc 1742141534); its criterion rule payload is empty/not populated for this case, so no MUTYH-specific BS3 rule or strength is available and generic ACMG/AMP 2015 (Richards et al., PMID:25741868) governs.
• Generic ACMG/AMP 2015 defines BS3 as 'well-established functional studies show no damaging effect'; a validated assay (including a transcript/splice assay for a deep intronic variant such as c.1186+17C>T) is required.
• No validated functional or transcript assay for MUTYH c.1186+17C>T was present in the case evidence; nothing in the seven triaged papers or the four extracted full texts reports functional data on this variant.
• SpliceAI lookup returned no scores (query timed out; max delta null). The case annotation states this is missing data and must not be used to support BP4, so it cannot be used to support BS3 either.
• Pangolin output (splice gain 0.041, splice loss -0.002) is uncalibrated in-silico prediction with no ACMG-strength calibration and is not a validated functional assay, so it cannot satisfy BS3.
• ClinVar's Benign/Likely benign record (8 submissions, 2-star, no expert-panel review) was not treated as functional evidence: the ClinVar submission audit reported zero criterion leads usable for classification, and a source's final classification must never be used to infer a criterion.

Still missing

• No validated functional or transcript assay demonstrating a non-damaging (normal splicing/normal function) effect was found for MUTYH c.1186+17C>T; BS3 cannot be evaluated without one.
• SpliceAI returned no scores (query timeout); a successful run is a prerequisite for any calibrated splice-based reasoning, though splice prediction alone would still not constitute BS3 assay evidence.
• The InSiGHT MUTYH VCEP criterion payload (doc 1742141534) is not populated locally, so no gene-specific BS3 strength calibration could be applied.

Sources

Reasoning

BS4 requires lack of segregation of the variant with the disease in affected family members, which presupposes an informative family in which several affected relatives have been genotyped and shown not to carry the variant. No family, pedigree or genotype data of any kind exist for c.1186+17C>T in the assembled evidence: the ClinVar record for variation 182700 consists of laboratory clinical-testing submissions with no criterion leads or family-level information, the governing InSiGHT MUTYH specification retrieval provided gene-level metadata only, and none of the screened publications report this variant in any patient or family. Absence of segregation is therefore not demonstrable rather than inferable, and it must not be substituted with other benign sources: the multiple 'Likely benign'/'Benign' ClinVar submissions and the population frequency of this variant are not BS4 evidence and were not used as such. Recorded as not assessed, with the absence of an informative pedigree named as the gap.

Rule

ACMG/AMP 2015 BS4: lack of segregation in affected members of a family.

Evidence considered

• BS4 is a benign criterion that requires demonstrated lack of co-segregation of the variant with disease among affected members of a family, i.e. affected relatives who do not carry the variant.
• The ClinVar record for variation 182700 reports eight germline clinical-testing submissions (six 'Likely benign', one 'Likely Benign', one 'Benign') with zero criterion leads and no family, pedigree, or segregation information; the record's overall classification was not used as BS4 evidence.
• The InSiGHT MUTYH VCEP specification retrieved for this case (CSPEC doc 1742141534, v1.0) provides no BS4 rule payload, so the generic ACMG/AMP 2015 definition was applied.
• The variant is not mentioned in any fetched full text (PMIDs 25741868, 34012068, 35802134, 24310308), and no publication or cohort was found that genotypes affected relatives of a carrier, so no non-segregation observation exists for this variant.
• Population frequency (gnomAD v2.1 total AF 0.000128, 36/280434 alleles, 0 homozygotes; gnomAD v4.1 total AF 0.000232, 374/1613286 alleles, 0 homozygotes; gnomAD-Canada AF 0.000543, 10/18410 alleles) and ClinVar's classification were deliberately not used to meet BS4, which is a family-segregation criterion.

Still missing

• No family with multiple affected members in which non-carrier status could be tested for c.1186+17C>T.
• No genotype/phenotype data for relatives, so non-segregation cannot be observed.
• ClinVar's aggregated Likely benign/Benign classifications and population frequency data are not acceptable substitutes for BS4 segregation evidence.

Sources

Reasoning

SCOPE AND PATH SELECTION: BP4 was evaluated only on the path matching the variant's consequence type. NM_001128425.2:c.1186+17C>T is an intronic single-nucleotide substitution (17 bp into intron 12 of the ClinGen-preferred MUTYH transcript, outside the canonical +/-1,2 donor site; consequence_class 'unknown', canonical_splice_consensus false), so BP4 here can only come from the SpliceAI path. It is not a missense variant, so the missense (REVEL) path was not used, and no splice prediction was used as BP4 support for a missense reading. GOVERNING FRAMEWORK: the MUTYH InSiGHT Hereditary Colorectal Cancer/Polyposis VCEP specification v1.0 (CSPEC doc 1742141534) was located but has an empty criteria list and rule payload (framework_complete=false), so no gene-specific BP4 rule exists and the generic ACMG/AMP in-silico calibration governs (BP4 supporting at SpliceAI max delta <= 0.1; Jaganathan et al. 2019, Cell, PMID:30661751). EVIDENCE: SpliceAI returned no score for this variant (spliceai_available false, max_delta_score null, error 'timeout: The read operation timed out'), and an independent retry of the SpliceAI lookup API from this session also timed out. The source's own annotation is explicit that this missing result is NOT evidence of absent splice impact and 'must not be used to support BP4'. bp4 therefore cannot be satisfied on its own terms: the <= 0.1 threshold requires an actual max delta score, and none exists. The only predictor with a value, Pangolin (splice gain 0.041, splice loss -0.002), is uncalibrated (no published ACMG-strength cutoff and none in the MUTYH VCEP specification) and so is not_available for BP4; it cannot be promoted into BP4 support simply because SpliceAI failed. CONCLUSION: BP4 is not_assessed, not not_met: the applicable predictor produced no result, so the threshold was never testable, and 'no SpliceAI score' is missing data rather than a benign prediction. No strength tier is assigned. Human review is flagged so SpliceAI can be re-run. No double-counting: no predictor value in this case was used for more than one of PP3/BP4/BP7.

Rule

Generic SpliceAI path for an intronic variant (Jaganathan et al. 2019, Cell, PMID:30661751): BP4 supporting at max delta <= 0.1. No gene-specific MUTYH BP4 rule exists in the InSiGHT CSPEC specification v1.0 (empty payload). Criterion could not be evaluated because no SpliceAI max delta score was returned.

Evidence considered

• BP4 path is the SpliceAI path only: NM_001128425.2:c.1186+17C>T is an intronic substitution 17 bp into intron 12 of the ClinGen-preferred MUTYH transcript (consequence_class 'unknown', canonical_splice_consensus false, protein NP_001121897.1:p.?), so it is not a missense variant and the REVEL path is not applicable to it.
• No missense-path score exists: the in-silico block records REVEL score null (found false) and BayesDel score null (found false); neither a REVEL gray-zone result nor any other missense score was available to be, or was, repurposed as BP4 support.
• SpliceAI lookup returned no scores for this variant (spliceai_available false, max_delta_score null, spliceai_error 'timeout: The read operation timed out'); the source's own sentence states this is missing data, is not evidence of absent splice impact, and must not be used to support BP4. A direct retry of the SpliceAI lookup API from this review session also timed out.
• Published SpliceAI threshold applied verbatim but untestable here: BP4 supporting requires max delta <= 0.1 (Jaganathan et al. 2019, Cell, PMID:30661751). With no max delta value, the threshold could not be evaluated either way.
• Pangolin (splice gain 0.041, splice loss -0.002) is the only available splice-model output; it carries no ACMG-strength calibration in the supplied published thresholds and none in the MUTYH VCEP specification, so it is not_available for BP4 and cannot substitute for the failed SpliceAI query.
• The MUTYH InSiGHT Hereditary Colorectal Cancer/Polyposis VCEP/CSPEC specification v1.0 (doc 1742141534) exposes no criteria or rule payload, so only the generic ACMG/AMP in-silico framework could be applied.
• The ACMG/AMP 2015 guideline (Richards et al.) describes in-silico splice predictions as a single line of evidence with limited specificity that cannot alone support a clinical assertion, and provides no computational result for this variant.

Still missing

• SpliceAI returned no score for c.1186+17C>T (lookup timeout; max_delta_score null, spliceai_available false), and an independent retry of the SpliceAI lookup API from this session also timed out. The generic BP4 cutoff (supporting at max delta <= 0.1) cannot be applied without a delta score, and the missing result must not be read as 0.00 or as benign. Re-running SpliceAI would close this gap.
• Pangolin (splice gain 0.041, splice loss -0.002) is the only splice-model value available and is uncalibrated for ACMG strength in the supplied published thresholds and in the MUTYH InSiGHT VCEP specification v1.0 (empty payload), so it cannot establish BP4.
• No MUTYH-specific BP4 rule exists in the InSiGHT CSPEC record (doc 1742141534, v1.0), whose criteria and rule payload are empty.

Sources

Reasoning

BP5 requires that the variant be observed in a patient who has an alternate molecular basis for the disease (for example, a pathogenic variant in another gene that explains the phenotype, or a clearly non-Mendelian/acquired explanation). No individual-level case data are available for this variant: no affected proband, no phenotype, no information about testing of other genes, and no second-variant or alternate-cause data accompany the case. None of the seven triaged publications reports NM_001128425.2:c.1186+17C>T (variant_mentioned = false for all), and the eight ClinVar submissions for variation 182700 record no case-level clinical narrative, alternate genetic diagnosis, or testing context beyond the submitting laboratory and condition label. With no case in which the variant was identified together with an alternate molecular basis for disease, the criterion's condition cannot be evaluated and is neither met nor affirmatively excluded.

Rule

Generic ACMG/AMP 2015 BP5: variant found in a case with an alternate molecular basis for disease.

Evidence considered

• ClinVar variation 182700: 8 single-submitter laboratory submissions with no case narrative, no co-occurring variant information, and no alternate-diagnosis detail; condition labels are 'not provided'/'not specified'.
• Literature screen for the exact variant across all 7 triaged PMIDs returned no individual-level case report (variant_mentioned = false for every paper).
• No proband phenotype, family history, or additional-gene testing results are available for this variant-level adjudication, so an alternate molecular basis for disease cannot be established or excluded.

Still missing

• No proband-level data (phenotype, other-gene testing, co-occurring pathogenic variants) exist for this case, which BP5 requires before the criterion can be assessed.
• No publication retrieved in this case reports the variant in an individual, with or without an alternate molecular basis for disease.

Sources