MSH3 c.1360C>T, p.(Arg454Ter)

Reasoning

PVS1 is met at full, un-downgraded strength, written as the strength tier 'very strong'. NM_002439.5:c.1360C>T (NC_000005.9:g.80021291C>T) is a nonsense substitution on NM_002439.5, the MANE Select and RefSeq Select MSH3 transcript, predicted to produce NP_002430.3:p.(Arg454Ter). Normalization confirms the premature stop truncates the 1137-residue MSH3 protein after codon 454, removing ~684 residues (~60% of the protein), and VariantValidator places the change in exon 9 of 24 exons (c.1341-1453), far from the transcript 3' end. MSH3 loss of function is the established disease mechanism: biallelic germline MSH3 null alleles (exon 7 frameshift p.Lys383Argfs*32, intron 16 c.2319-1G>A, exon 20 frameshift p.Tyr921Metfs*36, intron 21 c.3001-2A>C) cause autosomal-recessive colorectal adenomatous polyposis with complete loss of nuclear MSH3 and tumour EMAST (PMID:27476653), and biallelic MSH3 variants are corroborated as a recurrent polyposis cause in a second independent series (PMID:37402566). The governing framework is the generic ClinGen SVI PVS1 decision tree (PMC6185798), because MSH3 has no ClinGen CSPEC and no VCEP or local specification (cspec.found=false; vcep_materials.final_classification_framework.source=generic_acmg_fallback, framework_complete=false); the generic ACMG/AMP 2015 framework requires that null variants be a known mechanism consistent with the gene's inheritance pattern (PMID:25741868). Applying the tree: (1) the variant is a null allele (nonsense, non-canonical-splice, SpliceAI max delta 0.004 so the consequence is a protein truncation rather than a splice-altering event); (2) loss of function is an established mechanism for MSH3 disease; (3) NMD is predicted, because the stop lies in exon 9 of 24, neither in the last exon nor within the last 50 bp of the penultimate exon 23 (c.3131-3302, last exon 24 from c.3303); and (4) the truncation removes far more than 10% of the protein and an established functional region, namely the residues 774-812 implicated in DNA recognition and the dimerization domain disrupted by exon 22 alterations (PMID:27476653). The NMD-predicted branch of the tree therefore returns PVS1 at full strength, so the strength value is 'very strong'. No named SVI or VCEP downgrade rule applies: the stop codon does not escape NMD by position, the affected exon is not a non-critical exon (the truncation abolishes the DNA-recognition and dimerization machinery while retaining only the N-terminal mismatch-binding region and its low-complexity poly-alanine/poly-proline stretch), and the change is not a canonical splice-site or non-critical-exon event. Accordingly no 'strong', 'moderate' or 'supporting' tier is used, and no downgrade rationale is required. Caveats recorded in gaps_remaining: this specific allele (c.1360C>T / p.Arg454Ter) was not reported in any paper fetched for this case, so the NMD call is positional rather than assay-based; even under the NMD-escape branch the critical-region criterion would leave PVS1 at strong, so 'very strong' is not sensitive to that assumption; and because MSH3 disease is recessive and heterozygous truncating MSH3 alleles are population variants not established as causative alone (PMID:28528517), PVS1 documents the null effect of a single allele only.

Rule

svi_pvs1_decision_tree:null_variant|lof_mechanism_established|nmd_predicted(not_last_exon,not_last_50bp_penultimate)|removes_gt_10pct_protein_and_critical_region -> PVS1 very strong (no downgrade rule applied)

Evidence considered

• Normalization/mutalyzer: NM_002439.5:c.1360C>T -> NM_002439.5(NP_002430.3):p.(Arg454Ter); protein first position 453/454, last predicted position 454 versus last original position 1138; the reference MSH3 protein is 1137 residues with CDS c.1-3414.
• VariantValidator: NM_002439.5 is the MANE Select and RefSeq Select transcript for MSH3 (HGNC:7326, OMIM 600887, 5q14.1); the variant maps to exon 9 of 24 in NG_016607.2, NC_000005.9 and NC_000005.10.
• Transcript selector exon boundaries: exon 9 spans c.1341-1453, so c.1360 sits 20 nt into exon 9; the penultimate exon 23 spans c.3131-3302 and the terminal exon 24 spans c.3303-*953, placing the premature stop roughly 1.9 kb 5' of the last 50 bp of the penultimate exon, hence NMD is predicted.
• Protein length change: the codon 454 stop removes residues 454-1137 (about 684 residues, ~60% of the 1137-residue protein); the retained N-terminal portion contains only the mismatch-binding region and the low-complexity poly-alanine/poly-proline stretch visible in the normalized reference sequence, not the MMR catalytic or dimerization machinery.
• MSH3 loss-of-function mechanism and protein-level consequences of MSH3 truncating alleles from PMID:27476653: MSH3 spans 24 exons and 1,137 amino acids; the four confirmed patient alleles were all putative LoF; loss of exon 17 removes amino acids 774-812 involved in DNA recognition; loss of exon 22 alters the dimerization domain; patients showed complete loss of nuclear MSH3 by immunohistochemistry and EMAST in tumours.
• PMID:27476653 was additionally read in raw full text (publications/27476653.txt) beyond the pre-computed extraction to confirm the exon/domain statements and the recessive biallelic context, and to confirm the paper does not contain c.1360C>T.
• PMID:37402566 (output/pvs1_articles/PMID_37402566.md, abstract-level text only): five new unrelated patients with biallelic MSH3 germline variants and attenuated colorectal adenomatous polyposis; EMAST described as a hallmark of MSH3 deficiency; MSH3 confirmed as a polyposis-predisposing gene, corroborating biallelic LoF as the mechanism.
• PMID:28528517 raw full text: heterozygous clearly truncating MSH3 germline variants (including c.1035del p.Leu347*) have not been established as causative for Lynch syndrome/CRC, truncating MSH3 variants in ExAC are rare (allele frequencies at or below roughly 0.004%) with none homozygous in the control datasets they queried, and biallelic MSH3 defects are required for the polyposis phenotype; this constrains the recessive disease context but does not change the per-allele null effect assessed by PVS1.
• PMID:25741868 PVS1 null-variant guidance: null variants may be assumed to disrupt gene function, provided null is a known mechanism consistent with the established inheritance pattern of the disease, and provided caution is applied for last-exon or penultimate-exon (last 50 bp) truncations, alternate transcripts, and exons without previously described pathogenic variants.
• ClinVar record 644460, exact transcript match NM_002439.5(MSH3):c.1360C>T (p.Arg454Ter), 8 submissions (5 Pathogenic, 3 Likely pathogenic), no expert-panel submission; listed conditions include Familial adenomatous polyposis 4, hereditary cancer-predisposing syndrome and endometrial carcinoma. Consulted only for disease/gene context and to confirm that no VCEP-level MSH3 specification exists; the submitted classifications themselves were not used to infer PVS1.
• SpliceAI lookup (max delta 0.004; DS_AG 0.004, DS_AL 0.004, DS_DG 0.004, DS_DL 0.002; pangolin SG 0.003, SL -0.06; no canonical splice consensus at this position): no predicted splice effect, so the variant's functional consequence is a protein truncation rather than a splice-altering event.
• OncoKB curated biological effect for MSH3 R454* is 'Likely Loss-of-function' (variant-specific), used as supporting context for the LoF mechanism; the accompanying oncogenicity label was not used to infer PVS1.
• gnomAD v2.1 (7/251,176 alleles, AF 2.79e-05, 0 homozygotes) and v4.1 (20/1,613,002 alleles, AF 1.24e-05, 0 homozygotes): no homozygous carrier reported, which is consistent with the recessive LoF model; these frequencies play no role in the PVS1 determination.
• Framework precedence: cspec.found=false, gene_vcep_dir=null, and vcep_materials.final_classification_framework shows found=false with source=generic_acmg_fallback and framework_complete=false, so the generic ACMG/AMP 2015 framework and the ClinGen SVI PVS1 decision tree are the governing specification for this criterion.
• Two of the sources listed in the gene-context scaffold concern cohorts or cases in which MSH3 is incidental (a Hungarian FAP cohort study and a CYLD cutaneous-syndrome case report); they were reviewed and not relied upon for the loss-of-function mechanism.

Still missing

• No MSH3-specific ClinGen CSPEC or VCEP criteria specification exists, so the PVS1 call rests entirely on the generic ClinGen SVI null-variant decision tree (PMC6185798).
• No RNA or protein assay exists for this specific allele; the NMD prediction is positional (exon 9 of 24 exons, remote from the penultimate exon junction), and degradation of the p.Arg454Ter product was not measured. PMID:27476653 did detect shortened MSH3 proteins by western blot for other truncating alleles, so a stable but non-functional truncated fragment cannot be formally excluded.
• This variant (c.1360C>T / p.Arg454Ter) is not reported in any of the papers fetched for this case, so no patient-level, functional or segregation evidence specific to this allele was available (variant_mentioned=false in every MSH3 paper reviewed).
• Heterozygous truncating MSH3 alleles occur in population datasets and MSH3-associated disease is recessive (PMID:28528517); PVS1 addresses only the functional consequence of this allele, and no phase or second-allele information for the proband is available in this case.

Sources

Reasoning

PM2 is met at Supporting strength. PM2 applies when the variant is absent or at extremely low frequency in population controls; under the ClinGen SVI recommendation PM2 is downgraded to Supporting with a threshold of AF <= 0.0001 (PMID:25741868). The observed all-comers allele frequencies are below that threshold in both genome-wide datasets: gnomAD v2.1 AF 2.78689e-05 (7/251,176 alleles, 0 homozygotes) and gnomAD v4.1 AF 1.23992e-05 (20/1,613,002 alleles, 0 homozygotes). There is no MSH3 VCEP specifying a non-cancer or exome-only source, so the all-comers totals remain the default; nevertheless the non-cancer subsets are concordant and even rarer (v2.1 non-cancer exomes 2.53545e-05, 6/236,644; v3.1 non-cancer genomes 1.35868e-05, 2/147,202), and gnomAD-Canada v1.0 is negative. Threshold/ancestry artifact: the gnomAD v2.1 grpmax FAF of 0.000106360 marginally exceeds the 0.0001 cutoff, and this is driven entirely by the small East Asian Korean sub-cohort (5/3,796 = 0.13%); the corresponding gnomAD v4.1 grpmax FAF is 5.796e-05, below threshold, and the total frequency is at or below 2.8e-05 in every dataset. Because the marginal exceedance is a small-subpopulation sampling artifact rather than a genuine higher-frequency signal, and because no homozygotes are observed, PM2 is judged met at Supporting. Supporting context: other truncating MSH3 variants in ExAC are reported at AF no more than 0.00004 (PMID:28528517), consistent with extreme rarity of this variant class. Flagged for human review solely because of the borderline v2.1 grpmax FAF.

Rule

PM2 Supporting: absent or extremely low frequency in controls, AF <= 0.0001 (PMID:25741868, SVI downgrade) — observed total AF 0.0000279 (v2.1) / 0.0000124 (v4.1), met

Evidence considered

• PM2 Supporting threshold applied: AF <= 0.0001 (ClinGen SVI 2020 recommendation to downgrade PM2 to Supporting; PMID:25741868).
• gnomAD v2.1 all-comers: AF 2.78689e-05 (7/251,176 alleles), 0 homozygotes; highest ancestry East Asian AF 0.000272183 (5/18,370); grpmax FAF 0.00010636.
• gnomAD v4.1 all-comers: AF 1.23992e-05 (20/1,613,002 alleles), 0 homozygotes; highest ancestry East Asian AF 0.000133917 (6/44,804); grpmax FAF 5.796e-05.
• gnomAD v2.1 non-cancer exomes: AF 2.53545e-05 (6/236,644 alleles), 0 homozygotes; grpmax FAF 0.00011076. Concordant with and lower than the all-comers totals.
• gnomAD v3.1 non-cancer genomes: AF 1.35868e-05 (2/147,202 alleles), 0 homozygotes; grpmax FAF 8.05e-06. The nominal 1KG:MSL frequency of 0.595% rests on a single allele in 168 chromosomes and is a reference-panel sample-size artifact.
• Threshold/ancestry artifact: the gnomAD v2.1 grpmax FAF 0.00010636 exceeds the 0.0001 threshold only through the Korean sub-cohort (5/3,796 = 0.13%); gnomAD v4.1 grpmax FAF is 5.796e-05 and total AF is <= 2.8e-05 in every dataset, so the marginal exceedance is treated as a small-subpopulation artifact.
• No homozygotes in any dataset; the expected homozygote count at these frequencies is <0.001, so the absence of homozygotes is uninformative for benignity and does not offset PM2.
• The variant is absent from gnomAD-Canada v1.0 (HostSeq genomes).
• Other truncating MSH3 variants in the ExAC browser are reported at an allele frequency of no more than 0.00004 (PMID:28528517), consistent with extreme rarity of this variant class.
• No ClinGen CSPEC or MSH3 VCEP exists (generic ACMG/AMP framework), so no gene-specific population source or threshold overrides the all-comers gnomAD defaults.

Still missing

• gnomAD v2.1 grpmax FAF (0.00010636) marginally exceeds the 0.0001 PM2 threshold because of the small East Asian Korean sub-cohort (5/3,796); gnomAD v4.1 grpmax FAF (5.796e-05) is below threshold. Manual confirmation of the ancestry handling is recommended.
• No ClinGen CSPEC or VCEP exists for MSH3; PM2 was applied with the generic ClinGen SVI Supporting-strength threshold (PMID:25741868) rather than a gene-specific one.
• The case bundle reports gnomAD non-cancer subsets that have no matching source_registry keys, so those values are recorded in evidence_used prose rather than as citations.

Sources

Reasoning

PS3 requires a well-established functional assay (validated for the gene/mechanism, with controls) demonstrating that the variant itself has a damaging effect on protein or gene function. Variant identity is NM_002439.5:c.1360C>T, NP_002430.3:p.(Arg454Ter) - a nonsense (stop-gained) change at MSH3 codon 454 (ClinVar 644460). No functional assay of this specific allele was found in any source examined. The MSH3 functional literature retrieved for this case characterises other alleles and/or MSH3 loss of function at gene level, not codon 454: (i) biallelic MSH3 LoF alleles c.1148delA, c.2319-1G>A, c.2760delC, c.3001-2A>C were validated by RT-PCR transcript analysis, site-directed mutagenesis in HEK293T cells with western blot, MSH3 immunohistochemistry and EMAST/MSI tumour testing in unrelated adenomatous polyposis patients; (ii) MSH3 germline variants reported in MSH6-truncating Lynch-syndrome families were c.1035del p.(Leu347*), c.2732T>G p.(Leu911Trp), c.1148del p.(Lys383Argfs*32), c.693G>A p.(=) and c.2305del p.(Val769*), none at codon 454; (iii) mouse Msh3-null cells show defective insertion/deletion mismatch repair and Msh3/Msh6 cooperate in intestinal tumour suppression, a gene-level mechanism finding; (iv) in five further MSH3-associated polyposis patients, EMAST phenotype (elevated microsatellite alterations at selected tetranucleotide repeats) confirmed MSH3 deficiency, and a negative EMAST result excluded germline MSH3 deficiency in two patients - establishing the assay and its controls at gene level but with no result for p.Arg454Ter. The OncoKB record for this variant carries a curated biological effect of 'Likely Loss-of-function', but this is an annotation/curation inference from variant type and gene-level biology, not the output of a functional assay, and cannot satisfy PS3. The protein-truncating consequence of p.Arg454Ter is itself the appropriate input to PVS1 (null-variant criterion), not to PS3, which is reserved for assay-based functional evidence. PS3 is therefore not met on the available evidence; the appropriate MSH3 assays (mismatch-repair activity, MutSbeta/MSH3 protein expression or heterodimer formation, or tumour EMAST/MSI in a carrier) have not been performed on this variant.

Evidence considered

• Variant identity: NM_002439.5:c.1360C>T = NP_002430.3:p.(Arg454Ter), a nonsense (stop-gained) change at MSH3 codon 454; ClinVar record NM_002439.5(MSH3):c.1360C>T (p.Arg454Ter), variation ID 644460, classified Pathogenic (5 laboratories) / Likely pathogenic (3 laboratories).
• PS3 definition applied: a variant-specific functional assay, validated for the gene and assayed with appropriate controls, demonstrating a damaging effect on protein/gene function; the criterion is assay-based and is therefore distinct from PVS1, which scores the expected null consequence of a stop-gained change.
• No governing VCEP or CSPEC specification exists for MSH3 in this case (CSPEC lookup found = false; framework resolution is generic ACMG/AMP 2015 fallback, framework found = false, gene_vcep_dir = null), so the generic ACMG/AMP PS3 requirement was applied and no gene-specific PS3 strength calibration was available to substitute.
• PMID:27476653 (full text; functional theme) assayed MSH3 loss-of-function alleles c.1148delA, c.2319-1G>A, c.2760delC and c.3001-2A>C by RT-PCR transcript analysis, site-directed mutagenesis in HEK293T cells with western blot, MSH3 immunohistochemistry in normal and tumour tissue and EMAST/microsatellite analysis; the paper states it identified two unrelated individuals with compound-heterozygous MSH3 loss-of-function germline mutations and does not contain c.1360C>T or any variant at codon 454 (searched terms included c.1360C>T, p.(R454*), p.Arg454Ter).
• PMID:28528517 (full text; functional theme) screened MSH3 exons 1-24 in 11 unrelated Lynch-syndrome patients with truncating MSH6 variants and reports MSH3 germline variants c.1035del p.(Leu347*), c.2732T>G p.(Leu911Trp), c.1148del p.(Lys383Argfs*32), c.693G>A p.(=) and c.2305del p.(Val769*); it contains no codon 454 variant, and its position is that heterozygous clearly truncating germline MSH3 variants have not been reported to cause Lynch syndrome while biallelic truncating MSH3 variants have been reported in two adenomatous polyposis patients.
• PMID:10706084 (abstract) reports mouse and cell functional work: Msh3-null cells are defective in repair of insertion/deletion mismatches but can repair base-base mismatches, and combined Msh3/Msh6 loss reproduces the Msh2-null tumour phenotype - gene-level evidence that MSH3 loss impairs mismatch repair, with no data on any human codon 454 allele and no full text supplied for this case.
• PMID:37402566 (abstract) reports five further unrelated patients with MSH3-associated attenuated adenomatous polyposis in whom an EMAST (elevated microsatellite alterations at selected tetranucleotide repeats) phenotype confirmed MSH3 deficiency and a negative EMAST result ruled out germline MSH3 deficiency in two patients - demonstrating that MSH3-deficiency functional readout exists and discriminates, but providing no EMAST result for the c.1360C>T allele; no full text or extract was available for this PMID.
• OncoKB records for R454* a curated biological effect of 'Likely Loss-of-function' and oncogenicity label 'Likely Oncogenic'; these are database curation statements rather than results of a functional assay performed on this allele, so they were not counted as PS3 evidence.
• PMID:25741868 (ACMG/AMP 2015 standards, full text) was consulted for the PS3 requirement itself and for its caution that null variants may be assumed to disrupt gene function through absence of product (lack of transcription or nonsense-mediated decay) - an inference expressly routed to the null-variant criterion, not to PS3.

Still missing

• No variant-level functional result for MSH3 p.Arg454Ter: a mismatch-repair (MutSbeta) activity assay, MSH3 protein expression/heterodimerisation assay, or tumour EMAST/MSI-IHC readout from a carrier of this allele would be required before PS3 could be met at any strength.
• The MSH3 full texts available for this case (PMID:27476653, PMID:28528517) neither perform nor cite an assay on the c.1360C>T allele; no other retrieved source reports one.
• OncoKB's 'Likely Loss-of-function' biological effect is a curator inference, not assay evidence, and was not used to support PS3.

Sources

Reasoning

No ClinGen CSPEC, VCEP specification, or local custom framework exists for MSH3 (vcep_materials.json reports no cspec url, framework_complete=false), so PS4 is adjudicated against the generic ACMG/AMP 2015 definition (PMID:25741868): the prevalence of the variant in affected individuals must be significantly increased compared with controls. Governing operator, quoted literally from that source for this criterion: "In general, variants with a modest Mendelian effect size will have an OR of 3 or greater, while highly penetrant variants will have very high ORs" - i.e. enrichment strengthens as the value rises and the comparison is OR >= 3. No odds ratio, relative risk, or case-versus-control comparison is reported for NM_002439.5:c.1360C>T (p.Arg454Ter) in any source, so the inequality cannot even be formed: OR = not available; "OR >= 3" is unevaluable and therefore not satisfied. What the sources do provide is rarity and gene-level biology, neither of which is case enrichment: (a) gnomAD v4.1 AF = 1.2399240670501338e-05 (20/1613002 alleles, 0 homozygotes) and gnomAD v2.1 AF = 2.78689e-05 (7/251176 alleles, 0 homozygotes); (b) ClinVar variation 644460, where all 8 submitters are single clinical laboratories (expert_panel_submissions = 0) reporting Pathogenic (5 labs) / Likely pathogenic (3 labs) with no published enrichment statistic; (c) the MSH3 literature, in which this exact variant appears in no publication at all - publications.json computed _variant_mentioned_count = 0 across all 12 triaged papers. MSH3-associated adenomatous polyposis is reported as a biallelic, autosomal-recessive phenotype in small case series (PMID:27476653 two unrelated compound-heterozygous index individuals; PMID:37402566 five new unrelated patients), not as an enrichment signal for this heterozygous variant. Because no enrichment statistic exists, PS4 is not met.

Evidence considered

• Governing framework: no MSH3 CSPEC/VCEP or local gene framework was found (vcep_materials.json: 'no cspec url available', framework_complete=false), so generic ACMG/AMP 2015 definitions (PMID:25741868) apply.
• Operator quoted literally from the governing generic source for PS4: 'In general, variants with a modest Mendelian effect size will have an OR of 3 or greater, while highly penetrant variants will have very high ORs' - enrichment strengthens upward, met requires OR >= 3.
• Explicit comparison attempt: OR = not available for NM_002439.5:c.1360C>T; the inequality 'OR >= 3' cannot be written with a substituted value, so the criterion cannot be satisfied.
• gnomAD v4.1 (exome): AF = 1.2399240670501338e-05, AC = 20 / AN = 1613002, 0 homozygotes; popmax East Asian AF = 0.00013391661458798322, grpmax FAF = 5.796e-05. gnomAD v2.1 (exome): AF = 2.78689e-05, 7/251176 alleles, 0 homozygotes, popmax East Asian AF = 0.000272183. These values quantify rarity, not case enrichment; no case/control comparison is derivable from them.
• ClinVar variation 644460 (NM_002439.5(MSH3):c.1360C>T p.Arg454Ter): 8 submissions, expert_panel_submissions = 0, all trust_tier 'single_submitter'; classifications Pathogenic (Myriad Genetics, MGZ Medical Genetics Center, Ambry Genetics, Sinai Health System, Labcorp/Invitae) and Likely pathogenic (Baylor Genetics, Fulgent Genetics, University of Chicago). Submitter evidence signals contain only PS3 and PM2 leads; no submission reports a case-control statistic. Condition labels are empty except one 'not provided'.
• Literature triage: publications.json _variant_mentioned_count = 0 - none of the 12 triaged publications contains the terms for this variant or protein change. PMID:27476653 and PMID:37402566 describe MSH3 loss-of-function as an autosomal-recessive adenomatous polyposis mechanism in small case series (2 unrelated compound-heterozygous individuals; 5 new unrelated patients respectively); PMID:28528517 states that no heterozygous clearly truncating germline MSH3 variant has been reported in Lynch syndrome/colorectal cancer patients and that heterozygous MSH3 defects alone do not appear to induce a phenotype. None of these constitutes affected-versus-control enrichment for this variant.
• MSH3-associated disease in the gene-level context is recessive (biallelic), whereas this variant is reported heterozygous in gnomAD and in the germline clinical submissions; an enrichment argument for a heterozygous carrier state is not supported by any retrieved source.

Still missing

• No case-control or cohort study reports the frequency of c.1360C>T in affected individuals versus controls; no odds ratio or relative risk is available for this variant, so the generic PS4 requirement (OR >= 3 with statistical significance) could not be tested.
• The variant's carrier frequency among MSH3-associated polyposis cases is unknown; the case series retrieved (PMID:27476653, PMID:37402566) describe only small numbers of unrelated biallelic patients and neither names this variant in the available text.

Sources

Reasoning

PP5 (PMID:25741868) applies when a reputable source reports the variant as pathogenic. The governing rule for this case restricts PP5 to an exact-variant ClinVar expert-panel classification: Pathogenic/Likely pathogenic by an expert panel supports PP5, while ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions must never trigger it. ClinVar variation 644460 (NM_002439.5(MSH3):c.1360C>T, p.Arg454Ter) has expert_panel_submissions = 0; all 8 submissions are single clinical laboratories (Myriad Genetics, Baylor Genetics, Fulgent Genetics, MGZ Medical Genetics Center, Ambry Genetics, Sinai Health System, University of Chicago, Labcorp/Invitae, none flagged is_expert_panel), each with review status 'criteria provided, single submitter'. The overall ClinVar record review status is likewise 'criteria provided, single submitter'. The aggregate label of Pathogenic (5 laboratories) and Likely pathogenic (3 laboratories) is exactly the non-expert, aggregated signal this rule forbids from triggering PP5, and no publication in the triage set names the variant to substitute for it (_variant_mentioned_count = 0). No qualifying expert-panel assertion exists, so PP5 is not met.

Evidence considered

• PP5 definition (PMID:25741868): a reputable source reports the variant as pathogenic. Governing case rule: only an exact-variant ClinVar expert-panel Pathogenic/Likely pathogenic classification may trigger PP5; laboratory-only submissions and aggregate ClinVar labels must not.
• ClinVar variation 644460 audit (clinvar_submission_audit.json): exact_match = true (matched by transcript HGVS), submissions_total = 8, expert_panel_submissions = 0, every submission is_expert_panel = false and is_clinical_testing = true, trust_tier = single_submitter, review_status = 'criteria provided, single submitter'.
• Submitter list: SCV004189052 Myriad Genetics (Pathogenic), SCV004197543 Baylor Genetics (Likely pathogenic), SCV005666180 Fulgent Genetics (Likely pathogenic), SCV002580370 MGZ Medical Genetics Center (Pathogenic), SCV001171469 Ambry Genetics (Pathogenic), SCV006056512 Sinai Health System (Pathogenic), SCV006325324 University of Chicago (Likely pathogenic), SCV000937995 Labcorp/Invitae (Pathogenic).
• ClinVar record-level review status: 'criteria provided, single submitter'; aggregate label Pathogenic (5 clinical laboratories) and Likely pathogenic (3 clinical laboratories) - an aggregate, non-expert signal explicitly excluded from triggering PP5.
• No independent reputable source assertion for this exact variant: publications.json _variant_mentioned_count = 0 across 12 triaged papers, so the laboratory-only ClinVar submissions are the entirety of the 'reported pathogenic' evidence.

Still missing

• No ClinVar expert-panel (VCEP/ClinGen) submission exists for this variant; if one is deposited later, PP5 would need re-evaluation at supporting strength.

Sources

Reasoning

BA1 requires a stand-alone benign allele frequency of at least 0.05 (Richards et al. 2015, PMID:25741868). This nonsense MSH3 variant is extremely rare in every gnomAD dataset: gnomAD v2.1 all-comers AF 2.78689e-05 (7/251,176 alleles, 0 homozygotes) and gnomAD v4.1 AF 1.23992e-05 (20/1,613,002 alleles, 0 homozygotes). The non-cancer subsets are concordant (v2.1 non-cancer exomes 2.53545e-05, 6/236,644; v3.1 non-cancer genomes 1.35868e-05, 2/147,202). No VCEP specification exists for MSH3, so the generic all-comers totals are the default population source, but the non-cancer subsets give the same picture. The highest ancestry-specific value anywhere is East Asian/Korean: 0.000282965 in gnomAD v2.1 non-cancer EAS (5/17,670) and 0.00131718 in the v2.1 EAS_KOR sub-cohort (5/3,796), i.e. roughly 0.13%; even that extreme value is about 380-fold below the 0.05 stand-alone threshold. The variant is also absent from gnomAD-Canada v1.0. A frequency this low cannot be stand-alone benign evidence under any ancestry or dataset choice.

Rule

BA1 stand-alone benign: AF >= 0.05 (PMID:25741868) — observed max AF 0.00132, threshold not reached

Evidence considered

• gnomAD v2.1 all-comers: AF 2.78689e-05 (7/251,176 alleles), 0 homozygotes; highest ancestry East Asian AF 0.000272183 (5/18,370); grpmax FAF 0.00010636.
• gnomAD v4.1 all-comers: AF 1.23992e-05 (20/1,613,002 alleles), 0 homozygotes; highest ancestry East Asian AF 0.000133917 (6/44,804); grpmax FAF 5.796e-05.
• gnomAD v2.1 non-cancer exomes: AF 2.53545e-05 (6/236,644 alleles), 0 homozygotes; grpmax FAF 0.00011076.
• gnomAD v3.1 non-cancer genomes: AF 1.35868e-05 (2/147,202 alleles), 0 homozygotes; grpmax FAF 8.05e-06.
• Highest ancestry-specific frequency observed anywhere is the gnomAD v2.1 EAS_KOR sub-cohort at 0.00131718 (5/3,796); v2.1 non-cancer EAS is 0.000282965 (5/17,670).
• BA1 threshold applied: stand-alone benign AF >= 0.05 (Richards et al. 2015, PMID:25741868).
• The observed maximum allele frequency is approximately 380-fold below the 0.05 threshold.
• The variant is absent from gnomAD-Canada v1.0 (HostSeq genomes).
• No MSH3 CSPEC/VCEP is available; the generic ACMG/AMP framework and the all-comers gnomAD totals were used as defaults, with the non-cancer subsets as corroboration.

Still missing

• No ClinGen CSPEC or gene-specific VCEP exists for MSH3; generic ACMG/AMP 2015 thresholds were used.
• No MSH3-specific benign allele-frequency threshold has been published, so the generic 0.05 stand-alone value is applied.

Sources

Reasoning

BS1 requires an allele frequency greater than expected for the disorder. No MSH3-specific per-gene derivation exists (no ClinGen CSPEC, no VCEP), so the generic strong threshold of AF >= 0.01 derived from Whiffin et al. 2017 (PMID:28518168) is applied. Observed frequencies are gnomAD v2.1 2.78689e-05 (7/251,176), gnomAD v4.1 1.23992e-05 (20/1,613,002), v2.1 non-cancer exomes 2.53545e-05 (6/236,644) and v3.1 non-cancer genomes 1.35868e-05 (2/147,202) — roughly 360-fold below the 0.01 threshold. MSH3 germline loss-of-function causes autosomal-recessive familial adenomatous polyposis 4 (PMID:27476653), a very rare, attenuated condition; a recessive model would permit a higher pathogenic carrier frequency than a dominant one, so the strictly appropriate comparator would be the square root of disease prevalence rather than the prevalence itself. However, no FAP4 prevalence figure is available in the case materials, and the observed allele frequency is orders of magnitude below 0.01 in all datasets and all ancestries (maximum 0.13%, EAS_KOR), so BS1 is not met regardless of the prevalence value chosen. This criterion was considered at strong strength; no lower strength is met either.

Rule

BS1 strong benign: AF >= 0.01 (PMID:28518168) — observed AF 0.0000279, threshold not reached

Evidence considered

• gnomAD v2.1 all-comers AF 2.78689e-05 (7/251,176 alleles, 0 homozygotes); gnomAD v4.1 AF 1.23992e-05 (20/1,613,002 alleles, 0 homozygotes).
• gnomAD v2.1 non-cancer exomes AF 2.53545e-05 (6/236,644 alleles); gnomAD v3.1 non-cancer genomes AF 1.35868e-05 (2/147,202 alleles) — concordant with the all-comers totals.
• Generic BS1 strong threshold applied: AF >= 0.01 (Whiffin et al. 2017, PMID:28518168); no MSH3-specific derivation and no VCEP exists to supply one.
• Observed allele frequency is approximately 1/360 of the 0.01 threshold; the highest ancestry-specific value (0.13%, gnomAD v2.1 EAS_KOR) is still more than 7-fold below threshold.
• MSH3 germline loss-of-function causes autosomal-recessive, attenuated familial adenomatous polyposis 4 (PMID:27476653), so the expected pathogenic allele frequency is bounded by disease prevalence; the observed frequency is nonetheless far below the threshold under any plausible prevalence.
• The variant is absent from gnomAD-Canada v1.0 (HostSeq genomes).

Still missing

• No per-gene BS1 allele-frequency derivation for MSH3 exists; the generic 0.01 strong threshold was used.
• No disease-prevalence estimate for MSH3-associated familial adenomatous polyposis 4 is present in the case materials; a prevalence-based recessive comparator could not be derived.

Sources

Reasoning

BS2 requires the variant to be observed in a healthy adult in the genotype relevant to the disorder's inheritance model — homozygous (or compound heterozygous) for a recessive disorder — with full penetrance expected at an early age. MSH3 germline loss-of-function is an established autosomal-recessive cause of familial adenomatous polyposis 4, defined by biallelic/compound-heterozygous truncating variants (PMID:27476653), with MSH3-associated polyposis described as attenuated. Heterozygous carriers are therefore the expected genotype for a recessive allele and do not constitute BS2 evidence. In this case there is not a single homozygous observation: gnomAD v2.1 reports 7/251,176 alleles with 0 homozygotes, gnomAD v4.1 20/1,613,002 with 0 homozygotes, v2.1 non-cancer exomes 6/236,644 with 0 homozygotes, and v3.1 non-cancer genomes 2/147,202 with 0 homozygotes. The variant is also absent from gnomAD-Canada v1.0. With an allele frequency of about 2.8e-05, the expected number of homozygotes in gnomAD v2.1 is on the order of 2e-7, so the absence of homozygotes is statistically uninformative rather than supportive of benignity. Additionally, MSH3-associated polyposis does not meet BS2's premise of full penetrance at an early age. No BS2-qualifying observation exists in any dataset.

Rule

BS2 strong benign: healthy adult in disorder-relevant genotype (homozygous for recessive) with full early penetrance — 0 homozygotes observed, criterion not met

Evidence considered

• BS2 requires observation in a healthy adult in the disorder-relevant genotype (homozygous or compound heterozygous for a recessive disorder) with full penetrance expected at an early age.
• MSH3 germline loss-of-function causes autosomal-recessive familial adenomatous polyposis 4, established in compound-heterozygous patients (PMID:27476653); heterozygous carriers are the expected genotype and are not BS2 evidence.
• Zero homozygotes in every gnomAD dataset: v2.1 7/251,176; v4.1 20/1,613,002; v2.1 non-cancer exomes 6/236,644; v3.1 non-cancer genomes 2/147,202.
• Expected homozygote count at AF 2.79e-05 in gnomAD v2.1 is approximately 2e-7, so the absence of homozygotes is uninformative and provides no benign support.
• MSH3-associated polyposis is described as an attenuated phenotype, so the 'full penetrance expected at an early age' premise of BS2 is not satisfied.
• The variant is absent from gnomAD-Canada v1.0 (HostSeq genomes).

Still missing

• No homozygous or compound-heterozygous adult observation is available in any population dataset; gnomAD homozygote counts are zero, so a BS2-eligible observation cannot be evaluated.
• No ClinGen CSPEC or VCEP exists for MSH3; generic ACMG/AMP 2015 criteria were used.

Sources

Reasoning

BS3 requires a well-established functional assay showing that the variant has no damaging effect on protein/gene function (normal or near-normal activity, with assay controls). No functional assay result of any kind exists for NM_002439.5:c.1360C>T, NP_002430.3:p.(Arg454Ter); the allele is a nonsense (stop-gained) change at MSH3 codon 454 (ClinVar 644460) that removes the C-terminal portion of the MSH3 protein, and no source examined reports preserved MSH3 expression, preserved MutSbeta mismatch-repair activity, absence of EMAST/MSI, or any other normal-function readout for this allele. The functional MSH3 studies retrieved for this case concern other alleles (biallelic LoF alleles c.1148delA, c.2319-1G>A, c.2760delC, c.3001-2A>C in adenomatous polyposis patients; c.1035del, c.2732T>G, c.1148del, c.693G>A, c.2305del in MSH6-truncating families) and gene-level mechanism (mouse Msh3-null cells defective in insertion/deletion mismatch repair; EMAST readouts confirming MSH3 deficiency in patients and excluding it in two others). Those results describe loss of MSH3 function, not preservation of it, and none applies to p.Arg454Ter. Accordingly BS3 is not met: there is no functional evidence that this variant is benign, and the criterion is not satisfied by the absence of data. As no MSH3-specific VCEP/CSPEC exists for this case, the generic ACMG/AMP BS3 requirement was applied.

Evidence considered

• Variant identity: NM_002439.5:c.1360C>T = NP_002430.3:p.(Arg454Ter), nonsense (stop-gained) at MSH3 codon 454 (ClinVar 644460); no assay in any retrieved source was performed on this allele.
• BS3 definition applied: a validated functional assay demonstrating no damaging effect on protein/gene function; evidence of normal function is absent here rather than present, so the criterion is not met.
• PMID:27476653 (full text) reported functional consequences (exon skipping, truncated protein on western blot, complete loss of nuclear MSH3 on immunohistochemistry, EMAST-high tumours) for other MSH3 alleles only; searched for c.1360C>T, p.(R454*), p.Arg454Ter with no match. These results demonstrate loss of function, the opposite of the BS3 direction.
• PMID:28528517 (full text) reports other MSH3 variants (c.1035del, c.2732T>G, c.1148del, c.693G>A, c.2305del) and states that no clearly truncating heterozygous germline MSH3 variant has been reported in Lynch-syndrome/colorectal-cancer patients; it contains no codon 454 variant and reports no normal-function assay for this allele.
• PMID:10706084 (abstract) shows Msh3-null mouse cells are defective in repair of insertion/deletion mismatches - gene-level loss-of-function data, not a normal-function readout for p.Arg454Ter.
• PMID:37402566 (abstract) used EMAST to confirm MSH3 deficiency in five patients and to exclude germline MSH3 deficiency in two patients; the assay and its ability to be negative are documented, but no EMAST or equivalent result is reported for this variant.
• OncoKB annotates this variant with a biological effect of 'Likely Loss-of-function'; the only variant-level biological annotation available therefore points away from preserved function, and nothing in the retrieved evidence supports normal MSH3 activity for p.Arg454Ter.
• PMID:25741868 (ACMG/AMP 2015 standards, full text) was consulted for the BS3 requirement; no MSH3-specific VCEP/CSPEC framework exists for this case (CSPEC found = false, generic ACMG/AMP 2015 fallback).

Still missing

• No normal-function functional readout (mismatch-repair activity, MSH3 protein expression/heterodimerisation, or negative tumour EMAST/MSI) has been reported for MSH3 p.Arg454Ter; such a result would be required for BS3 to be met.
• No MSH3-specific VCEP/CSPEC specification or gene-specific BS3 calibration exists for this case, so the generic ACMG/AMP BS3 definition applies.
• Full texts were unavailable for PMID:10706084 and PMID:37402566 (abstract-level evidence only); neither abstract reports this variant.

Sources

Reasoning

BP6 (PMID:25741868) applies when a reputable source reports the variant as benign. The governing rule for this case restricts BP6 to an exact-variant ClinVar expert-panel classification: a Benign/Likely benign expert-panel assertion supports BP6, while ordinary laboratory submissions, aggregate ClinVar labels, and non-expert assertions must never trigger it. ClinVar variation 644460 (NM_002439.5(MSH3):c.1360C>T, p.Arg454Ter) has expert_panel_submissions = 0, and every one of the 8 submissions is a single clinical laboratory (Myriad Genetics, Baylor Genetics, Fulgent Genetics, MGZ Medical Genetics Center, Ambry Genetics, Sinai Health System, University of Chicago, Labcorp/Invitae, none flagged is_expert_panel) with review status 'criteria provided, single submitter'. Every submission classifies the variant as Pathogenic or Likely pathogenic - none as Benign or Likely benign - and no expert-panel benign assertion exists in any retrieved source. The criterion's evidentiary prerequisite is therefore absent, and BP6 is not met.

Evidence considered

• BP6 definition (PMID:25741868): a reputable source reports the variant as benign. Governing case rule: only an exact-variant ClinVar expert-panel Benign/Likely benign classification may trigger BP6; laboratory-only submissions and aggregate ClinVar labels must not.
• ClinVar variation 644460 audit: expert_panel_submissions = 0; all 8 submissions is_expert_panel = false, trust_tier = single_submitter, review_status = 'criteria provided, single submitter'; classifications are Pathogenic (5 laboratories) and Likely pathogenic (3 laboratories) - no benign classification of any kind is present.
• No benign assertion from any other retrieved source: publications.json _variant_mentioned_count = 0 across the 12 triaged papers, and no database record (gnomAD, ClinVar, OncoKB, COSMIC) asserts the variant is benign.

Still missing

• No ClinVar expert-panel Benign/Likely benign submission exists for this variant; absent such an assertion BP6 has no evidentiary basis. Population rarity (gnomAD v4.1 AF 1.24e-05, 0 homozygotes) is likewise not BP6 evidence.

Sources

Reasoning

PS2 requires a de novo occurrence with BOTH maternity and paternity confirmed, in a patient with the disease and no family history (ACMG/AMP 2015, PMID:25741868). The case evidence contains no trio or parental testing, no statement of parental origin, and no proband phenotype assertion, so de novo status for NM_002439.5:c.1360C>T p.(Arg454Ter) can be neither confirmed nor excluded. The only case-level clinical source, ClinVar VCV000644460 (8 submissions, 2-star, criteria provided, no expert panel), records submission origin as germline for six submitters and unknown for two, with no parental genotypes, no phase information, and no statement of whether the variant is heterozygous or biallelic in the proband; ClinVar classification itself is not evidence for this criterion. Gene-level context also argues that a de novo route is not the expected mechanism: MSH3-associated adenomatous polyposis (FAP4) is an autosomal recessive condition in which biallelic disease-causing alleles are typically inherited from heterozygous carrier parents rather than arising de novo (PMID:27476653, PMID:37402566). No MSH3 CSPEC/VCEP specification exists, so generic ACMG/AMP 2015 semantics apply. Verdict: not assessed - the criterion cannot be evaluated without parental testing.

Rule

PS2 (ACMG/AMP 2015, PMID:25741868): De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. No MSH3-specific VCEP/CSPEC specification available (generic ACMG/AMP applied).

Evidence considered

• PS2 definition applied: de novo (both maternity and paternity confirmed) in a patient with the disease and no family history (ACMG/AMP 2015, PMID:25741868); no MSH3 CSPEC or VCEP specification exists (cspec found = false; vcep_materials reports no cspec url available), so the generic ACMG/AMP 2015 criterion is used.
• Case-level review of every evidence file present for this case (review_manifest, case_summary, source_registry, evidence, literature_pass, fulltext_map, paper_extracts, plus prefetch.json and vcep_materials.json read directly) found no proband record, no pedigree, no parental genotypes, and no de novo statement.
• ClinVar VCV000644460 (variation 644460) for NM_002439.5:c.1360C>T: Pathogenic/Likely pathogenic, 2-star (criteria provided, multiple submitters, no conflicts), 8 submissions from 8 submitters, 0 expert-panel submissions; submitter origin germline (6) or unknown (2); conditions listed include Familial adenomatous polyposis 4 and Hereditary cancer-predisposing syndrome. No parental testing, phase, or segregation fields are reported, and the classification itself is not used as evidence for this criterion.
• PMID:27476653 established MSH3 biallelic loss-of-function as an autosomal recessive cause of colorectal adenomatous polyposis; the affected individuals carried compound-heterozygous MSH3 LoF alleles (c.1148delA, c.2319-1G>A, c.2760delC, c.3001-2A>C) rather than de novo alleles, framing MSH3 disease as inherited from carrier parents.
• PMID:37402566 reports four patients from two families with biallelic MSH3 germline variants and attenuated colorectal adenomatous polyposis, again consistent with recessive inheritance rather than de novo occurrence.
• Variant identity confirmed: NM_002439.5:c.1360C>T, NP_002430.3:p.(Arg454Ter), nonsense in exon 9, MANE Select, NC_000005.10:g.80725472C>T (GRCh38).

Still missing

• No trio or parental testing (targeted parental genotyping of c.1360C>T) is present anywhere in the case evidence.
• No statement of parental origin, phase, or proband zygosity (heterozygous vs biallelic) for c.1360C>T.
• No proband phenotype/family-history assertion, so the 'patient with the disease and no family history' precondition cannot be verified.
• ClinVar VCV000644460 records germline origin but no parental genotypes or segregation data.

Sources

Reasoning

PM3 for a recessive disorder requires the variant to be detected in trans with a pathogenic variant (ACMG/AMP 2015, PMID:25741868). MSH3 germline loss of function is an established autosomal-recessive cause of adenomatous polyposis (familial adenomatous polyposis 4), so PM3 is the applicable allelic criterion in principle (PMID:27476653, PMID:37402566). No source in this case, however, provides any phase information for NM_002439.5:c.1360C>T p.(Arg454Ter): there is no proband record, no zygosity statement, and no genotype for a second MSH3 allele. ClinVar VCV000644460 (8 submissions from 8 laboratories, all single-submitter, 2 stars, no expert panel; conditions familial adenomatous polyposis 4, endometrial carcinoma, hereditary cancer-predisposing syndrome, not provided) reports submission origin as germline (6) or unknown (2) but contains no zygosity, phase, cis/trans or second-allele data; the only criterion-level signals recorded from those submissions are a functional-study caveat (PS3 lead) and population rarity (PM2 lead), never a pathogenic allele in trans. The variant is not reported at all in either full text retrieved for this case: PMID:27476653 instead describes compound-heterozygous MSH3 loss-of-function alleles (c.1148delA, c.2319-1G>A, c.2760delC, c.3001-2A>C) in two unrelated individuals, and PMID:28528517 reports the MSH3 variants c.1035del p.(Leu347*), c.2732T>G p.(Leu911Trp), c.1148del p.(Lys383Argfs*32), c.693G>A p.(=) and c.2305del p.(Val769*), with no codon-454 variant. PMID:37402566, the paper that confirms MSH3 as a polyposis gene, is available for this case only at abstract level and its abstract carries no variant-level detail. Population data likewise contain no biallelic observation: 0 homozygotes in gnomAD v2.1 (7/251,176 alleles) and gnomAD v4.1 (20/1,613,002 alleles). Verdict: not assessed - the criterion needs a proband-level trans observation that the evidence set does not contain, so PM3 can be neither satisfied nor excluded on the available data.

Rule

PM3 (ACMG/AMP 2015, PMID:25741868): For recessive disorders, detected in trans with a pathogenic variant. No MSH3-specific VCEP/CSPEC specification available (generic ACMG/AMP applied).

Evidence considered

• PM3 definition applied: for recessive disorders, detected in trans with a pathogenic variant (ACMG/AMP 2015, PMID:25741868). No MSH3 CSPEC/VCEP specification exists (cspec found = false; vcep_materials reports 'no cspec url available'; framework mode generic_acmg), so the generic criterion is used.
• Inheritance mode established from the retrieved literature: biallelic MSH3 loss-of-function germline variants cause autosomal-recessive colorectal adenomatous polyposis (PMID:27476653), restated for four patients from two families in PMID:37402566, and heterozygous clearly truncating MSH3 variants are not established as causative (PMID:28528517). MSH3 disease is therefore a recessive trait, which is the setting in which PM3 is the relevant allelic criterion.
• ClinVar VCV000644460 for NM_002439.5:c.1360C>T: Pathogenic/Likely pathogenic, 2 stars ('criteria provided, multiple submitters, no conflicts'), 8 submissions from 8 submitters, 0 expert-panel submissions; conditions familial adenomatous polyposis 4, endometrial carcinoma, hereditary cancer-predisposing syndrome and 'not provided'; submission origin germline (6) or unknown (2). No phase, zygosity, second-allele, or family-genotype information is reported in any submission, and the classifications themselves were not used as evidence for this criterion.
• Criterion-level signals extracted from the ClinVar submissions contain no trans/phase claim: one laboratory (Genetic Services Laboratory, University of Chicago) mentions only that functional studies have not been performed (PS3 lead) and one (Labcorp/Invitae) mentions only population rarity (PM2 lead); the remaining submissions carry only the ACMG/AMP guideline citation (PMID:25741868).
• PMID:27476653 (full text retrieved) provides the trans-allele precedent for MSH3 but not for this variant: the two unrelated affected individuals carried compound-heterozygous MSH3 loss-of-function alleles c.1148delA, c.2319-1G>A, c.2760delC and c.3001-2A>C, and c.1360C>T is not reported anywhere in the paper.
• PMID:28528517 (full text retrieved) screened MSH3 in Lynch-syndrome patients and lists the MSH3 variants c.1035del p.(Leu347*), c.2732T>G p.(Leu911Trp), c.1148del p.(Lys383Argfs*32), c.693G>A p.(=) and c.2305del p.(Val769*); no codon-454 variant is reported, so it supplies no trans observation for c.1360C>T.
• PMID:37402566 was retrievable only as an abstract for this case (no full text; the stored record is a Sci-Hub 'not available' page with similar-article listings). The abstract confirms biallelic MSH3 germline variants in patients from two families and MSH3's role in attenuated adenomatous polyposis, but gives no variant identities, so it neither supports nor excludes a trans configuration for c.1360C>T.
• Population data contain no biallelic genotype that could substitute for a proband observation: 0 homozygotes in gnomAD v2.1 (7/251,176 alleles, AF 2.78689e-05) and 0 homozygotes in gnomAD v4.1 (20/1,613,002 alleles, AF 1.23992e-05).
• Sources retrieved for this case that were checked and found to carry no MSH3 or phase content: PMID:24493721 and PMID:25394175 (family-history/predisposition guidelines that do not name MSH3), PMID:25741868 (criteria definitions only) and PMID:10706084 (mouse Msh3/Msh6 intestinal tumour-suppression study, abstract level only).
• Net assessment: the criterion's required observation - this variant in trans with a pathogenic MSH3 variant in an affected individual - is absent from every available source, and no source makes a claim either way about the phase of c.1360C>T; the criterion is therefore recorded as not assessed rather than evaluated and failed.

Still missing

• No proband record, zygosity, or phase information exists for this case, so the cis/trans configuration of c.1360C>T relative to a second MSH3 allele is unknown.
• ClinVar VCV000644460 reports submission origin (germline/unknown) only, with no second-allele, zygosity, or phase fields in any of its 8 submissions.
• No full text is available for PMID:37402566 (MSH3: a confirmed predisposing gene for adenomatous polyposis); only the abstract was retrievable, so any biallelic MSH3 genotypes it reports could not be checked for c.1360C>T.
• No ClinGen CSPEC or VCEP specification exists for MSH3, so generic PM3 semantics (trans with a pathogenic variant in a recessive gene) were applied with no gene-specific strength guidance.

Sources

Reasoning

PM6 requires that the variant be assumed de novo without confirmation of maternity and paternity - i.e. there must be at least a clinical assertion of apparent de novo occurrence, such as the variant being absent from tested parents (ACMG/AMP 2015, PMID:25741868). No such assertion exists anywhere in this case's evidence: there are no parental samples, no trio, and no proband record from which an apparent de novo occurrence could even be assumed. ClinVar VCV000644460 reports submitter origin as germline or unknown, which says nothing about whether c.1360C>T was inherited and therefore cannot substitute for a de novo assertion. Gene-level data point the other way: heterozygous clearly truncating MSH3 germline variants have not been reported in Lynch-syndrome/colorectal-cancer patients and heterozygous MSH3 defects alone do not appear to produce that phenotype, whereas biallelic truncating MSH3 variants have been reported in adenomatous polyposis (PMID:28528517, PMID:27476653) - so the expected origin of an MSH3 disease allele is inheritance from a carrier parent, not de novo occurrence. Verdict: not assessed - no parental-origin information exists to support even the weaker, unconfirmed de novo assumption.

Rule

PM6 (ACMG/AMP 2015, PMID:25741868): Assumed de novo, but without confirmation of maternity and paternity. No MSH3-specific VCEP/CSPEC specification available (generic ACMG/AMP applied).

Evidence considered

• PM6 definition applied: assumed de novo (no confirmation of maternity and paternity), used at supporting strength in ACMG/AMP 2015 (PMID:25741868); no MSH3 CSPEC/VCEP specification exists, so generic ACMG/AMP semantics apply.
• Searched the whole case evidence set (including prefetch.json and vcep_materials.json read directly) for any parental, trio, or de novo statement relating to c.1360C>T: none is present.
• ClinVar VCV000644460 lists 8 submissions with origin germline (6) or unknown (2) and no parental genotypes, phase, or de novo assertion; the ClinVar classification is not used as evidence for this criterion.
• PMID:28528517: the authors state that to their knowledge no heterozygous clearly truncating germline MSH3 variant has been reported in the literature for Lynch-syndrome/colorectal-cancer patients and that heterozygous MSH3 defects alone do not appear to induce that phenotype, while biallelic truncating MSH3 germline variants have been reported in two patients with adenomatous polyposis - i.e. MSH3 disease alleles are expected to be inherited, not de novo.
• PMID:27476653: the MSH3 variants underlying recessive adenomatous polyposis were compound-heterozygous LoF alleles in two unrelated individuals, consistent with inheritance from carrier parents.

Still missing

• No parental testing or trio data, and no clinical statement that c.1360C>T is absent from the parents, so the 'assumed de novo' premise cannot be established.
• Proband zygosity and phase (inherited vs de novo, maternal vs paternal) for c.1360C>T are not documented.
• ClinVar VCV000644460 origin fields (germline/unknown) carry no parental-origin information.

Sources

Reasoning

PP1 requires co-segregation of the variant with disease in multiple affected family members, evaluated across meioses in the patient's own family (ACMG/AMP 2015, PMID:25741868). Two conditions are therefore needed: a pedigree with affected relatives, and genotypes for c.1360C>T in those relatives. Neither exists in this case: there is no pedigree, no family history, no affected relative genotypes, and no meiosis count anywhere in the case evidence, and ClinVar VCV000644460 provides no segregation data. Literature retrieved for this case reports MSH3 families (PMID:27476653: two unrelated individuals with compound-heterozygous biallelic MSH3 LoF and adenomatous polyposis; PMID:37402566: four patients from two families with biallelic MSH3 variants) but never the c.1360C>T variant and never this patient's family, so those published families cannot supply co-segregation evidence for this variant; family-history-focused sources retrieved for this case (PMID:24493721, PMID:25394175) contain no MSH3 content at all. Verdict: not assessed - the criterion is unassessable because no family genotypes exist for this variant.

Rule

PP1 (ACMG/AMP 2015, PMID:25741868): Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease. No MSH3-specific VCEP/CSPEC specification available (generic ACMG/AMP applied).

Evidence considered

• PP1 definition applied: co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease (ACMG/AMP 2015, PMID:25741868); no MSH3 CSPEC/VCEP specification exists, so no gene-specific meiosis threshold is available and none was applied.
• Searched all case evidence files for pedigree, proband, family-history, parental, segregation and meiosis content: the only occurrences of these terms are papers' titles/notes, none relating to this case or to c.1360C>T.
• ClinVar VCV000644460 (8 submissions, 2-star, no expert panel) reports no family members, no segregation data, and no co-segregation counts; its classification was not used as evidence.
• PMID:27476653: two unrelated individuals with compound-heterozygous MSH3 loss-of-function variants (c.1148delA, c.2319-1G>A, c.2760delC, c.3001-2A>C) causing recessive colorectal adenomatous polyposis; c.1360C>T is not reported, so no co-segregation data exist for this variant.
• PMID:37402566: four patients from two families with biallelic MSH3 germline variants and attenuated colorectal adenomatous polyposis (abstract only; no full text retrieved for this PMID); provides gene-level recessive inheritance context, not variant-level segregation.
• PMID:24493721 and PMID:25394175 were retrieved as family-history/pedigree-related sources but contain no MSH3 content and no segregation data (see paper_findings).

Still missing

• No pedigree or family history for the proband is provided in any case file.
• No genotypes for c.1360C>T in affected or unaffected relatives, so no meiosis can be counted toward co-segregation.
• ClinVar VCV000644460 submissions report no segregation information.
• Literature MSH3 families (PMID:27476653, PMID:37402566) do not include this variant or this family and cannot substitute for in-family co-segregation.

Sources

Reasoning

PP4 (PMID:25741868) requires that the patient's phenotype or family history be highly specific for a disease with a single genetic etiology. This case is a variant-level request: the case evidence contains no proband phenotype, no HPO terms, no family history, and no referral indication (case_summary.json, evidence_brief.json and prefetch.json contain no phenotype/pedigree field), and the ClinVar submissions for variation 644460 carry no condition labels (all empty except one 'not provided'). Phenotype specificity therefore cannot be evaluated for the individual being tested, which is a data gap rather than a negative finding on the variant. Gene-level specificity does exist: MSH3 biallelic loss of function is associated with an autosomal-recessive attenuated colorectal adenomatous polyposis with elevated microsatellite alterations at selected tetranucleotide repeats (EMAST) (PMID:27476653, PMID:37402566), and heterozygous MSH3 changes have minimal evidence for a Lynch-syndrome phenotype (PMID:28528517). However, this exact variant is not named in any of the 12 triaged publications (_variant_mentioned_count = 0), and no retrieved source matches this variant to a proband phenotype, so no variant-level PP4 evidence exists either. Returned as not_assessed and flagged rather than scored.

Evidence considered

• PP4 definition (PMID:25741868): the patient's phenotype or family history must be highly specific for a disease with a single genetic etiology.
• Case-level verification for a proband phenotype: case_summary.json (compact_evidence keys: variant_input, extracted, normalization, cspec, clinvar, gnomad, spliceai, revel, bayesdel, hci_prior, oncokb, cosmic, hotspots) and evidence_brief.json contain no phenotype, HPO, pedigree, or clinical-history field; prefetch.json likewise; hci_prior 'found' = false.
• ClinVar variation 644460 condition labelling: all 8 submissions have empty condition_labels except SCV006325324 ('not provided'), so no phenotype is recoverable from ClinVar either.
• Gene-level phenotype specificity (context only, not applied): MSH3 biallelic loss of function causes autosomal-recessive attenuated colorectal adenomatous polyposis with EMAST (PMID:27476653, PMID:37402566); heterozygous truncating MSH3 variants have not been established as causative for Lynch syndrome (PMID:28528517).
• Variant-level phenotype match: publications.json _variant_mentioned_count = 0 across 12 triaged papers; no publication links c.1360C>T to a described phenotype in a tested individual.

Still missing

• No proband clinical phenotype, HPO term list, age at diagnosis, or family history is present anywhere in the case evidence; PP4 cannot be scored without the patient's own phenotypic data.
• The exact variant c.1360C>T (p.Arg454Ter) is not reported in any of the 12 triaged publications, so even a gene-level phenotype match cannot be tied to this variant.

Sources

Reasoning

BS4 requires positive documentation of non-segregation: the variant is shown to be absent from affected members of a family, or affected relatives are demonstrated not to carry it (ACMG/AMP 2015, PMID:25741868). It is a benign criterion, so it can only be met by actual observation of non-segregation in a genotyped family. No pedigree, no affected relatives, and no family genotypes exist anywhere in this case's evidence, and ClinVar VCV000644460 reports no segregation information; therefore non-segregation for NM_002439.5:c.1360C>T has neither been observed nor can it be inferred. Absence of segregation data is not evidence of absence of segregation, so BS4 is not supported; because the required family data are wholly missing rather than informative, the criterion is recorded as not assessed rather than as an evaluated-and-failed criterion. Verdict: not assessed - no segregation study exists in which non-segregation could have been documented.

Rule

BS4 (ACMG/AMP 2015, PMID:25741868): Lack of segregation in affected members of a family. No MSH3-specific VCEP/CSPEC specification available (generic ACMG/AMP applied).

Evidence considered

• BS4 definition applied: lack of segregation in affected members of a family (ACMG/AMP 2015, PMID:25741868); no MSH3 CSPEC/VCEP specification exists, so the generic criterion semantics apply.
• Searched every case evidence file for family, pedigree, parental and segregation content: none pertains to this case, so no non-segregation observation exists.
• ClinVar VCV000644460 (8 submissions, no expert panel, 2-star) contains no family or segregation data; ClinVar classifications were not used as evidence for this criterion.
• PMID:27476653, the principal MSH3 inheritance paper retrieved for this case, describes families with biallelic MSH3 LoF but does not genotype this variant or provide any non-segregation observation applicable to c.1360C>T.

Still missing

• No family data of any kind (pedigree, affected relatives, or their genotype results) is available for this case.
• Non-segregation cannot be established or refuted without genotypes for c.1360C>T in affected family members.
• ClinVar VCV000644460 submissions provide no segregation or family-genotype information.

Sources

Reasoning

BP2 requires either that the variant be observed in cis with a pathogenic variant (in any inheritance pattern) or that it be observed in trans with a pathogenic variant in a fully penetrant dominant gene/disorder (ACMG/AMP 2015, PMID:25741868). Both clauses presuppose phase information for NM_002439.5:c.1360C>T p.(Arg454Ter) relative to a known pathogenic allele, and none exists in this case: there is no proband record, no zygosity statement, no parental data, and no second-allele testing. ClinVar VCV000644460 reports no phase or zygosity information for any of its 8 submissions (origin recorded as germline for 6 and unknown for 2). The second clause is additionally unsupported by the gene context - MSH3-associated adenomatous polyposis is autosomal recessive and attenuated rather than a fully penetrant dominant disorder (PMID:27476653, PMID:37402566), and heterozygous truncating MSH3 variants have not been established as causative (PMID:28528517) - but the first clause (cis with a pathogenic variant, any inheritance pattern) still requires an observation that the evidence does not contain. Population data likewise show no biallelic genotype (0 homozygotes in gnomAD v2.1 and v4.1). Verdict: not assessed - no cis/trans phase observation exists, so BP2 can be neither met nor formally excluded on the available data.

Rule

BP2 (ACMG/AMP 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. Requires phase information. No MSH3-specific VCEP/CSPEC specification available (generic ACMG/AMP applied).

Evidence considered

• BP2 definition applied: 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 (ACMG/AMP 2015, PMID:25741868). No MSH3 CSPEC/VCEP specification exists, so the generic definition is used.
• Searched all case evidence for phase, cis/trans, zygosity, parental and second-allele content relating to c.1360C>T: none is present. ClinVar VCV000644460, the only case-level clinical source, reports no phase or zygosity data in any of its 8 submissions (origin germline 6 / unknown 2) and no co-occurring pathogenic variant.
• MSH3 inheritance context: biallelic loss-of-function variants cause autosomal-recessive, attenuated colorectal adenomatous polyposis (PMID:27476653), and PMID:37402566 confirms MSH3 as a polyposis predisposition gene on the basis of biallelic germline variants; the fully penetrant dominant-disorder clause of BP2 therefore has no gene-level support, while the cis clause still requires data that are absent.
• PMID:28528517 adds that heterozygous clearly truncating germline MSH3 variants have not been reported as causative in Lynch-syndrome/colorectal-cancer patients, further indicating that a heterozygous MSH3 allele is not a dominant-acting pathogenic allele that could be configured in cis or trans with c.1360C>T.
• Neither full text retrieved for this case (PMID:27476653, PMID:28528517) mentions c.1360C>T or any codon-454 variant, so no published cis/trans observation exists for this variant; PMID:37402566 is abstract-only for this case and carries no variant-level detail.
• Population data show no biallelic observation: 0 homozygotes in gnomAD v2.1 (7/251,176 alleles) and gnomAD v4.1 (20/1,613,002 alleles).
• Because no phase observation exists in any direction, BP2 is recorded as not assessed rather than evaluated and failed; the criterion was checked against the same sources used for PM3 in this group.

Still missing

• No cis/trans phase observation exists for c.1360C>T in any source, so neither BP2 clause can be evaluated.
• ClinVar VCV000644460 records origin (germline/unknown) only, with no zygosity, phase, or co-occurring pathogenic variant data.
• No proband, parental, or family genotype data are present in this case with which a cis configuration could be identified.
• No ClinGen CSPEC or VCEP specification exists for MSH3, so generic BP2 semantics were applied with no gene-specific guidance.

Sources

Reasoning

BP5 (PMID:25741868) is a case-based criterion: the variant must be found in a case with an alternate molecular basis for disease (for example, a pathogenic variant in another gene that explains the phenotype, indicating this variant is not the cause). No VCEP-specific or likelihood-ratio-style BP5 definition exists for MSH3 (no CSPEC/VCEP, and no local gene framework), so only the generic case-based definition applies and no numeric threshold or comparison operator is available to evaluate. This case provides no proband-level data from which an alternate molecular basis could be judged: there is no phenotype, no co-occurring variant report, no pedigree, and no segregation or second-hit information in the case evidence (case_summary.json, evidence_brief.json, prefetch.json). The ClinVar submissions for variation 644460 all report c.1360C>T itself as the (Likely) pathogenic finding, and none reports a second, alternative causal variant; COSMIC records the variant once as a somatic event (COSV99435195), which is not an alternate germline molecular basis. Because the presence or absence of an alternate molecular basis cannot be determined for the tested individual, BP5 is returned as not_assessed rather than scored.

Evidence considered

• BP5 definition (PMID:25741868): variant found in a case with an alternate molecular basis for disease (e.g. a pathogenic variant in a different gene explaining the phenotype).
• Framework check: vcep_materials.json and final_classification_framework.json show no MSH3 CSPEC/VCEP (found=false, source=generic_acmg_fallback), so no VCEP-specific or likelihood-ratio-style BP5 threshold exists; no numeric operator comparison is applicable.
• Case-level check: no phenotype, pedigree, co-occurring variant, or second-hit data anywhere in case_summary.json, evidence_brief.json or prefetch.json; hci_prior 'found' = false.
• ClinVar variation 644460: all 8 single-submitter submissions report c.1360C>T as the finding itself (Pathogenic/Likely pathogenic); none reports an alternate molecular basis.
• COSMIC COSV99435195 records this variant once as a somatic alteration; that is not an alternate germline molecular basis for the patient's disease and does not satisfy BP5.

Still missing

• No proband-level genotype information (co-occurring pathogenic variants, second hits, or the test indication) exists in the case evidence, so whether an alternate molecular basis explains the phenotype cannot be determined.
• No MSH3-specific BP5 definition or numeric threshold exists; a VCEP/likelihood-ratio-based BP5 cannot be applied if none was specified.

Sources