MSH3 c.1897-12T>A, p.?

Reasoning

No applicable VCEP or gene-specific framework was available, so the generic ClinGen SVI PVS1 framework was considered. NM_002439.5:c.1897-12T>A is a noncanonical intronic single-nucleotide variant, not a nonsense, frameshift, or canonical ±1/2 splice-site variant; its normalized protein consequence is NP_002430.3:p.?. SpliceAI reports a maximum delta of 0.163, which is below the supplied PP3 splice threshold of 0.2 and within the 0.1–0.2 indeterminate interval, providing no established null splice consequence. Therefore PVS1 is not met; there is no PVS1 strength tier to assign.

Rule

Generic ClinGen SVI PVS1 framework (PMC6185798); no PVS1 bucket applies to this noncanonical intronic SNV without demonstrated splice disruption.

Evidence considered

• The generic ClinGen SVI PVS1 framework applies to established null-variant consequences; this variant is intronic at c.1897-12 and is outside the default nonsense, frameshift, and canonical ±1/2 splice-site buckets.
• SpliceAI scores are DS_AG 0.163, DS_AL 0.022, DS_DG 0.000, and DS_DL 0.005; the maximum delta is 0.163, below the supplied 0.2 splice-impact threshold and in the indeterminate 0.1–0.2 interval.
• The normalized consequence is NP_002430.3:p.? with no demonstrated transcript or protein truncation.

Still missing

• No variant-specific RNA or protein assay establishes abnormal splicing or loss of function.

Sources

Reasoning

No applicable MSH3-specific VCEP or gene framework was identified, so the supplied generic PM2 threshold of AF <= 0.0001 applies. The default all-comers gnomAD v4.1 overall AF is 0.0003354600, above the threshold; the explicitly relevant non-cancer values are also above it, including gnomAD v2.1 non-cancer exomes AF 0.0004513327 and gnomAD v3.1 non-cancer genomes AF 0.0016497187. Therefore the variant is not sufficiently rare for PM2 under the supplied threshold.

Rule

Generic ACMG/AMP PM2: allele frequency <= 0.0001; no applicable MSH3-specific rule was identified.

Evidence considered

• The case has no applicable MSH3-specific VCEP or gene-specific PM2 rule, so the supplied generic AF <= 0.0001 threshold was used.
• gnomAD v4.1 all-comers AF is 0.0003354600 (528/1,573,958 alleles; 3 homozygotes).
• The relevant non-cancer subsets are also above the threshold: gnomAD v2.1 non-cancer exomes AF 0.0004513327 (106/234,860 alleles; 1 homozygote) and gnomAD v3.1 non-cancer genomes AF 0.0016497187 (244/147,904 alleles; no homozygotes).
• The supplied generic PM2 threshold is AF <= 0.0001 (ClinGen SVI recommendation; PMID:25741868).

Sources

Reasoning

The variant NM_002439.5:c.1897-12T>A is intronic, so PP3 is evaluated only through the SpliceAI path. SpliceAI reports a maximum delta score of 0.163. This is below the supplied generic PP3 supporting threshold of >=0.2 from the SpliceAI recommended cutoff (Jaganathan et al. 2019, PMID:30661751), and therefore does not support PP3. No applicable MSH3-specific VCEP specification was identified.

Rule

SpliceAI PP3 supporting requires max delta >=0.2; variant max delta = 0.163.

Evidence considered

• The variant is intronic (NM_002439.5:c.1897-12T>A), so SpliceAI is the sole applicable PP3/BP4 predictor path; REVEL and BayesDel are not applicable.
• SpliceAI maximum delta score is 0.163 (DS_AG 0.163, DS_AL 0.022, DS_DG 0.000, DS_DL 0.005).
• The supplied generic SpliceAI PP3 supporting threshold is >=0.2, from Jaganathan et al. 2019, Cell (PMID:30661751).

Sources

Reasoning

PP5 is not met because the exact-variant ClinVar record has zero expert-panel submissions. The record contains only ordinary laboratory submissions, including benign and likely benign classifications, plus uncertain classifications; these are explicitly ineligible to trigger PP5 under the required rule.

Evidence considered

• The exact ClinVar variation 995588 has expert_panel_submissions = 0 and review status limited to criteria provided, single submitter.
• The exact-variant laboratory classifications are uncertain, benign, and likely benign; no exact-variant expert-panel Pathogenic or Likely pathogenic classification is present.

Sources

Reasoning

No applicable MSH3-specific VCEP or gene framework was identified, so the supplied generic ACMG/AMP BA1 threshold of AF >= 0.05 applies. The highest observed allele frequency is 0.0061080848 in the gnomAD v4.1 African/African American population, with overall AF 0.0003354600; both are below 0.05. The non-cancer subsets likewise remain below this threshold, with gnomAD v2.1 non-cancer exomes AF 0.0004513327 and gnomAD v3.1 non-cancer genomes AF 0.0016497187.

Rule

Generic ACMG/AMP BA1: allele frequency >= 0.05; no applicable MSH3-specific rule was identified.

Evidence considered

• The case has no applicable MSH3-specific VCEP or gene-specific population rule, so the supplied generic ACMG/AMP BA1 threshold of AF >= 0.05 was used.
• gnomAD v4.1 reports overall AF 0.0003354600 and maximum ancestry AF 0.0061080848 in African/African American individuals; these are below 0.05.
• The supplied generic BA1 threshold is the ACMG/AMP 2015 stand-alone threshold of AF >= 0.05 (PMID:25741868).

Sources

Reasoning

No applicable MSH3-specific VCEP or gene framework was identified, so the supplied generic BS1 threshold of AF >= 0.01 applies. The highest observed allele frequency is 0.0061080848 in gnomAD v4.1 African/African American individuals, and the highest reported non-cancer subgroup value is 0.0058200595 in gnomAD v2.1 non-cancer African/African American exomes; neither reaches 0.01. The gnomAD v3.1 non-cancer HGDP:YORUBA value of 0.0263158 is based on only 1 of 38 alleles and is a small-subgroup outlier, while its reported grpmax FAF is 0.00502799 and does not meet the threshold.

Rule

Generic ACMG/AMP BS1: allele frequency >= 0.01; no applicable MSH3-specific rule was identified.

Evidence considered

• The case has no applicable MSH3-specific VCEP or gene-specific BS1 rule, so the supplied generic AF >= 0.01 threshold was used.
• The highest standard gnomAD v4.1 ancestry AF is 0.0061080848; the gnomAD v2.1 non-cancer exome ancestry AF is 0.0058200595 and grpmax FAF is 0.00517448.
• gnomAD v3.1 non-cancer reports a 1/38 HGDP:YORUBA subgroup value of 0.0263158, but its grpmax FAF is 0.00502799; the isolated 1/38 observation is treated as a small-sample threshold artifact rather than representative population frequency.
• The supplied generic BS1 threshold is AF >= 0.01, with per-gene derivation preferred (PMID:28518168).

Sources

Reasoning

The variant NM_002439.5:c.1897-12T>A is intronic, so BP4 is evaluated only through the SpliceAI path. SpliceAI reports a maximum delta score of 0.163. This exceeds the supplied generic BP4 supporting threshold of <=0.1 from the SpliceAI recommended cutoff (Jaganathan et al. 2019, PMID:30661751), and therefore does not support BP4. No applicable MSH3-specific VCEP specification was identified.

Rule

SpliceAI BP4 supporting requires max delta <=0.1; variant max delta = 0.163.

Evidence considered

• The variant is intronic (NM_002439.5:c.1897-12T>A), so SpliceAI is the sole applicable PP3/BP4 predictor path; REVEL and BayesDel are not applicable.
• SpliceAI maximum delta score is 0.163 (DS_AG 0.163, DS_AL 0.022, DS_DG 0.000, DS_DL 0.005).
• The supplied generic SpliceAI BP4 supporting threshold is <=0.1, from Jaganathan et al. 2019, Cell (PMID:30661751).

Sources

Reasoning

BP6 is not met because the exact-variant ClinVar record has zero expert-panel submissions. Although ordinary laboratories submitted Benign and Likely benign classifications, those non-expert assertions cannot trigger BP6 under the required rule.

Evidence considered

• The exact ClinVar variation 995588 has expert_panel_submissions = 0.
• Benign and likely benign classifications are from single clinical laboratories rather than an exact-variant ClinVar expert panel, so they are ineligible for BP6.

Sources

Reasoning

No case-level evidence documents that NM_002439.5:c.1897-12T>A occurred de novo in an affected proband, and no parental genotyping or maternity/paternity confirmation is provided.

Evidence considered

• No case-level proband, parental-testing, or confirmed de novo observation was identified for NM_002439.5:c.1897-12T>A.

Still missing

• Affected-proband clinical documentation is not provided.
• Both biological parents require variant testing with documented maternity/paternity confirmation to assess de novo status.

Reasoning

No applicable MSH3-specific VCEP specification was identified, so generic ACMG/AMP functional-evidence principles apply. The reviewed evidence contains no validated RNA, protein, minigene, cellular, or other disease-relevant functional assay for NM_002439.5:c.1897-12T>A, with appropriate controls and a result interpretable as abnormal. SpliceAI prediction (maximum delta 0.163) is computational evidence and cannot establish PS3. The available literature extraction likewise reports that the exact variant is not mentioned and provides no variant-specific functional result; therefore PS3 cannot be assessed.

Rule

PS3 requires well-established functional studies showing a damaging effect; no qualifying variant-specific functional study is available.

Evidence considered

• The case evidence identifies NM_002439.5:c.1897-12T>A as an intronic MSH3 variant and provides no direct RNA, protein, minigene, cellular, or other functional assay result for it.
• SpliceAI reports a maximum delta score of 0.163, but this is an in-silico prediction and is not a validated functional assay result.
• The ACMG/AMP guidance extraction states that predicted splice effects require confirmation by RNA or protein analysis; it does not report a functional result for this variant (PMID:25741868).
• The splicing-validation paper describes a genome-wide RNA-seq resource but does not mention MSH3 c.1897-12T>A or provide a variant-specific assay result (PMID:31275557).

Still missing

• A validated, disease-relevant assay directly testing NM_002439.5:c.1897-12T>A is needed, with appropriate positive, negative, and assay-performance controls.
• Variant-specific RNA or protein evidence demonstrating an abnormal effect is unavailable.

Sources

Reasoning

No case-control enrichment analysis, affected-case series, odds ratio, likelihood ratio, or statistically supported excess of this exact MSH3 variant in affected individuals is available. The literature findings reviewed do not mention NM_002439.5:c.1897-12T>A, and ClinVar contains no expert-panel or cohort-level phenotype evidence that could establish PS4.

Evidence considered

• ClinVar variation 995588 is an exact transcript-level match but reports conflicting single-submitter classifications and no expert-panel submission; it does not provide case-control enrichment evidence.
• The reviewed publications do not mention MSH3 c.1897-12T>A or provide variant-specific affected-case counts, controls, odds ratios, or likelihood ratios.

Still missing

• A case-control or validated affected-case enrichment dataset for NM_002439.5:c.1897-12T>A is unavailable.
• No applicable MSH3-specific PS4 specification was retrieved.

Sources

Reasoning

PM3 cannot be assessed because the available evidence does not document an affected proband with a second pathogenic or likely pathogenic MSH3 variant, nor does it establish biallelic inheritance or phase (trans). The variant is intronic (NM_002439.5:c.1897-12T>A), but no case-level genotype, family segregation, or phase observation is provided. Population homozygotes in gnomAD are not affected-proband observations and do not satisfy PM3.

Rule

Generic ACMG/AMP PM3: pathogenic or likely pathogenic variant observed in trans with the variant under assessment in an affected individual with a recessive disorder; phase and disease relevance must be established.

Evidence considered

• ClinVar provides an exact variant record but no affected-proband genotype, second MSH3 variant, family segregation, or phase information.
• gnomAD reports 1 homozygote in v2.1 and 3 homozygotes in v4.1, but these are population observations without affected-proband or disease-status evidence and therefore do not establish PM3.

Still missing

• Affected-proband observations are unavailable.
• A second pathogenic or likely pathogenic MSH3 variant is not documented.
• Cis/trans phase and inheritance are not established.

Sources

Reasoning

No case-level observation supports an assumed de novo occurrence of NM_002439.5:c.1897-12T>A in an affected proband when parental testing is unavailable; therefore PM6 cannot be applied.

Evidence considered

• No case-level de novo observation without parental testing was identified for NM_002439.5:c.1897-12T>A.

Still missing

• A documented affected proband carrying the variant is required.
• The clinical report must explicitly support de novo status despite unavailable or incomplete parental testing.

Reasoning

No familial case series, informative meioses, affected relatives, or genotype-phenotype segregation data are provided for NM_002439.5:c.1897-12T>A; segregation evidence is therefore insufficient to apply PP1.

Evidence considered

• No familial segregation, informative meiosis, or affected-relative genotype data were identified for NM_002439.5:c.1897-12T>A.

Still missing

• A pedigree or family structure with informative meioses is required.
• Genotypes and relevant phenotypes for affected and unaffected relatives are required to evaluate cosegregation.

Reasoning

No individual-level phenotype, disease-specific clinical presentation, or highly specific phenotype-to-gene correlation is supplied for the variant. The available records identify MSH3 and the intronic variant but do not document a phenotype sufficiently specific to support PP4.

Evidence considered

• The exact ClinVar record provides no patient phenotype or disease-specific clinical description for NM_002439.5:c.1897-12T>A.
• The reviewed literature does not report this exact variant in a clinically characterized patient.

Still missing

• Individual clinical phenotype and diagnostic indication are unavailable.
• No MSH3-specific PP4 phenotype framework was retrieved.

Sources

Reasoning

The variant is observed in homozygotes in gnomAD v2.1 non-cancer exomes (1 homozygote) and gnomAD v4.1 overall data (3 homozygotes), but BS2 requires documented observation in healthy adults and interpretation depends on the disease inheritance model, age-related penetrance, and phenotype ascertainment. No MSH3-specific framework establishes those requirements here, and the supplied case evidence does not document the homozygotes' ages, clinical status, or whether MSH3 disease is dominant or recessive for this assessment. The homozygote counts therefore cannot be converted into BS2 without human adjudication.

Rule

Generic ACMG/AMP BS2 requires the variant in healthy adults, with interpretation dependent on the relevant inheritance model; human review is required.

Evidence considered

• gnomAD v2.1 non-cancer exomes report 1 homozygote, and gnomAD v4.1 reports 3 homozygotes overall.
• The available population records do not provide individual ages, clinical phenotypes, or penetrance data for those homozygotes.
• The supplied ACMG/AMP framework requires healthy-individual observations to support BS2; the inheritance-specific interpretation cannot be resolved from the available evidence alone (PMID:25741868).

Still missing

• Confirm the applicable MSH3 disease inheritance model for this variant and condition.
• Determine whether the reported homozygotes were healthy adults with adequate age and phenotype ascertainment.
• Assess disease penetrance and whether the observed homozygote count is incompatible with pathogenicity.

Sources

Reasoning

No applicable MSH3-specific VCEP specification was identified, so generic ACMG/AMP functional-evidence principles apply. The reviewed evidence contains no validated RNA, protein, minigene, cellular, or other disease-relevant functional assay for NM_002439.5:c.1897-12T>A, with appropriate controls and a result interpretable as normal. SpliceAI prediction (maximum delta 0.163) is computational evidence and cannot establish BS3. The available literature extraction likewise reports that the exact variant is not mentioned and provides no variant-specific functional result; therefore BS3 cannot be assessed.

Rule

BS3 requires well-established functional studies showing no damaging effect; no qualifying variant-specific functional study is available.

Evidence considered

• The case evidence identifies NM_002439.5:c.1897-12T>A as an intronic MSH3 variant and provides no direct RNA, protein, minigene, cellular, or other functional assay result for it.
• SpliceAI reports a maximum delta score of 0.163, but this is an in-silico prediction and is not a validated functional assay result.
• The ACMG/AMP guidance extraction states that splice-effect predictions require confirmation by RNA or protein analysis; it does not report a normal functional result for this variant (PMID:25741868).
• The splicing-validation paper describes a genome-wide RNA-seq resource but does not mention MSH3 c.1897-12T>A or provide a variant-specific assay result (PMID:31275557).

Still missing

• A validated, disease-relevant assay directly testing NM_002439.5:c.1897-12T>A is needed, with appropriate positive, negative, and assay-performance controls.
• Variant-specific RNA or protein evidence demonstrating a normal effect is unavailable.

Sources

Reasoning

No family data document unaffected relatives who were tested and do not carry NM_002439.5:c.1897-12T>A, and no phenotype-qualified non-segregation analysis is available; BS4 cannot be assessed.

Evidence considered

• No tested unaffected relatives, negative familial genotypes, or phenotype-qualified non-segregation evidence were identified for NM_002439.5:c.1897-12T>A.

Still missing

• Tested unaffected relatives with reliable clinical phenotyping are required.
• Their negative variant results and the expected disease model must be documented to evaluate non-segregation.

Reasoning

BP2 cannot be assessed because no reliable observation documents this variant with another pathogenic variant in cis or trans in a relevant individual. The available ClinVar record contains no paired-variant or phase data, and the population databases provide allele counts and homozygote counts without linked haplotype or clinical genotype information.

Rule

Generic ACMG/AMP BP2: variant observed in trans with a pathogenic variant for a fully penetrant dominant disorder, or in cis with a pathogenic variant for a recessive disorder; phase and relevant clinical context are required.

Evidence considered

• ClinVar provides an exact variant record but no paired pathogenic variant, phase, or inheritance information.
• gnomAD provides population allele and homozygote counts but does not provide linked haplotypes or clinical observations needed for BP2.

Still missing

• A relevant individual carrying this variant with another pathogenic variant is not documented.
• Cis/trans phase and clinical context are unavailable.

Sources

Reasoning

No patient-level evidence shows that the phenotype is explained by an alternative molecular etiology while this MSH3 variant is present. The available ClinVar and literature records do not document a confirmed alternate pathogenic variant in an affected individual, so BP5 cannot be assessed.

Evidence considered

• The exact ClinVar record contains no patient-level alternate molecular diagnosis or paired-variant clinical evidence.
• The reviewed publications do not report this exact variant in a case with an independently established alternative molecular cause.

Still missing

• Patient-level testing and a confirmed alternative molecular diagnosis are unavailable.
• No BP5-specific numeric likelihood-ratio or other governing threshold is applicable to the available evidence.

Sources