Five-step scale from benign to pathogenic. This variant is pathogenic. Select a step to read what it means.
This variantScenario
BenignPathogenic
Pathogenic.A pathogenic classification under the recorded framework.
In plain language
Gene-level background from the recorded interpretation.
Words with a dotted underline have a short definition. Hover or tap them.
About MLH1
MLH1 encodes a DNA mismatch repair protein that, together with partners such as PMS2, corrects errors that arise during DNA replication. Loss of this repair function leads to the accumulation of mutations, particularly in repetitive microsatellite sequences, and tumors with defective MLH1 typically show high microsatellite instability. Germline mutations in MLH1 cause Lynch syndrome (hereditary non-polyposis colorectal cancer), which strongly predisposes to colorectal, endometrial, ovarian, and other cancers, while biallelic mutations cause constitutional mismatch repair deficiency (CMMRD). Because it normally protects against cancer by maintaining genomic stability, MLH1 acts as a tumor suppressor, and mismatch-repair-deficient tumors often respond well to immunotherapy.
LYFE Sciences is an AI system, and it can make mistakes.
Criteria may be applied incorrectly, sources may be misread, and a
confident-looking classification can still be wrong.
Double-check every criterion and its underlying evidence
before relying on any call.
01The reasoning
How it was classified
The recorded framework is ClinGen MLH1 VCEP v2.0. The criteria and strengths shown below come from the saved assessment; this page does not recalculate the classification.
Recorded applied criteria
The applied criteria and their strengths are shown as assessed under the stated framework.
Met (Very Strong): disrupts the canonical splice-acceptor -2 position; SpliceAI predicts near-total acceptor loss (max delta 0.996), with exon skipping predicted to trigger nonsense-mediated decay. Caveat for review: the frameshift/NMD outcome was assumed, as RNA confirmation was not available.
PM2Supporting
PM2 · supporting
Met (Supporting): absent from gnomAD v4.1, below the 0.00002 allele-frequency threshold.
PP5Supporting
PP5 · supporting
Met (Supporting): the InSiGHT expert panel classified this variant as Likely pathogenic.
Benign criteria applied · 0
No benign criterion marked as applied in the recorded assessment.
Recorded classification: Pathogenic
Under ClinGen InSiGHT MLH1 VCEP v2.0, PVS1 (Very Strong) plus PM2 and PP5 (Supporting) satisfies Rule4 (1 Very Strong + at least 2 Supporting), yielding Pathogenic.
02The sources
Evidence at a glance
Evidence summaries and captured source records, where available. Missing entries should not be interpreted as negative findings.
gnomAD v4.1
gnomAD v2.1
Population frequency
Not recorded
gnomAD v4.1 allele counts were not recorded for this variant.
ACMG/AMP criteria under ClinGen MLH1 VCEP v2.0, with the data and sources behind each one.
28 criteria with recorded statuses: 3 met · 2 not met · 10 not assessed · 13 not applicable. Open any row for the full reasoning, the rule and the evidence behind it.
PVS1 applied at Very Strong strength based on the InSiGHT MLH1 CSPEC v2.0 canonical splice-site (IVS+/-1/+/-2) rule, supported by near-complete SpliceAI-predicted loss of the natural acceptor (max delta 0.996) and MLH1's established germline loss-of-function mechanism; flagged for human review because the CSPEC Appendix table needed to independently confirm the exon 16 skip reading-frame/NMD outcome was not extractable as text from the case's local VCEP materials.
PM4 and BP3 are both explicitly listed as Not Applicable in the InSiGHT MLH1 CSPEC v2.0 and are independently inapplicable to this canonical splice-site substitution, which is not an in-frame protein-length-change variant.
NM_000249.4:c.1732-2A>G is a canonical splice-acceptor variant (position -2 of the AG dinucleotide) with no missense/amino-acid consequence (protein_1l/protein_3l = NP_000240.1:p.?); it is evaluated by other criteria groups (e.g., PVS1) rather than by codon/residue/domain-based codes.
The InSiGHT MLH1 VCEP Specification v2.0 explicitly designates PM1, PP2, and BP1 as Not Applicable for MLH1.
PS1's non-canonical splice-nucleotide branch does not apply because this variant occupies the canonical (invariant) splice-acceptor -2 position, not a non-canonical splice nucleotide.
PM5 requires a missense residue comparator, which does not exist for this splice variant; the pm5_candidates.json artifact confirms ineligibility for the classic same-residue missense PM5 search.
MLH1 VCEP (InSiGHT) v2.0 cspec defines PS3/BS3 calibration on functional odds-for-pathogenicity thresholds and lists gene-specific calibrated assays in the VCEP Functional-assay-SVI-documentation-MMR.xlsx reference; none of those calibrated studies, nor any publication in the case bundle, report a functional assay performed on the exact allele NM_000249.4:c.1732-2A>G.
A closely related variant at the same splice-acceptor nucleotide position, c.1732-2A>T, was functionally characterized (RT-PCR, exon 16 skipping, classified pathogenic) in PMID 24090359, but per strict citation-grounding rules this distinct nucleotide substitution cannot be used to ground PS3/BS3 for c.1732-2A>G.
Both PS3 and BS3 are therefore returned as not_assessed pending direct functional/RNA evidence on this exact allele.
MLH1 is governed by the ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis VCEP specification (cspec), which defines gene-specific PP3/BP4 pathways: HCI-prior lookup for missense substitutions, and SpliceAI delta-score thresholds for non-canonical splicing nucleotides (intronic/synonymous) only.
NM_000249.4:c.1732-2A>G is a canonical splice-acceptor consensus variant (-2 position), which the case's pvs1_variant_assessment explicitly routes to the PVS1 pathway and instructs not to double-count with PP3 for the same splice-prediction evidence.
Because the variant is neither missense nor a non-canonical splice-site variant, and its position (-2) does not meet BP7's -21/+7 boundary, none of PP3, BP4, or BP7 have an applicable pathway for this variant under the governing VCEP framework; all three are scored not_applicable rather than not_met to avoid implying a computational prediction was weighed and found insufficient.
Population evidence under the applicable MLH1 InSiGHT VCEP specification supports PM2 at Supporting strength only; BA1 and BS1 frequency thresholds are not met, and BS2 is not assessed because the required phased co-occurrence and phenotype evidence is absent.
No criteria match.
Met · 3
PVS1Very strongMet (Very Strong): disrupts the canonical splice-acceptor -2 position; SpliceAI predicts near-total acceptor loss (max delta 0.996), with exon skipping predicted to trigger nonsense-mediated decay. Caveat for review: the frameshift/NMD outcome was assumed, as RNA confirmation was not available.
Details ▾Hide ▴
Reasoning
NM_000249.4:c.1732-2A>G disrupts the canonical splice acceptor consensus at intron 15/exon 16 (IVS-2, i.e. the -2 position of the 3' splice acceptor immediately preceding the affected exon), a position explicitly covered by the InSiGHT MLH1 CSPEC v2.0 PVS1 rule for 'Variants at IVS+/-1 or IVS+/-2 where exon skipping or use of a cryptic splice site disrupts reading frame and is predicted to undergo NMD' (Very Strong tier). VariantValidator places the variant at the intron-15/exon-16 boundary ('15i'), consistent with disruption of the natural acceptor for the downstream exon. SpliceAI predicts near-certain loss of the natural splice acceptor (DS_AG=0.732, DS_AL=0.996; max delta score 0.996) with negligible competing donor signal, indicating the natural acceptor is essentially abolished and splicing will proceed either via exon skipping or a cryptic acceptor rather than normal splicing. MLH1 loss-of-function is an established disease mechanism for Lynch syndrome, and the CSPEC includes explicit PVS1 guidance for MLH1 (pvs1_gene_context: lof_mechanism_supported=true, pvs1_gene_gate=eligible), so PVS1 is applicable at full strength for canonical splice-site disruption in this gene per gene-specific guidance rather than falling back to generic ACMG. Per the CSPEC footnote structure, the tier (Very Strong vs Strong vs Moderate) for an IVS+/-1/+/-2 variant depends on whether the resulting exon-skipping event disrupts the reading frame and is predicted to trigger NMD, versus preserving frame in a functionally important region (Strong) or disrupting frame without NMD (Moderate). The CSPEC Appendix reportedly contains a curated list of MLH1 exons that would be in-frame versus out-of-frame if skipped, but that Appendix content is not extractable as text from the case's VCEP materials (PVS1_DecisionTree_MMR.pdf and MMR-Functional-domains.pdf are image-based PDFs with no OCR/text layer available to this pipeline), so the exon 16 skip frame/NMD status could not be independently confirmed from case materials alone. Given the near-total predicted loss of the natural acceptor and no evidence of a competing in-frame cryptic acceptor or donor signal, the modal outcome (exon skipping causing frameshift and NMD) is applied as the default per the CSPEC rule, consistent with the pipeline's own generic PVS1 scaffold, which independently flags this variant as a canonical +/-1,2 splice consensus disruption warranting full-strength PVS1 pending exon-frame confirmation.
Rule
InSiGHT MLH1 CSPEC v2.0 PVS1, Very Strong tier: nonsense/frameshift PTC at/before codon 753 OR large genomic alteration OR variant at IVS+/-1 or IVS+/-2 where exon skipping/cryptic splice site use disrupts reading frame and is predicted to undergo NMD OR variant at the initiation codon.
Evidence considered
InSiGHT MLH1 CSPEC v2.0 PVS1 rule (Very Strong tier): variants at IVS+/-1 or IVS+/-2 where exon skipping or cryptic splice site use disrupts reading frame and is predicted to undergo NMD; not to be combined with PP3 and not for use with a confirmed splice defect from RNA data.
VariantValidator exonic position annotation places the variant at '15i' (intron 15/exon 16 boundary) for both GRCh37 and GRCh38, confirming the variant lies at the canonical splice acceptor consensus immediately 5' of an MLH1 exon.
SpliceAI lookup (queried at NM_000249.4:c.1732-2A>G) reports DS_AG=0.732, DS_AL=0.996 (acceptor loss), DS_DG=0.0, DS_DL=0.114, with max delta score 0.996, indicating near-certain loss of the natural splice acceptor and no meaningful competing donor-site signal.
pvs1_gene_context.json: MLH1 germline loss-of-function is an established disease mechanism per the official CSPEC/VCEP (lof_mechanism_supported=true, pvs1_gene_gate=eligible), supporting PVS1 applicability in this gene.
pvs1_variant_assessment.json: variant classified as consequence_class='canonical_splice' with canonical_splice_consensus=true and suggested_default_strength='PVS1', noting PP3 should not be double-counted for the same splice-prediction evidence and that downgrade should be considered only if the truncation escapes NMD or the affected region is non-critical/enriched for population LoF variation.
The CSPEC Appendix referenced in the v2.0 release notes ('List of exons that would be in-frame and out-of-frame added to Appendix') could not be read as text from the local VCEP PDF file in this case, so the exon 16-specific frame call was not independently confirmed.
Still missing
The InSiGHT MLH1 CSPEC v2.0 Appendix listing which MLH1 exons are in-frame versus out-of-frame if skipped (referenced in the CSPEC release notes) could not be read as text from the case's local VCEP files (PVS1_DecisionTree_MMR.pdf is an image-only PDF with no extractable text in this pipeline), so exon 16 frame/NMD status was not independently verified against that table.
No patient-derived RNA/minigene splicing assay data are present in this case bundle to directly confirm the predicted exon-skipping or cryptic-splice-site outcome; the assessment relies on SpliceAI in silico prediction and the canonical splice-site position per CSPEC guidance.
Human review recommended to confirm exon 16 reading-frame/NMD status against the CSPEC Appendix before finalizing the Very Strong versus Strong/Moderate tier.
PM2SupportingMet (Supporting): absent from gnomAD v4.1, below the 0.00002 allele-frequency threshold.
Details ▾Hide ▴
Reasoning
The variant is absent from gnomAD v4.1, satisfying the MLH1 InSiGHT PM2_Supporting requirement for an allele frequency below 0.00002.
Rule
MLH1 InSiGHT VCEP v2.0 PM2_Supporting: absent/extremely rare allele frequency <0.00002 (<1 in 50,000 alleles) in gnomAD v4.
Evidence considered
The applicable MLH1 InSiGHT VCEP specification (version 2.0) assigns PM2 at Supporting strength for an absent or extremely rare gnomAD v4 allele frequency below 0.00002 (less than 1 in 50,000 alleles).
gnomAD v4.1 reports NM_000249.4:c.1732-2A>G (GRCh38 chr3:37047517 A>G) as absent, satisfying the below-0.00002 threshold.
PP5SupportingMet (Supporting): the InSiGHT expert panel classified this variant as Likely pathogenic.
Details ▾Hide ▴
Reasoning
Expert panel International Society for Gastrointestinal Hereditary Tumours (InSiGHT) classified as Likely pathogenic.
Rule
PP5 Supporting: exact-variant ClinVar expert-panel Pathogenic or Likely pathogenic classification.
Evidence considered
ClinVar variation 89861 is an exact transcript-HGVS match and includes SCV000106332 from InSiGHT, reviewed by expert panel, with Likely pathogenic significance for Lynch syndrome.
BA1Not met: absent from gnomAD v4.1, so the BA1 population-frequency threshold cannot be reached.
Details ▾Hide ▴
Reasoning
The variant is absent from gnomAD v4.1 and therefore cannot meet the MLH1 InSiGHT BA1 threshold.
Rule
MLH1 InSiGHT VCEP v2.0 BA1: gnomAD v4 grpmax filtering allele frequency >=0.001 (0.1%) and excluded as founder pathogenic variant.
Evidence considered
The applicable MLH1 InSiGHT VCEP specification (version 2.0) assigns BA1 only when gnomAD v4 grpmax filtering allele frequency is at least 0.001 (0.1%) and the variant is excluded as a founder pathogenic variant.
gnomAD v4.1 reports NM_000249.4:c.1732-2A>G (GRCh38 chr3:37047517 A>G) as absent; its allele frequency is therefore below the BA1 threshold.
BS1Not met: absent from gnomAD v4.1, so the BS1 population-frequency range cannot be reached.
Details ▾Hide ▴
Reasoning
The variant is absent from gnomAD v4.1 and therefore cannot meet the MLH1 InSiGHT BS1 frequency range.
Rule
MLH1 InSiGHT VCEP v2.0 BS1: gnomAD v4 grpmax filtering allele frequency >=0.0001 and <0.001 (0.01%-0.1%) and excluded as founder pathogenic variant.
Evidence considered
The applicable MLH1 InSiGHT VCEP specification (version 2.0) assigns BS1 only when gnomAD v4 grpmax filtering allele frequency is at least 0.0001 and below 0.001 (0.01%-0.1%), with exclusion of a founder pathogenic variant.
gnomAD v4.1 reports NM_000249.4:c.1732-2A>G (GRCh38 chr3:37047517 A>G) as absent; its allele frequency is below the BS1 range.
PS2Not assessed: no de novo observation or parental test results were available.
Details ▾Hide ▴
Reasoning
No proband-level de novo observation, parental test results, parentage confirmation, or qualifying Lynch syndrome tumor information is provided. Therefore no InSiGHT PS2 de novo points can be assigned.
Rule
MLH1 InSiGHT v2.0 PS2: aggregate de novo points; Supporting = 0.5, Moderate = 1-1.5, Strong = 2-3.5, Very Strong = at least 4.
Evidence considered
The MLH1 InSiGHT specification assigns PS2 points only to a proband with a de novo variant and an LS-spectrum tumor: 2 points when maternity and paternity are confirmed and MMR-deficient tumor evidence is concordant (with MLH1 methylation excluded), 1 point when both parents are confirmed but tumor data are unavailable, or 0.5 point when one or both parents are unconfirmed and tumor data are unavailable.
The case bundle provides variant-level database, prediction, and literature metadata but no proband, parental, parentage, or tumor data for a de novo-point calculation.
Still missing
Proband phenotype and Lynch syndrome tumor type.
Parental testing for c.1732-2A>G and confirmation of maternity and paternity.
For a 2-point observation, tumor MSI/IHC concordance with MLH1 and MLH1 promoter methylation result.
PS3Not assessed: no calibrated functional assay or RNA/cDNA study of this exact variant was available.
Details ▾Hide ▴
Reasoning
The InSiGHT MLH1 VCEP v2.0 specification requires PS3 be supported by calibrated functional assays with a documented functional odds-for-pathogenicity value, or an MMR functional assay flowchart result, or monoallelic-expression/full-length-transcript data, applied to the specific variant under review. The VCEP-approved calibrated assay table (Functional-assay-SVI-documentation-MMR.xlsx) lists calibrated protein-level MMR repair-capacity, cell-survival, and cDNA/full-length-transcript assays, but none of the listed studies (PMIDs 30504929, 31965077, 33357406, 36550560, 36054288, 24362816, 31332305, 30998989) report results for c.1732-2A>G or an equivalent nucleotide change at this position. No source in this case bundle reports an RNA/minigene splicing assay, cDNA analysis, or monoallelic-expression study performed specifically on the NM_000249.4:c.1732-2A>G allele.
Still missing
No calibrated functional assay (RNA/cDNA, minigene, or protein-based MMR assay) with a reported functional odds-for-pathogenicity value could be found for the exact variant c.1732-2A>G in the case bundle or reviewed publications.
A related but distinct nucleotide substitution at the identical splice-acceptor position (c.1732-2A>T, not c.1732-2A>G) was functionally characterized by RT-PCR in a Danish cohort study and shown to cause in-frame skipping of exon 16, but this is a different allele and cannot be used to ground PS3 for c.1732-2A>G per strict variant-match citation requirements.
SpliceAI predicts a high-confidence acceptor-loss effect (DS_AL=0.996) for this variant, but computational splice prediction is not equivalent to a validated functional assay and cannot be used to satisfy PS3 under this VCEP specification.
If direct RNA/mRNA studies on patient-derived material for this exact variant become available, PS3 should be re-evaluated against the VCEP odds-for-pathogenicity thresholds (Strong >18.7, Moderate >4.3-18.7, Supporting >2.08-4.3).
PM3Not assessed: no evidence of a second MLH1 variant or phase information indicating CMMRD was available.
Details ▾Hide ▴
Reasoning
The MLH1 InSiGHT VCEP specification permits PM3 only for a variant meeting PM2_Supporting that co-occurs with a pathogenic/likely pathogenic MLH1 variant in a person with clinical features consistent with CMMRD. No affected-person observation, second MLH1 variant, zygosity, or phase information is present in the consolidated evidence bundle; therefore no PM3 points can be assigned.
Rule
MLH1 InSiGHT VCEP PM3: pathogenic/likely pathogenic variant in trans = 1.0 point; phase unknown = 0.5 point; PM3_Supporting = 0.5 point, PM3_Moderate = 1-1.5 points, PM3_Strong = 2-3.5 points, PM3_VeryStrong = >=4 points; variant must meet PM2_Supporting and the patient must have CMMRD-compatible clinical features.
Evidence considered
The MLH1 InSiGHT VCEP PM3 specification requires PM2_Supporting and a co-occurrence with a pathogenic/likely pathogenic MLH1 variant in a patient with CMMRD-compatible clinical features; confirmed trans phase is 1.0 point and unknown phase is 0.5 point.
gnomAD v4.1 reports this variant absent, but population absence alone does not provide the required second-variant, phenotype, or phase evidence for PM3.
Still missing
Variant-specific affected-person reports with CMMRD-compatible clinical features.
Identity and pathogenic/likely pathogenic classification of any second MLH1 variant.
Phase determination by parental or offspring testing, or explicit documentation that phase is unknown for a qualifying co-occurrence.
Confirmation that PM2_Supporting is met, as required by the MLH1 VCEP PM3 specification.
PP1Not assessed: no pedigree genotypes or segregation data were available.
Details ▾Hide ▴
Reasoning
No pedigree-level genotypes, cancer phenotypes, ages, meioses, or calculated overall Bayes likelihood ratio are provided for the exact c.1732-2A>G variant. PP1 cannot be assigned.
Rule
MLH1 InSiGHT v2.0 PP1: combined pedigree Bayes likelihood ratio thresholds are Supporting >2.08 to 4.3, Moderate >4.3 to 18.7, and Strong >18.7 in at least two families.
Evidence considered
The MLH1 InSiGHT specification requires a combined pedigree Bayes likelihood ratio for PP1: Supporting >2.08 to 4.3, Moderate >4.3 to 18.7, and Strong >18.7 in at least two families.
No case-specific segregation dataset or Bayes likelihood ratio is present in the consolidated bundle.
Still missing
Pedigree(s) with tested relatives, variant genotypes, cancer diagnoses, and ages at diagnosis or last follow-up.
Combined overall Bayes likelihood ratio from an MMR-appropriate segregation analysis.
PP4Not assessed: an MSI-H tumor with MLH1/PMS2 loss was reported, but no primary tumor report confirmed the MSI method, tumor independence, or methylation exclusion.
Details ▾Hide ▴
Reasoning
A ClinVar single-submitter evidence signal describes an MSI-H Lynch syndrome-associated tumor with MLH1/PMS2 loss, which would be phenotypically consistent with an MLH1 variant. However, no primary tumor report was provided to establish the required MSI method, tumor independence, and exclusion of MLH1 promoter methylation; therefore PP4 cannot be applied.
Rule
MLH1 InSiGHT VCEP PP4: Supporting requires one qualifying CRC/endometrial MSI-H tumor and/or concordant MMR IHC with MLH1 promoter methylation excluded; Moderate requires two and Strong requires at least three independent qualifying tumors in at least two families.
Evidence considered
The MLH1 InSiGHT PP4 rule requires one or more CRC/endometrial MSI-H tumors and/or MMR protein loss consistent with MLH1, with MLH1 promoter methylation excluded; strength depends on the number of independent qualifying tumors.
ClinVar's exact-variant Ambry submission evidence signal reports a Lynch syndrome-associated tumor described as MSI-H with loss of MLH1/PMS2 expression, but the bundled record does not provide the underlying tumor report or methylation result.
Still missing
Primary tumor report(s) documenting CRC or endometrial tumor type, MSI-H testing by a standard 5-10-marker panel or tumor genome and/or concordant MMR IHC.
For each MLH1-associated tumor, MLH1 promoter methylation must be excluded.
If more than one tumor is proposed, documentation that tumors are independent is required.
BS2Not assessed: no confirmed in-trans occurrence with a known pathogenic MLH1 variant was documented.
Details ▾Hide ▴
Reasoning
No evidence documents a confirmed in-trans occurrence with a known pathogenic MLH1 variant in an appropriately phenotyped individual meeting the MLH1-specific BS2 requirements.
Rule
MLH1 InSiGHT VCEP v2.0 BS2: confirmed in-trans co-occurrence with a known pathogenic MLH1 variant in the specified later-onset Lynch syndrome cancer context and no CMMRD manifestations; parental or offspring testing required for phase.
Evidence considered
The applicable MLH1 InSiGHT VCEP specification requires confirmed in-trans co-occurrence with a known pathogenic variant in the same gene, phase confirmed by testing parents or offspring, in a patient with colorectal cancer after age 45 or another Lynch syndrome cancer above the relevant median age of onset, and no clinical manifestations of CMMRD. No such case-level evidence is present in the consolidated bundle.
Still missing
A report of confirmed parental- or offspring-established in-trans phase with a known pathogenic MLH1 variant, colorectal cancer after age 45 (or another Lynch syndrome cancer above its Lynch syndrome median age of onset), and no previous or current clinical manifestations of CMMRD is required.
BS3Not assessed: no calibrated functional assay demonstrated normal or proficient MLH1 function for this variant.
Details ▾Hide ▴
Reasoning
The InSiGHT MLH1 VCEP v2.0 specification requires BS3 be supported by calibrated functional assays showing a functional odds-for-pathogenicity <=0.48 (Strong <=0.05, Supporting >0.05-0.48), or variant-specific proficient function on the MMR functional assay flowchart. No calibrated functional assay or RNA/protein study addressing the exact variant c.1732-2A>G was found in the case bundle, the VCEP-approved assay documentation, or the reviewed publications; therefore there is no evidence of normal/proficient function to support BS3.
Still missing
No functional assay data (calibrated or otherwise) exists for c.1732-2A>G to evaluate a benign/proficient-function functional odds value.
This is a canonical splice-acceptor (IVS-2) position with a very high SpliceAI acceptor-loss delta (0.996), which is inconsistent with a benign functional profile, though bioinformatic prediction alone does not satisfy or refute BS3 under this VCEP specification.
If a validated RNA/mRNA or protein-based assay directly testing this allele demonstrates normal splicing/function, BS3 should be re-evaluated against the VCEP odds-for-pathogenicity thresholds.
BS4Not assessed: no non-segregation observations or pedigree likelihood data were available.
Details ▾Hide ▴
Reasoning
No non-segregation observations or combined pedigree Bayes likelihood ratio are provided for the exact c.1732-2A>G variant. BS4 cannot be assigned.
Rule
MLH1 InSiGHT v2.0 BS4: combined pedigree Bayes likelihood ratio <0.05 is Strong; >0.05 to 0.48 is Supporting.
Evidence considered
The MLH1 InSiGHT specification requires a combined pedigree Bayes likelihood ratio below 0.05 for BS4 Strong or greater than 0.05 through 0.48 for BS4 Supporting.
No case-specific non-segregation data or Bayes likelihood ratio is present in the consolidated bundle.
Still missing
Pedigree(s) documenting affected non-carriers and/or unaffected carriers with ages and relevant cancer phenotypes.
Combined overall Bayes likelihood ratio demonstrating lack of co-segregation.
BP5Not assessed: no evidence of an alternative molecular basis, such as MLH1 promoter methylation or BRAF V600E-positive tumors, was available.
Details ▾Hide ▴
Reasoning
No tumor-level evidence was provided for an alternative molecular basis under the MLH1 VCEP BP5 framework. In particular, there are no documented qualifying MSS tumors, retained/discordant MMR IHC results, BRAF V600E-positive colorectal tumors, or MLH1-methylated Lynch-spectrum tumors.
Rule
MLH1 InSiGHT VCEP BP5: Supporting requires 2-3 qualifying tumors or one qualifying BRAF V600E/MLH1-methylated tumor; Strong requires at least 4 qualifying tumors or at least 2 qualifying BRAF V600E/MLH1-methylated tumors.
Evidence considered
The MLH1 InSiGHT BP5 specification requires two or three qualifying tumors for Supporting strength, or at least four qualifying tumors for Strong strength; one BRAF V600E-positive colorectal tumor or MLH1-methylated Lynch-spectrum tumor with MSI-H/MLH1 loss can support BP5.
Still missing
Tumor reports documenting qualifying MSS and/or retained or gene-inconsistent MMR IHC in two or more CRC/endometrial or Lynch-spectrum tumors, or a BRAF V600E-positive colorectal tumor or MLH1-methylated Lynch-spectrum tumor with MSI-H/MLH1 loss.
BP6Not assessed: no expert-panel benign classification exists; the only panel assertion is InSiGHT Likely pathogenic.
Details ▾Hide ▴
Reasoning
No exact-variant ClinVar expert-panel Benign or Likely benign classification is present. The only expert-panel assertion is InSiGHT Likely pathogenic; ordinary laboratory submissions and aggregate labels were not used for BP6.
Rule
BP6 Supporting: exact-variant ClinVar expert-panel Benign or Likely benign classification.
Evidence considered
ClinVar variation 89861 has an exact-variant InSiGHT expert-panel assertion of Likely pathogenic and no expert-panel Benign/Likely benign assertion.
Still missing
An exact-variant ClinVar expert-panel Benign or Likely benign classification would be required to apply BP6 under the requested rule.
Not applicable · 13· recorded as not applicableShow reasons ▾Hide reasons ▴PS1N/ANot applicable: this is a canonical splice-acceptor variant with no amino-acid change, outside both PS1 branches (missense or non-canonical splice nucleotide).
Details ▾Hide ▴
Reasoning
NM_000249.4:c.1732-2A>G is a canonical splice acceptor variant (IVS position -2, protein consequence NP_000240.1:p.?) with no defined amino-acid change; it is not a missense substitution. The InSiGHT MLH1 VCEP Specification v2.0 PS1 rule has two branches: (1) a missense substitution encoding the same amino acid change via a different nucleotide change as a VCEP-established Pathogenic missense variant, or (2) a variant affecting the SAME non-canonical splice nucleotide as a confirmed Pathogenic splice variant with similar/worse SpliceAI prediction. Branch 1 requires a missense change, which this variant does not have. Branch 2 requires a non-canonical splice nucleotide; c.1732-2A>G occurs at the canonical AG splice-acceptor dinucleotide (-1/-2 position), which is explicitly outside the scope of the 'non-canonical splice nucleotide' PS1 branch (that branch is intended for splice-region positions outside the invariant +1,+2/-1,-2 dinucleotides). This variant is evaluated instead under the VCEP PVS1 canonical-splice framework, not PS1.
MLH1 InSiGHT VCEP Specification v2.0 PS1 rule text: 'A predicted missense substitution that encodes the same amino acid change...' and 'Variants affecting the same non-canonical splice nucleotide as a confirmed Pathogenic splice variant...' Neither branch applies because c.1732-2A>G is a canonical (not non-canonical) splice-acceptor position and produces no missense change (protein_1l/protein_3l = NP_000240.1:p.?).
Still missing
PS1 governing rule (both branches) requires either a missense same-amino-acid-change comparator or a non-canonical splice nucleotide match; neither applies to this canonical splice acceptor variant.
PS4VCEPNot applicable: the MLH1 VCEP evaluates tumor immunohistochemistry through PP4 instead of PS4.
Details ▾Hide ▴
Reasoning
The applicable MLH1 InSiGHT VCEP specification does not use PS4, because tumor IHC data are available for MMR variant classification through PP4 rather than proband counting.
VCEP rule
MLH1 InSiGHT VCEP: PS4 is not applicable; tumor IHC evidence is assessed under PP4.
Evidence considered
The MLH1 InSiGHT specification marks PS4 not applicable and states that it has not been used for MMR variant classification using proband counting because tumor IHC data are addressed by PP4.
PM1VCEPNot applicable: the MLH1 VCEP defines no mutational-hotspot PM1 rule for this gene.
Details ▾Hide ▴
Reasoning
The InSiGHT MLH1 VCEP Specification v2.0 explicitly marks PM1 as 'Not Applicable' for this gene (no mutational hotspot/domain-based PM1 rule is defined in the CSPEC). No generic ACMG fallback is used because a governing VCEP specification exists and explicitly addresses this criterion.
VCEP rule
InSiGHT MLH1 VCEP Specification v2.0 (cspec) - PM1 Not Applicable
Evidence considered
case_summary.json cspec.criteria array lists {"label": "PM1", "applicability": "Not Applicable"} for the MLH1 InSiGHT VCEP Specification v2.0.
PM4VCEPNot applicable: the MLH1 VCEP does not use PM4, and this splice variant has no defined protein length change.
Details ▾Hide ▴
Reasoning
The InSiGHT MLH1 CSPEC v2.0 explicitly designates PM4 as Not Applicable for MLH1 ('Protein length change from an in-frame variant is not used due to lack of evidence'). This variant is also mechanistically not a candidate for PM4 regardless of that gene-specific exclusion: c.1732-2A>G is a canonical splice acceptor substitution, not an observed in-frame insertion/deletion or stop-loss variant with a directly determined protein length change, so the criterion's underlying evidentiary basis (a confirmed in-frame protein length alteration) is not met.
VCEP rule
InSiGHT MLH1 CSPEC v2.0: PM4 not applicable for this VCEP.
Evidence considered
InSiGHT MLH1 CSPEC v2.0: PM4 listed as 'Not applicable' for all strength tiers, with instructionsToUse stating protein length change from an in-frame variant is not used for MLH1 due to lack of evidence.
The variant is a splice acceptor substitution (c.1732-2A>G) rather than an in-frame indel or stop-loss variant, so it does not present the protein-length-change scenario PM4 addresses even outside the gene-specific exclusion.
PM5N/ANot applicable: no amino-acid substitution results (p.?), so there is no residue to compare against other missense changes.
Details ▾Hide ▴
Reasoning
PM5 under the MLH1 VCEP requires a missense change at an amino-acid residue where a different missense change at that same residue was classified Pathogenic/Likely Pathogenic on the protein level. c.1732-2A>G is a canonical splice acceptor variant with no resulting amino-acid substitution (protein_1l/protein_3l = NP_000240.1:p.?; residue.reference_aa/position/observed_alt_aa are all null), so there is no residue to compare against other missense variants. The dedicated PM5 comparator search artifact confirms this variant is ineligible for the classic same-residue missense PM5 search.
pm5_candidates.json: eligible_for_classic_pm5_search=false, reason='Framework is classic same-residue missense PM5, but the variant consequence class (canonical_splice) is not missense.' recommendation.level='not_applicable'.
Still missing
Not a missense variant; no protein residue exists to search for same-residue Pathogenic/Likely Pathogenic missense comparators.
Sources
pm5_candidates
PM6VCEPNot applicable: the MLH1 VCEP does not use PM6; unconfirmed de novo observations are handled through PS2.
Details ▾Hide ▴
Reasoning
PM6 is explicitly not applicable in the MLH1 InSiGHT specification; unconfirmed de novo observations, if available, are handled through the PS2 point system instead.
VCEP rule
MLH1 InSiGHT v2.0 PM6: not applicable; see PS2.
Evidence considered
The MLH1 InSiGHT specification marks PM6 as not applicable and directs assessors to PS2.
PP2VCEPNot applicable: the MLH1 VCEP does not use PP2, and this is not a missense variant.
Details ▾Hide ▴
Reasoning
The InSiGHT MLH1 VCEP Specification v2.0 explicitly marks PP2 as 'Not Applicable' for this gene. PP2 (missense-in-a-gene-with-low-benign-missense-rate) is also not appropriate for c.1732-2A>G since it is a canonical splice variant with no missense consequence.
VCEP rule
InSiGHT MLH1 VCEP Specification v2.0 (cspec) - PP2 Not Applicable
Evidence considered
case_summary.json cspec.criteria array lists {"label": "PP2", "applicability": "Not Applicable"} for the MLH1 InSiGHT VCEP Specification v2.0.
PP3N/ANot applicable: PP3 is restricted to non-canonical splice sites and missense variants; canonical splice disruption is scored under PVS1 without double-counting the same splice evidence.
Details ▾Hide ▴
Reasoning
NM_000249.4:c.1732-2A>G disrupts the canonical -2 position of the MLH1 splice acceptor consensus (canonical_splice bucket per pvs1_variant_assessment), not a non-canonical splice nucleotide. The governing InSiGHT MMR VCEP PP3 rule text explicitly restricts SpliceAI-based PP3 to 'non-canonical splicing nucleotides' and to missense substitutions scored via the HCI-priors lookup; a canonical +/-1,2 splice-consensus variant is evaluated under the VCEP's PVS1 pathway instead. The case's own pvs1_variant_assessment artifact states directly: 'For canonical splice variants, do not stack PP3 for the same splice-effect prediction evidence' and 'PP3 should not be double-counted for the same splice prediction evidence.' The variant is also not missense (protein_1l = p.?), so the HCI-prior missense pathway for PP3 does not apply. PP3 is therefore not applicable for this variant under the governing VCEP framework.
Rule
InSiGHT MMR VCEP PP3: Missense variant with HCI prior probability for pathogenicity >0.68 & <=0.81, OR predicted splice defect for non-canonical splicing nucleotides using SpliceAI with delta score >= 0.2 (Walker et al 2023). Not applicable to canonical +/-1,2 splice-consensus variants, which fall under this VCEP's PVS1 pathway instead.
Evidence considered
cspec VCEP PP3 rule (label PP3): 'Missense variant with HCI prior probability for pathogenicity >0.68 & <=0.81... OR Predicted splice defect for non-canonical splicing nucleotides using SpliceAI with delta score >= 0.2 as per Walker et al 2023' - this variant is neither missense nor a non-canonical splice nucleotide.
pvs1_variant_assessment classifies the variant as consequence_class 'canonical_splice' at the -2 acceptor position and instructs: 'For canonical splice variants, do not stack PP3 for the same splice-effect prediction evidence.'
SpliceAI lookup (source key spliceai) shows DS_AG=0.732, DS_AL=0.996, max_delta_score=0.996, confirming a canonical acceptor-disrupting effect already captured by the PVS1 pathway rather than an independent non-canonical splice prediction eligible for PP3.
BP1VCEPNot applicable: the MLH1 VCEP does not use BP1, and this is a splice variant, not a missense change.
Details ▾Hide ▴
Reasoning
The InSiGHT MLH1 VCEP Specification v2.0 explicitly marks BP1 as 'Not Applicable' for this gene. BP1 (missense variant in a gene where only truncating variants cause disease) is also not appropriate here since the variant is a canonical splice acceptor change, not a missense substitution.
VCEP rule
InSiGHT MLH1 VCEP Specification v2.0 (cspec) - BP1 Not Applicable
Evidence considered
case_summary.json cspec.criteria array lists {"label": "BP1", "applicability": "Not Applicable"} for the MLH1 InSiGHT VCEP Specification v2.0.
BP2VCEPNot applicable: the MLH1 VCEP does not use BP2, directing use of BS2 instead.
Details ▾Hide ▴
Reasoning
BP2 is explicitly not applicable in the MLH1 InSiGHT VCEP specification; its instructions direct use of BS2 instead. BP2 is therefore not evaluated under the governing gene-specific framework.
VCEP rule
MLH1 InSiGHT VCEP: BP2 not applicable; use BS2 instead.
Evidence considered
The MLH1 InSiGHT VCEP specification marks BP2 as not applicable and states: 'BS2 is used instead.'
BP3VCEPNot applicable: the MLH1 VCEP does not use BP3, and this is not an in-frame deletion/insertion.
Details ▾Hide ▴
Reasoning
The InSiGHT MLH1 CSPEC v2.0 explicitly designates BP3 as Not Applicable for MLH1. This variant is additionally not a candidate for BP3 on its own merits: it is a canonical splice acceptor substitution rather than an in-frame deletion/insertion in a repetitive/non-functional region, so it does not meet the underlying scenario BP3 is designed to address.
VCEP rule
InSiGHT MLH1 CSPEC v2.0: BP3 not applicable for this VCEP.
Evidence considered
InSiGHT MLH1 CSPEC v2.0 case_summary criteria listing: BP3 applicability = 'Not Applicable' for this VCEP.
The variant is a splice acceptor substitution, not an in-frame indel in a repeat region, so it is mechanistically outside the scope BP3 addresses regardless of the gene-specific exclusion.
BP4N/ANot applicable: BP4 covers missense or non-canonical splice variants with SpliceAI delta <=0.1; this canonical splice variant scores 0.996.
Details ▾Hide ▴
Reasoning
The governing InSiGHT MMR VCEP BP4 rule applies only to (a) missense variants via the HCI-priors lookup, or (b) intronic/synonymous variants at non-canonical splicing nucleotides with SpliceAI delta score <=0.1. c.1732-2A>G is neither missense (protein_1l = p.?) nor a non-canonical splice-site variant - it disrupts the canonical -2 acceptor position (consequence_class 'canonical_splice'). In addition, SpliceAI predicts a strong splice-disrupting effect (max delta = 0.996), the opposite of the 'no splicing impact' pattern BP4 requires, so BP4 would not be met on the evidence even if the rule pathway applied. BP4 is not applicable for this variant.
Rule
InSiGHT MMR VCEP BP4: Missense variant with HCI-prior probability of pathogenicity <0.11, OR for intronic and synonymous variants SpliceAI predicts no splicing impact with delta score <=0.1 (Walker et al 2023). Not applicable to canonical splice-consensus variants or to variants with a high SpliceAI delta score.
Evidence considered
cspec VCEP BP4 rule (label BP4): 'Missense variant with HCI-prior probability of pathogenicity <0.11... OR For intronic and synonymous variants: SpliceAI predicts no splicing impact with delta score <= 0.1 as per Walker et al 2023' - variant is canonical splice, not missense/intronic-non-canonical/synonymous, and SpliceAI does not predict 'no impact'.
SpliceAI lookup (spliceai) reports DS_AG=0.732, DS_AL=0.996, max_delta_score=0.996 for this canonical acceptor position, inconsistent with the 'no splicing impact' pattern BP4 requires.
BP7N/ANot applicable: BP7 covers synonymous or deep-intronic variants at or beyond -21/+7, not a canonical -2 splice position.
Details ▾Hide ▴
Reasoning
The governing VCEP BP7 rule applies to 'a synonymous (silent) or intronic variant at or beyond -21/+7 (5'/3' exonic)'. NM_000249.4:c.1732-2A>G sits at intronic position -2, well within the canonical splice-consensus region and far from the -21/+7 deep-intronic/near-exonic boundary the rule requires. It is also not a synonymous variant. BP7 is not applicable to this variant regardless of its computational splice-impact score.
Rule
InSiGHT MMR VCEP BP7: A synonymous or intronic variant at or beyond -21/+7 (5'/3' exonic).
Evidence considered
cspec VCEP BP7 rule (label BP7): 'A synonymous (silent) or intronic variant at or beyond -21/+7 (5'/3' exonic). Variants may satisfy both BP7 and BP4.' The variant's position (-2) is inside the splice-consensus region, not at or beyond -21/+7, and it is not synonymous.
Every interpreted variant in MLH1, mapped onto the exon structure of NM_000249.4. Select a row to find it on the plot.
MLH1NM_000249.4
Fetching transcript structure from UCSC…
Exons
—
Transcript span
—
Strand
—
Variants mapped
—
Source
UCSC ncbiRefSeqCurated
All variants in MLH1—Select a row to find it on the plot; use the link to open its page.
Protein
Location
Classification
Link
gnomAD v4.1Allele count data not recorded
gnomAD v2.1Allele count data not recorded
gnomAD-Canada v1.0Allele count data not recorded
Statements
gnomAD v4.1
Absent from gnomAD v4.1.
gnomAD v2.1
Absent from gnomAD v2.1.
gnomAD-Canada
Absent from gnomAD-Canada v1.0.
SpliceAI and Pangolin
SpliceAI acceptor lossat +2 bp
0.996
SpliceAI acceptor gainat +42 bp
0.732
SpliceAI donor lossat +166 bp
0.114
SpliceAI donor gainat +46 bp
0.000
Pangolin splice lossmagnitude shown
-0.893
Pangolin splice gain
0.592
Other predictors
REVEL
Not available
BayesDel
0.660
How the predictions were used
PP3N/ANot applicable: PP3 is restricted to non-canonical splice sites and missense variants; canonical splice disruption is scored under PVS1 without double-counting the same splice evidence.
Details ▾Hide ▴
Reasoning
NM_000249.4:c.1732-2A>G disrupts the canonical -2 position of the MLH1 splice acceptor consensus (canonical_splice bucket per pvs1_variant_assessment), not a non-canonical splice nucleotide. The governing InSiGHT MMR VCEP PP3 rule text explicitly restricts SpliceAI-based PP3 to 'non-canonical splicing nucleotides' and to missense substitutions scored via the HCI-priors lookup; a canonical +/-1,2 splice-consensus variant is evaluated under the VCEP's PVS1 pathway instead. The case's own pvs1_variant_assessment artifact states directly: 'For canonical splice variants, do not stack PP3 for the same splice-effect prediction evidence' and 'PP3 should not be double-counted for the same splice prediction evidence.' The variant is also not missense (protein_1l = p.?), so the HCI-prior missense pathway for PP3 does not apply. PP3 is therefore not applicable for this variant under the governing VCEP framework.
Rule
InSiGHT MMR VCEP PP3: Missense variant with HCI prior probability for pathogenicity >0.68 & <=0.81, OR predicted splice defect for non-canonical splicing nucleotides using SpliceAI with delta score >= 0.2 (Walker et al 2023). Not applicable to canonical +/-1,2 splice-consensus variants, which fall under this VCEP's PVS1 pathway instead.
Evidence considered
cspec VCEP PP3 rule (label PP3): 'Missense variant with HCI prior probability for pathogenicity >0.68 & <=0.81... OR Predicted splice defect for non-canonical splicing nucleotides using SpliceAI with delta score >= 0.2 as per Walker et al 2023' - this variant is neither missense nor a non-canonical splice nucleotide.
pvs1_variant_assessment classifies the variant as consequence_class 'canonical_splice' at the -2 acceptor position and instructs: 'For canonical splice variants, do not stack PP3 for the same splice-effect prediction evidence.'
SpliceAI lookup (source key spliceai) shows DS_AG=0.732, DS_AL=0.996, max_delta_score=0.996, confirming a canonical acceptor-disrupting effect already captured by the PVS1 pathway rather than an independent non-canonical splice prediction eligible for PP3.
BP4N/ANot applicable: BP4 covers missense or non-canonical splice variants with SpliceAI delta <=0.1; this canonical splice variant scores 0.996.
Details ▾Hide ▴
Reasoning
The governing InSiGHT MMR VCEP BP4 rule applies only to (a) missense variants via the HCI-priors lookup, or (b) intronic/synonymous variants at non-canonical splicing nucleotides with SpliceAI delta score <=0.1. c.1732-2A>G is neither missense (protein_1l = p.?) nor a non-canonical splice-site variant - it disrupts the canonical -2 acceptor position (consequence_class 'canonical_splice'). In addition, SpliceAI predicts a strong splice-disrupting effect (max delta = 0.996), the opposite of the 'no splicing impact' pattern BP4 requires, so BP4 would not be met on the evidence even if the rule pathway applied. BP4 is not applicable for this variant.
Rule
InSiGHT MMR VCEP BP4: Missense variant with HCI-prior probability of pathogenicity <0.11, OR for intronic and synonymous variants SpliceAI predicts no splicing impact with delta score <=0.1 (Walker et al 2023). Not applicable to canonical splice-consensus variants or to variants with a high SpliceAI delta score.
Evidence considered
cspec VCEP BP4 rule (label BP4): 'Missense variant with HCI-prior probability of pathogenicity <0.11... OR For intronic and synonymous variants: SpliceAI predicts no splicing impact with delta score <= 0.1 as per Walker et al 2023' - variant is canonical splice, not missense/intronic-non-canonical/synonymous, and SpliceAI does not predict 'no impact'.
SpliceAI lookup (spliceai) reports DS_AG=0.732, DS_AL=0.996, max_delta_score=0.996 for this canonical acceptor position, inconsistent with the 'no splicing impact' pattern BP4 requires.
BP7N/ANot applicable: BP7 covers synonymous or deep-intronic variants at or beyond -21/+7, not a canonical -2 splice position.
Details ▾Hide ▴
Reasoning
The governing VCEP BP7 rule applies to 'a synonymous (silent) or intronic variant at or beyond -21/+7 (5'/3' exonic)'. NM_000249.4:c.1732-2A>G sits at intronic position -2, well within the canonical splice-consensus region and far from the -21/+7 deep-intronic/near-exonic boundary the rule requires. It is also not a synonymous variant. BP7 is not applicable to this variant regardless of its computational splice-impact score.
Rule
InSiGHT MMR VCEP BP7: A synonymous or intronic variant at or beyond -21/+7 (5'/3' exonic).
Evidence considered
cspec VCEP BP7 rule (label BP7): 'A synonymous (silent) or intronic variant at or beyond -21/+7 (5'/3' exonic). Variants may satisfy both BP7 and BP4.' The variant's position (-2) is inside the splice-consensus region, not at or beyond -21/+7, and it is not synonymous.
SpliceAI predicts possible splice impact for this variant (max delta score = 1.00). BayesDel score = 0.66.
8 paper records listed. No publication is explicitly linked to an applied criterion in the recorded metadata. None of the listed paper records explicitly indicates that the variant is named.
Recorded publications
Systematic mRNA analysis for the effect of MLH1 and MSH2 missense and silent mutations on aberrant splicing.
Methylation Tolerance-Based Functional Assay to Assess Variants of Unknown Significance in the MLH1 and MSH2 Genes and Identify Patients With Lynch Syndrome.
The evidence statements recorded by the interpretation pipeline. They have not been reinterpreted by the page.
gnomAD v4.1
Absent from gnomAD v4.1.
gnomAD v2.1
Absent from gnomAD v2.1.
gnomAD-Canada
Absent from gnomAD-Canada v1.0.
ClinVar
This variant has been reported in ClinVar as Pathogenic (2 clinical laboratories) and as likely pathogenic (1 clinical laboratory) and as Likely pathogenic (1 clinical laboratory) and as Likely pathogenic by International Society for Gastrointestinal Hereditary Tumours (InSiGHT) (expert panel). (ClinVarID = 89861)
In silico
SpliceAI predicts possible splice impact for this variant (max delta score = 1.00). BayesDel score = 0.66.
Somatic
This variant has not previously been reported in somatic cancers (COSMIC).
This variant was interpreted by pipeline 7.1.5.
The current version is 11.0.07.
Re-running queues a fresh interpretation; the result replaces this page when complete.
Re-runs are not free. Blank draws on today's public interpretation allowance; with a code, it uses one of that code's uses.