DICER1 c.2536A>G, p.(Ile846Val)

Reasoning

Per ClinGen DICER1 VCEP CSpec v1.4, BP4 (Supporting) applies to missense variants with REVEL < 0.500 and agreement in splicing predictors that no splicing effect is predicted. Local REVEL score is 0.191, well under the 0.500 threshold. SpliceAI predicts no significant splice impact for this variant (max delta score = 0.029, far below the actionable ~0.2 range), consistent with 'no splicing effects predicted.' MaxEntScan output was not separately computed in this case bundle, but SpliceAI (the VCEP's primary splice-impact tool, applicable per the VCEP's own instructionsToUse for scores outside/without a MaxEntScan-flagged region) shows no splicing signal, and this missense substitution does not fall in a canonical splice region requiring MES cross-check. Both REVEL and splicing-predictor prongs of the VCEP BP4 rule are satisfied; BP4 Supporting applies.

Rule

DICER1 VCEP CSpec v1.4 BP4: missense REVEL < 0.500 and splicing-predictor agreement on no splicing effect

Evidence considered

• DICER1 VCEP CSpec v1.4 BP4 rule: 'For missense variants, REVEL score < 0.500 and agreement in splicing predictors that no splicing effects are predicted.'
• Local REVEL v1.3 lookup: score = 0.191 for NM_177438.2:c.2536A>G, below the 0.500 benign-supporting threshold specified by the VCEP.
• SpliceAI Lookup: max delta score = 0.029 (DS_AG=0.002, DS_AL=0.029, DS_DG=0.0, DS_DL=0.008), indicating no predicted splicing impact, consistent with the 'no splicing effects predicted' component of the VCEP BP4 rule.
• REVEL score calibration for the 0.5 boundary is supported by the ClinGen SVI-endorsed calibration in Pejaver et al. 2022 (PMID 36413997).

Sources

Reasoning

The DICER1 VCEP specifies PM1 applies only to missense variants within the RNase IIIb domain (p.Y1682-p.S1846): Moderate for the seven metal ion-binding hotspot codons (p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, p.E1813) and Supporting for other RNase IIIb domain residues. The variant under evaluation, p.(Ile846Val), affects residue 846, which lies well outside the RNase IIIb domain (1682-1846) and is not one of the metal-binding hotspot codons. No VCEP-recognized structural/functional hotspot or domain criterion is met at this residue.

Rule

DICER1 VCEP v1.4 PM1: Moderate for metal ion-binding hotspot codons; Supporting for other RNase IIIb domain residues (p.Y1682-p.S1846).

Evidence considered

• DICER1 cspec v1.4 PM1 rule: 'Putative missense variants at residues affecting metal ion-binding: codons p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, p.E1813' (Moderate) and 'Putative missense variants at residues in the RNase IIIb domain (p.Y1682-p.S1846), besides the metal ion-binding residues' (Supporting).
• Variant residue is p.Ile846, numerically and structurally outside the RNase IIIb domain boundary (p.Y1682-p.S1846) and not among the listed hotspot codons.

Sources

Reasoning

PM2 is not met. The DICER1 VCEP requires allele frequency <0.000005 across gnomAD, no more than one allele in any subpopulation, and at least 20x coverage. gnomAD v4.1 reports 50/1,613,432 total alleles (AF 3.0990e-05), with 3/60,004 alleles in the highest-frequency subpopulation; these observations exceed both the overall frequency and per-subpopulation allele-count requirements.

Rule

Allele frequency <0.000005 across gnomAD with no more than one allele in any subpopulation and at least 20x coverage.

Evidence considered

• The ClinGen DICER1 VCEP specification version 1.4 defines PM2 Supporting as allele frequency <0.000005 across gnomAD, with no more than one allele in any subpopulation and at least 20x coverage; it instructs reviewers to use the most recent/most comprehensive gnomAD version in general.
• gnomAD v4.1 reports 50 alternate alleles among 1,613,432 total alleles (AF 3.09898e-05), 3 alternate alleles among 60,004 Admixed American alleles, and zero homozygotes; the frequency and allele-count observations fail the PM2 rule.
• gnomAD v2.1 independently reports 10 alternate alleles among 251,156 total alleles (AF 3.98159e-05) and 3 alternate alleles among 34,592 Admixed American alleles, also inconsistent with PM2.

Sources

Reasoning

Variant is missense (NP_803187.1:p.(Ile846Val)). Per ClinGen DICER1 VCEP CSpec v1.4, PP3 (Supporting) requires REVEL >= 0.750 OR agreement in splicing predictors predicting a splicing effect. Local REVEL lookup gives 0.191, far below the 0.750 threshold. SpliceAI predicts no significant splice impact (max delta score = 0.029, well below the conventional 0.2 SpliceAI actionable threshold), so there is no splicing-predictor support for a deleterious effect either. Neither prong of the VCEP PP3 rule is satisfied.

Rule

DICER1 VCEP CSpec v1.4 PP3: missense REVEL >= 0.750 OR splicing-predictor agreement on splicing effect

Evidence considered

• DICER1 VCEP CSpec v1.4 PP3 rule: 'For missense variants, REVEL score >= 0.750 OR agreement in splicing predictors predict splicing effects.'
• Local REVEL v1.3 lookup: score = 0.191 for NM_177438.2:c.2536A>G (chr14 zip lookup), well below the 0.750 pathogenic-supporting threshold specified by the VCEP.
• SpliceAI Lookup (broadinstitute) scores: DS_AG=0.002, DS_AL=0.029, DS_DG=0.0, DS_DL=0.008; max delta score = 0.029, indicating no predicted splicing effect.

Sources

Reasoning

BA1 is not met. The DICER1 VCEP requires a gnomAD subpopulation frequency greater than 0.003, with more than 2,000 alleles tested and at least 5 alternate alleles. In the most recent/comprehensive available dataset, gnomAD v4.1, the highest subpopulation frequency is 4.9997e-05 in Admixed American (3/60,004 alleles), far below 0.003.

Rule

Frequency >0.003 (0.3%) in gnomAD subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.

Evidence considered

• The ClinGen DICER1 VCEP specification version 1.4 defines BA1 as frequency >0.003 in a gnomAD subpopulation, requiring >2,000 alleles tested and at least 5 alleles present.
• gnomAD v4.1 reports the highest observed subpopulation frequency as 4.9997e-05 in Admixed American, based on 3 alternate alleles among 60,004 tested alleles; this is below the BA1 threshold.

Sources

Reasoning

BS1 is not met. The DICER1 VCEP requires a frequency greater than 0.0003 in a gnomAD subpopulation, with more than 2,000 alleles tested and at least 5 alternate alleles. The highest gnomAD v4.1 subpopulation frequency is 4.9997e-05 in Admixed American (3/60,004 alleles), and therefore does not reach the threshold. The gnomAD v2.1 subpopulation maximum is also below 0.0003.

Rule

Frequency >0.0003 (0.03%) in gnomAD subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.

Evidence considered

• The ClinGen DICER1 VCEP specification version 1.4 defines BS1 as frequency >0.0003 in a gnomAD subpopulation, requiring >2,000 alleles tested and at least 5 alleles present.
• gnomAD v4.1 reports a maximum subpopulation frequency of 4.9997e-05 in Admixed American, based on 3 alternate alleles among 60,004 tested alleles; this is below the BS1 threshold.
• gnomAD v2.1 reports a maximum subpopulation frequency of 1.3965e-04 in the AMR_XY subgroup, based on 2 alternate alleles among 14,322 tested alleles; this is also below 0.0003 and does not meet the minimum of 5 alternate alleles.

Sources

Reasoning

PS1 (DICER1 VCEP v1.4) requires identifying another variant with the same amino acid change (p.Ile846Val) that has been asserted pathogenic (not likely pathogenic) by the ClinGen DICER1 VCEP, with splicing ruled out. The PM5 comparator harvesting artifact (pm5_candidates.json) attempted this search but failed softly with an HTTP 429 (Too Many Requests) error, returning 0 same-residue candidates out of only 3 candidates seen before the search was interrupted. No usable comparator data exists to adjudicate PS1 in either direction.

Rule

DICER1 VCEP v1.4 PS1: For same AA change, must confirm no difference in splicing (RNA data or concordant MaxEntScan/SpliceAI); comparator must be asserted pathogenic (not likely pathogenic) by ClinGen DICER1 VCEP.

Evidence considered

• pm5_candidates.json reports eligible_for_classic_pm5_search=true, candidate_summary.same_residue_candidates=0, total_candidates_seen=3, with errors=['HTTP Error 429: Too Many Requests'] and warning 'PM5 candidate harvesting failed softly'.
• Variant under evaluation is NP_803187.1:p.(Ile846Val), residue I846, well outside the RNase IIIb domain (p.Y1682-p.S1846) defined by the DICER1 VCEP, so any PS1 comparator would also need independent verification of domain/mechanism relevance.

Still missing

• Same-amino-acid-change comparator search (ClinVar for p.Ile846Val classified pathogenic by DICER1 VCEP) was rate-limited (HTTP 429) and did not complete; needs re-run to confirm presence/absence of a qualifying comparator.

Sources

Reasoning

The governing DICER1 specification makes PS2 applicable and requires tallying de novo points from proband-level evidence, with both maternity and paternity confirmed. The case-evidence bundle contains no proband-specific de novo observation, parental genotypes, maternity/paternity confirmation, or de novo point assignment for NM_177438.2:c.2536A>G (p.Ile846Val). Therefore PS2 cannot be assessed.

Rule

De novo points should be tallied using the simplified table for tallying proband points. Strength: Very Strong for ≥4 points; Strong for ≥2 but <4 points; Moderate for ≥1 but <2 points; Supporting for ≥0.5 but <1 point.

Evidence considered

• The DICER1 and miRNA-Processing Gene Specification, version 1.4, states that PS2 is applicable, that de novo points should be tallied using the proband point table, and that the panel uses PS2 exclusively rather than PM6 for de novo evidence.
• The consolidated case-evidence bundle provides variant normalization and general database/literature results but no case-level parental testing, maternity/paternity confirmation, or de novo segregation record for this variant.

Still missing

• Proband-specific report of the variant as de novo is absent.
• Maternal and paternal testing results and confirmation of maternity and paternity are absent.
• No phenotype-qualified de novo point tally is available.

Sources

Reasoning

The DICER1 VCEP specification (v1.4) defines PS3 via either (a) an RNA/splicing assay showing splicing impact of defined severity, or (b) an in vitro pre-miRNA cleavage assay demonstrating failure or severely reduced capacity to produce 5p/3p microRNAs, with positive and negative controls. No such assay data exists for NM_177438.2:c.2536A>G (p.Ile846Val) in this case's evidence bundle: no functional/splicing/cleavage study for this specific variant was retrieved, none of the 5 associated ClinVar-linked PMIDs (24761742, 25394175, 25741868, 28492532, 38084291) is a functional study of this variant (they are guideline/methods/GeneReview papers), and SpliceAI predicts no significant splice impact (max delta 0.029), which is computational, not functional, evidence and does not itself satisfy PS3. The variant is a missense change outside the RNase IIIb domain (p.Y1682-p.S1846), so no domain-specific cleavage-assay literature was identified as applicable. PS3 cannot be assessed for lack of any validated functional assay data on this variant.

Evidence considered

• SpliceAI predicts no significant splice impact for c.2536A>G (max delta score = 0.029), but this is a computational prediction, not a functional assay, and therefore cannot be used to satisfy or refute PS3 per the DICER1 VCEP framework, which requires RNA or in vitro cleavage assay data
• The five PMIDs referenced from ClinVar submissions for this variant (24761742 DICER1 GeneReview, 25394175 ACMG/NSGC referral guideline, 25741868 ACMG/AMP 2015 standards, 28492532 Sherloc classification framework, 38084291 DICER1 VCEP specification paper) were reviewed and none contains a variant-specific functional cleavage or splicing assay for c.2536A>G/p.Ile846Val
• Variant is a missense substitution (p.Ile846Val) located outside the RNase IIIb domain (p.Y1682-p.S1846), so no domain hotspot cleavage-assay data source applies

Still missing

• No in vitro pre-miRNA cleavage assay (5p/3p microRNA production) data available for this specific variant
• No RNA/splicing minigene or patient-derived transcript assay data available for this specific variant
• No functional study identified in the retrieved literature set that examines this variant experimentally

Sources

Reasoning

The DICER1 VCEP specification makes PS4 applicable using phenotype points: supporting for 1–1.5 points, moderate for 2–3.5 points, and strong for at least 4 points. The case bundle contains no validated exact-variant case series, case-control enrichment analysis, or sufficiently detailed unrelated germline proband phenotypes from which phenotype points can be assigned. ClinVar submissions are ordinary single-submitter assertions and do not provide usable phenotype evidence for this criterion. Therefore PS4 cannot be assigned.

Rule

DICER1 VCEP PS4: unrelated probands may contribute up to 1 point each based on phenotype; supporting = 1–1.5 phenotype points, moderate = 2–3.5 points, strong = ≥4 phenotype points. Do not apply if BA1/BS1 criteria are met or if the proband has a qualifying alternative cause or excluded sporadic tumor context.

Evidence considered

• The DICER1 VCEP rules specify PS4 point-based thresholds of 1–1.5 points for supporting, 2–3.5 points for moderate, and at least 4 points for strong evidence.
• The exact-variant ClinVar record is conflicting and single-submitter only, with no expert-panel submission and no sufficiently detailed phenotype dataset suitable for PS4 point assignment.
• The reviewed DICER1 specification paper states that PS4 was adapted into a proband-counting code with variable strength based on affected unrelated probands, but it does not mention NM_177438.2:c.2536A>G or p.Ile846Val.

Still missing

• Exact-variant clinical reports or a case-control study with affected and control counts are needed.
• For each potentially countable unrelated germline proband, the phenotype, genotype, tumor context, and exclusion of an alternative genetic cause are needed to calculate DICER1 VCEP phenotype points.

Sources

Reasoning

PM5 (DICER1 VCEP v1.4) requires identification of a different missense change at the same residue (Ile846) with equal or worse Grantham score, splicing ruled out, and the comparator variant asserted pathogenic (not likely pathogenic) by the ClinGen DICER1 VCEP. The comparator harvesting artifact (pm5_candidates.json) attempted this search but was interrupted by an HTTP 429 rate-limit error and returned 0 same-residue candidates from only 3 candidates seen, explicitly flagging that PM5 cannot be adjudicated from this artifact alone.

Rule

DICER1 VCEP v1.4 PM5: Moderate if a different missense change at the same residue with equal/worse Grantham score is asserted pathogenic by DICER1 VCEP and splicing effect is ruled out; likely pathogenic comparators do not qualify; cannot combine with PM1 or PS1.

Evidence considered

• pm5_candidates.json: residue.position=846, reference_aa=I, observed_alt_aa=V; candidate_summary.same_residue_candidates=0; errors=['HTTP Error 429: Too Many Requests']; recommendation.hermes_instruction states 'do not apply PM5 from this artifact alone'.

Still missing

• Same-residue alternate-amino-acid comparator search (ClinVar for other p.Ile846X changes classified pathogenic by DICER1 VCEP) was rate-limited (HTTP 429) and did not complete; needs re-run to identify any qualifying comparator and compare Grantham scores.

Sources

Reasoning

The governing DICER1 specification makes PP1 applicable, but the case-evidence bundle contains no affected relatives, genotype-positive relatives, family structure, or countable meioses for NM_177438.2:c.2536A>G (p.Ile846Val). Phenotype-positive relatives and segregation data are therefore insufficient to assess PP1.

Rule

Phenotype-positive individuals should have high-, moderate-, or low-specificity phenotypes; do not use segregation of a single low-specificity phenotype across multiple individuals. Strength: Strong for ≥7 meioses across ≥2 families; Moderate for 5–6 meioses across ≥1 family; Supporting for 3–4 meioses across ≥1 family.

Evidence considered

• The DICER1 and miRNA-Processing Gene Specification, version 1.4, requires phenotype-positive individuals and defines PP1 strength by meioses: Supporting for 3–4 meioses, Moderate for 5–6 meioses, and Strong for ≥7 meioses across ≥2 families.
• The consolidated case-evidence bundle contains no family segregation observations or meiosis count for this variant.

Still missing

• No family pedigree or relationship information is provided.
• No genotype results for affected or unaffected relatives are provided.
• No phenotype-qualified affected relatives or countable meioses are provided.

Sources

Reasoning

The DICER1 VCEP restricts PP4 to a specific tumor phenotype pattern: a neoplasm associated with DICER1, a retained germline variant under evaluation, a somatic second hit in one of the specified RNase IIIb hotspot codons, and no additional somatic DICER1 loss-of-function/non-hotspot variant. The bundle does not provide paired tumor sequencing demonstrating that pattern for this variant. A single somatic database observation is not sufficient to establish the required tumor genotype or retention, so PP4 cannot be assigned.

Rule

DICER1 VCEP PP4: supporting when tumor testing in a DICER1-associated neoplasm shows a listed RNase IIIb hotspot somatic second hit and retention of the germline variant under evaluation, with no additional somatic DICER1 non-hotspot variants; not applicable for a germline variant in a listed hotspot codon or when additional somatic variants suggest sporadic tumorigenesis.

Evidence considered

• The DICER1 VCEP PP4 rule requires a somatic RNase IIIb hotspot second hit plus retention of the germline variant, with no additional somatic DICER1 non-hotspot variants consistent with sporadic tumorigenesis.
• The case bundle does not contain the paired tumor genotype, germline-variant retention status, or complete somatic variant profile required by that rule.
• The reviewed DICER1 specification paper describes PP4 as a tumor phenotype code under the DICER1 two-hit model but does not mention NM_177438.2:c.2536A>G or p.Ile846Val.

Still missing

• Paired tumor sequencing or an equivalent validated tumor report demonstrating retention of the germline variant is needed.
• The somatic second-hit allele must be identified as one of the VCEP-listed RNase IIIb hotspot codons, and additional somatic DICER1 variants must be excluded.

Sources

Reasoning

BS2 cannot be fully assessed from the available evidence. gnomAD v4.1 and v2.1 report zero homozygotes for this variant, so the homozygosity routes are not supported. However, the case bundle does not provide the VCEP-required data for 10 or more unrelated females from a single source who are tumor-free through age 50, nor does it establish clinically healthy status and parental confirmation for a homozygous observation.

Rule

40+ unrelated females from a single source are tumor-free through age 50 (caveat: ratio of BS2-eligible females to PS4-eligible probands must be >=40:1) OR 2+ observations of homozygosity in healthy individuals OR 1+ observation(s) of homozygosity in a healthy individual with status confirmed by parental testing. Supporting: 10+ unrelated females from a single source are tumor-free through age 50 (ratio >=10:1) OR 2+ observations of homozygosity in individuals lacking clinical information.

Evidence considered

• The ClinGen DICER1 VCEP specification permits BS2 Strong for 40 or more unrelated females from one source tumor-free through age 50, two or more homozygous observations in healthy individuals, or one homozygous observation in a healthy individual confirmed by parental testing; BS2 Supporting permits 10 or more such females or two homozygous observations in individuals lacking clinical information.
• gnomAD v4.1 reports 0 homozygotes for the variant across the dataset.
• gnomAD v2.1 reports 0 homozygotes for the variant across the dataset, but population database absence of homozygotes does not establish the VCEP-defined healthy-individual or tumor-free-female observations.

Still missing

• No qualifying cohort of at least 10 or 40 unrelated females tumor-free through age 50 from a single source was provided.
• No healthy homozygous individuals with clinical status and, where required, parental testing were provided.

Sources

Reasoning

The DICER1 VCEP specification (v1.4) defines BS3 via either (a) an RNA assay for intronic/synonymous variants showing no splicing impact (observed more than once), or (b) an in vitro cleavage assay demonstrating the variant still produces both 5p and 3p microRNAs from a pre-miRNA, with positive and negative controls, citing Wu et al. 2018 as an exemplar assay. This variant is a missense substitution (not intronic/synonymous), so the RNA-assay-for-splicing rule branch does not apply on its face, and no in vitro cleavage assay data for this specific variant was found anywhere in the case evidence bundle. No functional publication in the retrieved literature set (24761742, 25394175, 25741868, 28492532, 38084291) reports a cleavage assay result for this variant. BS3 cannot be assessed for lack of any validated functional assay data.

Evidence considered

• Variant is a missense substitution, not an intronic or synonymous change, so the BS3 RNA-splicing-assay branch of the DICER1 VCEP rule does not directly apply; only the in vitro cleavage assay branch is potentially relevant, and no such assay data exists for this variant in the case bundle
• The five PMIDs referenced from ClinVar submissions for this variant were reviewed and none reports an in vitro cleavage assay result for c.2536A>G/p.Ile846Val

Still missing

• No in vitro pre-miRNA cleavage assay data (demonstrating normal 5p/3p microRNA production) available for this specific variant
• No functional study identified in the retrieved literature set that examines this variant experimentally

Sources

Reasoning

The governing DICER1 specification makes BS4 applicable when phenotype-positive first-, second-, or third-degree relatives of the proband are genotype-negative. The case-evidence bundle contains no phenotype-positive relatives, relationship information, or genotype-negative familial observations, so lack of segregation cannot be assessed.

Rule

Strong: family members should be phenotype-positive with a high- or moderate-specificity phenotype and genotype-negative as first-, second-, or third-degree relatives of the proband.

Evidence considered

• The DICER1 and miRNA-Processing Gene Specification, version 1.4, requires phenotype-positive, genotype-negative first-, second-, or third-degree relatives of the proband for BS4 at Strong strength.
• The consolidated case-evidence bundle contains no non-segregation observations for this variant.

Still missing

• No affected relatives with high- or moderate-specificity DICER1 phenotypes are documented.
• No genotype-negative relatives and no relationship degrees are documented.
• No family-level evidence addressing phenocopies or another pathogenic variant is documented.

Sources

Reasoning

The DICER1 VCEP permits BP2 at Supporting strength for at least one observation in trans with a P/LP DICER1 variant, or at least three observations in cis or with phase unknown involving two or more different P/LP DICER1 variants. The case-evidence bundle contains no affected-proband, family, co-occurring-variant, phase, cis, or trans observations for NM_177438.2:c.2536A>G (p.Ile846Val), so the threshold cannot be evaluated.

Rule

≥1 observation in trans with P/LP DICER1 variant or ≥3 observations in cis or phase unknown with 2+ different P/LP DICER1 variants.

Evidence considered

• The VCEP BP2 rule requires at least one trans observation with a P/LP DICER1 variant, or at least three cis/phase-unknown observations involving at least two different P/LP DICER1 variants.
• The consolidated case-evidence bundle reports no phase, cis, trans, affected-proband, family, or co-occurring DICER1-variant observations for this variant.

Still missing

• Affected-proband or family-level observations containing the evaluated variant and another P/LP DICER1 variant
• Validated phase information establishing cis or trans, or sufficient observations meeting the VCEP phase-unknown threshold
• Pathogenic/likely pathogenic status of any co-occurring DICER1 variant

Sources