Resources for CNV variant interpretation

Determining a dose-sensitive gene or region

The ClinGen Dosage Sensitivity working group collects evidence supporting/refuting the haploinsufficiency and triplosensitivity of genes and genomic regions, making the data publicly available and indicating dosage sensitivity rating for each given gene or region.

Table 1. Rate of dosage sensitivity and its possible clinical interpretation.

DECIPHER also provides a list of curated microdeletion and microduplication syndromes involved in developmental disorders.

Curated haploinsufficiency/triplosensitivity lists are not exhaustive and other gene/disease association resources can be used. The ACGS Best Practice Guidelines for variant classification in rare disease 2024 recommend using additional resources for determining an established pathogenic dosage-sensitive gene/region such as ClinGen gene-disease validity with a ‘definitive’ or ‘strong’ status and Gene2Phenotype (G2P) genes with a ‘definitive’ or ‘strong’ status. Note, that the mechanism of disease must be curated as loss-of-function.

If the observed CNV does not overlap a known dose-sensitive region or gene, haploinsufficiency prediction scores may provide evidence for pathogenicity for deletions. The CNV loss guidelines state that two or more predictors must suggest that at least one gene in the interval is haploinsufficient. Common predictors include pLI, LOEUF, sHET and pHaplo.

Determining a benign gene or region

For determining if a gene/region is benign, the following resources are available:

• gnomAD Structural Variants: population structural variants from gnomAD.
• Database of Genomic Variants (DGV) Gold: DGV Gold Standard Variants are a select number of high-resolution and high-quality studies in DGV, which is a curated catalogue of structural variation found in the genomes of control individuals.
• DECIPHER’s Population Copy-Number Variants dataset: the CNV consensus set consists of a number of carefully selected, high-resolution control sets.

Defining the number of protein-coding genes within the CNV

There are multiple resources which can be used to determine the number of protein-coding genes within a genomic region. These include DECIPHER and the CNV-ClinViewer.

Assessing similar cases

A detailed evaluation of cases with similar genomic content is important evidence for pathogenicity classification, although not required if the observed CNV overlaps a known dose-sensitive gene or region. For CNV gains, cases with a complete duplication of one or more genes within the observed CNV, or cases with an overlapping CNV gain similar in genomic content to the observed CNV gain, can be counted. For CNV losses, cases with a complete deletion or a loss of function variant within a gene encompassed by the observed CNV or cases with an overlapping CNV loss similar in genomic content to the observed copy-number loss can be counted. The point scoring for each case depends on the inheritance of the CNV and the specificity of the phenotype. Case data from published literature, public databases, and/or internal lab data can be used, including from the following resources:

• ClinVar Structural Variants
• Patient CNV data shared through DECIPHER

When assessing a CNV identified in an individual whose ancestry is not well represented in global genomic datasets, the use of ancestry-matched and/or internal lab data will increase interpretation accuracy.

CNV classification calculators

Multiple resources provide CNV classification calculators that automatically calculate the final point value and corresponding CNV classification, once the appropriate evidence and score for each evidence criterion have been selected.

• ClinGen provides a web-based CNV classification calculator which is publicly available.
• DECIPHER’s CNV classification interface is available to registered users. Registering with DECIPHER is free and is available to individuals who work at academic centres and intend to share consented patient data through the platform.