20.02.08 Interpretation of CNVs Flashcards
What is a CNV
- Copy number variant
- Segment of DNA that is present at a variable copy number in comparison with a reference genome. Can be gains or losses.
- Commonly larger than 1kb, although due to advances in technology, now typically >50bp
Categories of CNVs
- Benign= likely to be of no clinical significance.
- Uncertain clinical significance= not previously reported or if reported, insufficient evidence for its classification
- Pathogenic= involve regions of known clinical significance or large changes.
Basics of CNV classification
- Assessment of regions involved in comparison to internal and external databases and literature.
- Comparison to clinical indication of patient
- Parent follow up
Analysis software examples
-Bluefuse or cytosure. Align data with annotation tracks of external and internal databases. Disease and control population datasets.
Why must caution be used for public datasets
- Different platforms used. Identical CNVs could be reported at different sizes.
- Sex of individuals not always given. CNV on X chromosome could be reported as benign in female when pathogenic in males
- Majority of CNVs in large population studies have not been validated
- Factors like incomplete penetrance/variable expressivity, age-of-onset, parent of origin imprinting effects are not considered.
- CNVs reported in multiple publications may represent the same individual.
What database is used for control population data
- DGV (database of genomic variants)
- Curated collection of CNVs.
- Comprises >2.5 million entries from >22,300 genomes
Databases for gene content information
- OMIM (online mendelian inheritance in man)= catalogues all human diseases with their gene content.
- OMIM morbid genes= List of genes linked to a particular disorder
- ENSEMBL= gene database
- HGNC (Hugo gene nomenclature committee)
- Haploinsufficiency scores. Predicted probability of exhibiting haploinsufficiency
- ClinGen dosage sensitivity map. Database of dosage sensitive genes/regions
Databases for pathogenic CNVs
- Decipher (database of chromosome imbalances and phenotypes in humans using ensembl resources). Breakpoints of previously recorded imbalances with phenotypic information.
- ClinVar
- ISCA database
- ECARUCA
- Internal lab databases
Are larger CNVs more likely to be pathogenic?
- Yes.
- Although very large CNVs can be benign. Need to consider the content (gene rich, repetitive elements etc).
Assessing if a CNV is pathogenic
- is the phenotype associated with haploinsufficiency? A gain may have effect
- Disorders due to gain of function variants rather than dosage imbalances (FGFR1 in skeletal dysplasia)
- Gains involving part of a gene may cause disruption
- Dels associated with recessive disease may suggest a mutation on the other allele
- Variants only involving introns could still affect gene function
Are benign CNVs common in general population
- Yes
- 4.8-9.5% of genome is copy number variable
- 100 genes can be completely deleted with no phenotypic consequences
What is the criteria for a benign CNV
- High frequency in population databases (>1%)
- CNV overlaps, similar to one seen in healthy population dataset
- No known clinical associations (absence of gene content)
What is the criteria for a pathogenic CNV
- Previously reported in literature/ disease database
- Gene content= genes with a clinical association relevant to phenotype and mode of inheritance.
- Larger CNVs more likely to be pathogenic (often >1Mb)
- Often de novo
Who created a list of criteria to assess pathogenicity of CNVs
-ISCA (International Standards for Cytogenomic Array) 2010
What follow up studies can be done
-Parental studies. To determine if de novo or inherited from a parent with a balanced translocation (fish and metaphase analysis can aid)
