Variant interpretation Flashcards
What QA methods are implemented for variant interpretation?
Internal-
- Checking and reviewing compliance with SOPs (ISO15189)
- Training and DOPs for staff
- Periodic auditing of training to ensure skills are up to date
External-
1. Participation in EQA schemes e.g. GenQA
Name 4 types of DNA sequences
- Genome- Inton + exon
- cDNA- Complimentary DNA synthesised from a single stranded RNA molecule
- Expressed sequence tags (ESTs)- subsequence of cDNA used to predict which transcript to select
- Functional non-coding sequences
a. non-coding RNAs e.g. microRNAs
b. Cis and trans elements e.g. enhancers
c. Pseudogenes
d. TF binding sites
What are the ACMG guidlines?
American College of Medical Genetics and Genomics (ACMG)
Best practice guidelines for the interpretation of sequence variants. Includes terminology, evidence required for classification and classification system
What are the 5 classifications a variant be?
- Pathogenic
- Likely pathogenic
- Uncertain significance
- Likely benign
- Benign
What pieces of evidence can be used to classify a variant?
- Population data
- Variant databases
- Functional data
- Segregation data
What build should be used as a reference genome?
GRCh37 (hg19)
GRCh38 is the newest version, but is not currently used clinically
What are the 2 types of transcripts used to label variants?
- Refseq (most commonly used)- Open access and currated collection of DNA, RNA and protein
a. Number of labels to describe the transcript:
i. NC- complete genomic molecule
ii. NG- incomplete genomic region
iii. NM- mRNA
iiii. NP -Protien
iiiii. XM, XP, XR- predicted eukaryotic molecule - Ensembl ID- genes, transcripts, exons
What is SNP?
A variant with a population frequency of >1% (tolerated within a population)
What is the classification of a variant based on?
The application of evidence codes e.g:
PVS1- nonsense, frameshift, splice site variant
PS1- Same AA change previously identified as pathogenic
PM2- Absent from controls or extremely low in freq, ESP, 1000 genomes or EXAC
PP3- Multiple computational evidence to support deleterious effect
BS3- well established in vitro or vivo study showing no damaging effect
What should be done for all class 4/5 variant?
Variants should be confirmed using Sanger seuqencing
What 8 things should be included on a variant form?
- Patient information
- Gene information e.g. inheritance pattern
- In-house database information (optional)
- Disease specific database information about variant
e.g. LOVD (Lynch) - Generic variant databases
e.g. HGMD (germline), ClinVar (Somatic and germline), DECIPHER (rare disease) - Population frequency information
e.g. ESP, 1000 genomes, EXAC, gnomad - In silico tools
Based on either conservation or difference in AA or both
a. protein function- AlignGVGD, SIFT, Panther, Polyphen (machine learning- HumDiv- rare HumVar- any )
b. Changes in splicing- Gene splicer, MaxEntscan, HSF - Results from the literature
Why do/would we need to investigate a variant?
To assess clinical impact and to aid in the diagnosis of patient. accurate diagnosis > better clinical management > better clinical outcomes!
We also need to gather evidence as the change may not impact function! so the effect needs to be investigated
What are three types of additional testing?
Functional-
- Enzymatic e.g. a-glactosidase enzyme function test
- RNA study e.g. measure the impact of the variant at the mRNA level when evaluating the effect of variants in intronic, coding or UTR regions
Segregation testing-
- Testing of parents (unaffected)
a. if present in parents –> evidence that the variant is benign
b. If not in parents –> evidence of pathogenicity (de novo variant)
