Topic 8b: gene panels Flashcards
Sequencing is not perfect, what are certain considerations you might want to be aware of?
do you want a more specific or sensitive test (e.g. in a clinical setting, you want very high sensitivity)
what are the steps to next-gen sequencing in the clinical laboratory?
- gene panel design
- assay validation (make sure that what you get is actually true)
- variant calling (bioinformatics)
- nomenclature (harder than you think since we have many trasncripts which may be different)
- interpretation (Biological (Effects on protein function-trucation, in-frame deletion, etc)
and Clinical (Benign, Likely Benign, VUS, Likely Pathogenic, Pathogenic – is the effect actually important))
now a lot of AI to co through big libraries of data
what are other options, other than gene panel testing?
- whole genome
- whole exome
- whole transcriptome
what questions do you ask yourself when designing a panel?
- What genes to include?
- How many genes to include?
- Exons + part of introns
- Use correct transcripts
How do you actually choose/screen a gene panel?
- chosoe the right trasncript
- do target enrichment via capture: take either normal or cancer DNA and hybridize the target genes, which can be seperated via beads and then sequenced and alogned/compared
what do you look for after screening your gene panel?
- sequence coverage: germline variant vs somatic variant – there is much mroe somatic changes vs germline mutation and also easier to find
- types of genetic alterations: Nucleotide substitutions, Small insertions and deletions, Copy number variant: deletion/duplication (CNV), Structural rearrangements
what are different things you can use/do to interpret the data?
- Biology – does the function correspond to potential pathogenic diseases
- Mutation Databases (LOVD, Clinvar, etc.)
- Population Prevalence (ExAC, gNOMAD, 1000 Genomes, etc.) – if its seen many times, prob not pathogenic
- In-house datasets
- In silico Predictive softwares (SIFT, Polyphen, Mutation Taster, etc.)
- Segregation within family – is variant in two people with a rare disease the same? if it is, prob the cause
- Functional Studies – mutate every possible base pair, then run through an assay and see if the variant restore function, if not, its pathogenic
- 3D Protein Structure
what do you have to make sure is consistent throughout the sample when sequencing?
the number of reads
want at least 50 reads
why is it imporant tohave enough controls/multiple tests for the same thing when sequencing?
we are comparing two things to each other (healthy vs disease), it is not an absolute result per-say
what genes would you put and not put on a cancer panel?
we should not put most “risk” SNPs since a lot of them have high allele frequency but very low risk
we should put genes that have higher relative risk (in an relative risk vs allele frequency graph)
of the ten genes for breast cancer and ovarian cancer, which are associated with what type of cancer?
ATM, BARD1, BRCA1, BRCA2, BRIP1, CHEK2, PALB2, RAD51C, RAD51D, TP53
high risk breast cancer: BRCA1/2, PALB2, TP53
ER- breast cancer: BARD1, RAD51C/D
ER+ breast cancer: ATM, CHEK2
ovarian only: BRIP1
what genes would we put on a breast cancer panel? waht about ovarian?
breast: ATM, BARD1, BRCA1, BRCA2, CHEK2, PALB2, RAD51C, RAD51D, TP53
ovarian: BRCA1, BRCA2, BRIP1, RAD51C, RAD51D,
what are the different relatively common inherited susceptibility syndromes to colorectal cancer?
- Lynch syndrome
- Familial Adenomatous Polyposis
- MUTYH-associated Polyposis
what are distinguishable pathological features of lynch syndrome?
- Tumour infiltrating lymphocytes (“TILs”)
- Excess of signet ring cell features
- Villous polyps
- Microsatellite instability (MSI)
- Accelerated carcinogenesis
whats special about the APC gene?
is has a genotype-phenotype associations for its mutations – depending on where the mutation is, it has different pathological effects
what happens when the mutation is at the begining of the APC gene?
assocaited with milder version of polyptosis because the gene is translated but it starts a little further than the original start site, leading to a truncated but still functional protein
So what genes should be included on a colorectal cancer susceptibility gene panel?
hard to know because there are 200+ genes, so what if instead we do a large panel test
what is the phenotype of MUTYH mutations?
Varies from single CRC with no polyps
(unusual) to multiple polyps (AFAP-like, most
frequent presentation) to “full-blown” FAP
(very rare)
Broadens requirement for risk assessment in CRC
what suscpetibility does a NTHL1 gene mutation lead to?
is an autosomal recessive gene for colorectal cancer susceptibility
what has changed to make “large panel testing” possible?
- technical: The introduction of MPS was the key development in that allowed for the testing of many genes at the same time
- legal: overturning of the patenting of human DNA sequences by the Supreme Court of the United States in 2013 led to a profusion of commercial laboratories who were able to offer this testing for a broad set of breast cancer genes
- social: Commercial direct to consumer companies have offered recreational genetic testing (including ancestry analysis) sometimes combined with cancer genetic testing “thrown in”. This information has often been widely shared
what are possible approaches to large pannel testing?
- History first – Testing in the Cancer Genetics Clinic (referral to Genetics service)
- Tumor first – Oncology/Pathology (then refer to Genetics)
- Person first – population-based genetic testing (then refer to family doctor/genetics/oncology)
what are the pros of wider “panel” testing?
- The upside of genome and exome sequencing is the extraordinary biology that it has revealed, research has benefitted
- Papers have poured out, new mechanisms have been uncovered, therapies have been promoted
- it is possible that some cures have happened solely because of wider panel testing
- Genetically undiagnosed cases became diagnosed
- Persons maybe able to consider preventive actions that otherwise they would not have considered
what are cons of wider panel testing?
- when it comes to genetic testing for cancer susceptibility, i.e. the specific question of assessing cancer risk in unaffected persons, the picture is much less clear
- The main reason “panel testing” has been so enthusiastically taken up by clinicians is for reasons of convenience, rather than science
- Moreover, what used to be the domain of research is now “accepted clinical practice” and informed consent has been abandoned. Perhaps we shouldn’t care too much about this. But we should be aware that it has happened
- The rule of “do no harm” is at risk of being broken
what is the standard model (i.e. standard testing) for cancer families? what is the challenge?
gerline genetic testing, but then we have to problem that pathogenicity of specific variants cannot be established and any assumption made on pathogenicity is based on class of variants
and this even for genes that are well characterized
what is the rationale for genetic testing looking at the tumor first?
what is the challenge?
is primarily to identify potential treatment targets but a
secondary result is the availability of germline data, and this leading to the identification of many unaffected people who were not aware of their at-risk status, which we could then test
only possible if we have hundreds or even thousands of samples
what is the rationale for person first based genetic screening? what are its challenges?
you test everyine and ask questions later (just get a bunch of data); the issues are the same with any population based approaches (what do you test for? how do you test it? and when do you test?)
With all the gebetic testing, what do people (i.e. regulatory bodies and general population) actually care about?
- Immediate relevance to core concerns
- Consistent with values
- Actionable
how does second generation sequencing work (i.e. next gen sequencing)?
- take genomic DNA
- fragment
- adaptor ligation
- amplification
- detect in cycles, allowing the addition of a single detector to see what the base is, with subsequent cycles showing the next nucleotide
what is the extra step that you need to do when doing whole exome sequencing?
before actually sequencing, have to take the sheared DNA and select for the exons and wash away the rest
what is the bioinformatics steps to whole exome sequencing?
- QC and trimming of raw data (FASTQ)
- mapping reads to a reference genome (BAM/SAM)
- variant calling (VCF)
- variant annotation
- data interpretation/visualization/pedigree information
what is the “data trimming do”?
takes all the reads of the sequences and puts a quality score on them, telling you the probability that you make a mistake (e.g. a score of 10 means that there is a 1 in 10 prob of an incorrect base call, but a score of 40 means that there is a 1 in 10,000 chance)
what sort of information/data would you add in the variant annotion step of bioinformatics?
- Genes and transcripts affected by the variants
- location of the variants (e.g. upstream of a transcript, in coding sequence, in non coding RNA, in regulatory regions)
- Consequence/Effect of variants on the protein sequence (e.g. stop gained, missense, stop lost, frameshift)
- known variants allele frequencies (1000G, ExAC, gnomAD)
- Disease databases: ClinVar, Cosmic, OMIM
- Prediction Scores: SIFT, PolyPhen, CADD
what do you with the VCF file once you get them (e.g. when you have many samples)?
can characterize/compare them via GEMINI, which takes these variants form each patient and puts them in a category based on genome annotations, sample genotypes, and sample relationship (i.e. different studies/databases)
what is an example of a VCF filtering approach?
- exclude variants that are nat shared between by infected individuals and that are present in unaffected individuals
- exclude non-coding variants and coding variants that are not rare
- test for seggregation of identified genetics variants with disease phenotype
