Medical Genomics Flashcards
When was the human genome mapped?
2003
When did next generation sequencing begin?
2010->
What is NGS?
Next generation sequencing
The rapidly advancing technology that makes it economically viable to sequence individuals rather than species
What % of the genome is the exome?
1%
What does studying the transcriptome RNA allow us to study?
Gene expression
Gene fusions
Splice variants
How many base pairs in a genome?
About 3200 million
How many base pairs in an exome?
5–70million
Negatives of exome sequencing
Only picks up 85% of disease causing variants
Requires additional sample prep
No assessment of non coding regions
Simple repeats, GC rich and highly homologous regions are poorly captured
Clinical exome develops over time
Negatives of genome sequencing
Much more expensive
Massive data (storage?)
Interpretation more expensive and difficult
Steps in exome sequencing
Genomic DNA Shotgun library-> fragments Rehybridisation Pull down and wash Captured DNA is sequenced Mapping, alignment and variant calling
What does variant calling rely on?
Accurate alignment to a reference sequence from the Human Genome Reference
No read is the full gene
Who should be sequenced?
Distantly related concordantly affected individuals (share very few mutations)
Closely related discordant individuals (very few differences)
What is a compound heterozygote?
The presence of two different mutant alleles at a particular gene locus, one on each chromosome of a pair
Different mutations in mum and dad but on same allele
Offspring has no functioning copies of that gene
What does consanguineous mean?
The quality of being descended from the same ancestor as another person.
What are the 6 main types of mutations?
Inherited (autosomal recessive, autosomal dominant, X linked recessive, consanguineous autosomal recessive)
De novo
Mosaic
What is needed for bioinformatics?
4Gb data for each sample
Biologists skilled in programming
IRIDIS4 supercomputer
What affects the quality of sequencing?
Read depths
Standard bias
PCR duplicates
What is tiered analysis?
Prioritised analysis on basis of clinical presentation
Use known targeted gene panels
Top candidate genes first then pathway, system, others
Minimises incidental findings
How can you filter data?
Mode of inheritance
Variant type
Concordance/discordance to affected relatives
Frequency in control populations
Give an example of an ‘in silico’ tool
Computerised predictors
Logit
Mutationtaster
Name 2 reference databases
UK10K, in-house, ESV
What do genetic libertarians believe about incidental findings in genome analysis?
“Return comprehensive data”
Patient has right to know
Return all data on known and unknown risk variants
What does the ACMG recommend about incidental findings in genome analysis?
“Return data on a limited number of conditions and genes”
20 diseases, 60 genes
What do genetic empiricists believe about incidental findings in genome analysis?
“Only return data that is truly significant”
Don’t create burden of ‘patient in waiting’
Penetrance for most variants is unknown
Who is involved in a medical genomics/bioinformatics MDT?
Genomicists Clinicians Immunologists Bioinformaticians Lab techs Registrars Research nurses PhD students Molecular biologists Pathologists
What are the issues with big data?
Storage space
Safety/confidentiality of data
Identifiable data refreshed for clinical life-course registry
Need data on cancer registries, rare diseases, mortality data
All data needs to be kept behind a firewall until processed
Then it can be given to clinicians, academics and industry
What are the options for a patient when they get their data back?
Info about patient’s main condition
Info about ‘serious and actionable’ conditions (optional)
Info on carrier status for non-affected parents of children with rare diseases (optional)
What can genomics be used for?
Making diagnoses
Identifying people and increased risk of developing a condition
Diagnosing infections and tracking epidemics
Personalising drug treatments
Identifying appropriate cancer treatments
Developing new therapies
How can genomics be used to personalise drug treatments?
Genetic variation can impact drug metabolism.
- > dosage
- > frequency
- > SEs
- > adverse reactions
How can NGS be used for monitoring of disease?
Measure minimal residual disease by liquid biopsy-> quantifies disease, monitors remission and detects relapse
How does a mutation become heritable?
If it is germline (egg/sperm)
Then all cells are affected in offspring
What do are proto-oncogenes?
Proto-oncogenes code for proteins that help to regulate cell growth and differentiation.
Could become an oncogene due to mutations or increased expression
What do tumour-suppressor genes usually do?
Prevent cell division and check cells for mutations before allowing them to progress in mitosis
What happens if a tumour-suppressor gene is mutated?
Loses function
Tumours are not suppressed
Apart from tumour-suppressor genes and proto-oncogenes, what other genes do we worry about mutations?
DNA repair genes
Mutation may lead to decreased repair of damaged genes
Increased likelihood of damaging mutations
What are the 2 types of mutation in cancer?
Drivers (give clones a selective advantage)
Passengers (have no effect)
What can be used for NGS samples?
Fresh frozen tissue biopsies
Formalin fixed paraffin embedded samples
Peripheral blood cells
Saliva
What does ‘cancer is clonal’ mean?
Cells are from one common ancestor cell characterised by one or more somatic driver mutations
What are targeted gene panels good for?
Cheap, quick, higher read depth Can estimate clonality Best for rare variants Restricted based on current knowledge Can predict prognosis
What are the 4 Ps of personalised healthcare?
Personalise
Predict
Prevent
Participate
How will clinical genomics stratify disease?
Classify disease based on causative pathways instead of signs and symptoms
