Identifying Genetic causes of Human disease

Question 1

Types of genetic variation and subtypes information

Answer 1

Numerical
- autossomes: can gain chr 13, 18 and 21 but can´t loose chr -> multiple abnormalities
- sex chr: can gain but loss of Y always lethal and mostly in X (minor effects if happens)

Structural: deletions, insertions, duplications - caused by incorrect DNA repair, replication errors and inapproriate recombination

Question 2

Detection of variations

Answer 2

karyotyping
array CGH - comparative genome hybridization (oligo DNA probes) with reference DNA (red- deletion; blue/green- duplication)
DNA sequencing

Question 3

Types of inheritance

Answer 3

Autossomal
Sex linked
mitochondrial

Question 4

Differences between monogenic and complex disease + examples

Answer 4

Monogenic
- single gene
- rare (10,000)
- function (gene encodes protein)
- follows mendelian (dominant, recessive, X-linked)
e.g- cystic fibrosis, sickle cell anaemia

Complex
- multiple genes
- common
- disease risk because affects regulatory mechanisms (non coding proteins) - genes expression
- behavioural and environmental facts
e.g- diabetes and coronary artery disease

Question 5

detection in monogenic diseases (2) and problems

Answer 5

candiadate gene approach: sequence genss predicted to be involved - biased

whole exome (20000 variations)/genome sequencing (4-5million variations): - unbiased but expensive
- GenomAD: remove common variations (allele freq) and variants observed in healthy individuals
- VEP/ANNOVAR: filter on predicted effect (LOF) - STOP codons, frameshift, splice variant
- look for two mutations in same gene on both alleles (dominant or recessive)

Question 6

Detection of complex traits, ways to reduce false + (2)

Answer 6

Genome Wide Association Studies (GWAS) - discovery of regions related to diseases by surveying genomes of many people and compare frequency of variants - search for nearby variants
p<5x10^-8 threshold
FDR - 1% (1 in every 100 loci incorrect) or 5%

Modern arrays for SNP´s, INDELS, CNVs
- regional association plot (linkage between SNPs and identify possible gene- high r2 and location)
- bioinformatics: identify variants and predict function of SNP (missense? altered transcription binding site? changes in gene expression linked to SNP?)

Question 7

Traditional Risk scores

Answer 7

Framingham: age, gender, diabetes, smoking, total cholesterol, HDL cholesterol and blood pressure
(mid to later life)

Question 8

Genetic risk scores

Answer 8

metaGRS for CAD- 1,7mi variants
early life prediction
alone or in combination

Question 9

46 CAD loci - traits

Answer 9

12 lipid traits
4 blood pressure traits
other 32?