Genome and SNP analysis

Question 1

What is a locus?

Answer 1

Region of DNA

Question 2

What is a gene?

Answer 2

Section of DNA that encodes a protein (molecule)

Question 3

What is a polymorphism/genetic variant?

Answer 3

A position in the genome that varies in the population (single nucleotide polymorphism => SNP)

Question 4

What is an allele?

Answer 4

Alternative versions of a polymorphism

Question 5

What is a genotype?

Answer 5

Combination of alleles a person has in each particular polymorphism

Question 6

What is a haplotype?

Answer 6

Combinations of alleles in a chromosome

Question 7

What is a trait (disease or quantitative)?

Answer 7

A phenotype

Question 8

How genetically similar are humans in %?

Answer 8

Human beings are 99.9 percent identical in their genetic makeup. Differences in the remaining 0.1 determine our differences, including our susceptibility to certain diseases

Question 9

What do we look for in genomic studies?

Answer 9

We look for changes in the genome that are associated with changes in the phenotype

Question 10

What are candidate gene studies good and bad for?

Answer 10

They are more accurate but give less information in general

They are hypothesis-driven

Question 11

What are genome-wide studies good and bad for?

Answer 11

They are less accurate but give better understanding in general
They are hypothesis-free and hypothesis generating

Question 12

What are DNA probes?

Answer 12

Small DNA sequence (40bp) that attaches to our DNA if it is complementary

Question 13

What are Chips?

Answer 13

Millions of DNA probes that is used to try out which probe will be complementary to our DNA

Question 14

What is genotyping?

Answer 14

Genotyping is going to specific SNPs and looking at their DNA sequence (need of a probe/chip to select the precise sequence)

Question 15

What is sequencing?

Answer 15

Sequencing is to read every single base pair of a genome or part of a genome (much more expensive)

Question 16

What is linkage disequilibrium and why is it important to take it into consideration?

Answer 16

In population genetics, linkage disequilibrium is the non-random association of alleles at different loci in a given population
This happens when random mutations in a given population happen and that it is passed down only to that precise population => this gives different haplotypes
This is very important to take into consideration when working with different populations because it can be miss-leading when we think that a specific SNP is the cause of a disease but it is actually just a haplotype of the population tested

Question 17

What is a imputation?

Answer 17

It is a statistical inference (deduction) of unobserved genotypes
Aim: to test initially-untyped genetic variants for association with a trait of interest
Reference panels:
- reference panel choice is population specific
- the bigger the better (computing power)

Question 18

What do we need when seeking to identify a genetic variant?

Answer 18

1. at least more than a thousand people, ideally 20’000 people
2. a lot of Chips to genotype everyone with the database to store it
3. a specific phenotype to look for (obviously)

Question 19

What is OR (odds ratio)?

Answer 19

How much more risk of disease do I have if I carry this risk allele
OR=2 means that having a particular risk allele DOUBLES my chance of having the disease

Question 20

What is P-value?

Answer 20

Probability of this statistical result to be found by chance (probability of false positive)

Question 21

What does Beta show?

Answer 21

The strength of an association

Question 22

What is Bonferroni correction?

Answer 22

It is a correction to do when you have a really big population
Genome wide significant = 5*10^-8

Question 23

Does having found an associated variant mean that you found the causal variant?

Answer 23

No! Becasue it could be a linkage disequilibrium

Question 24

Is the closest gene necessarily the causal gene?

Answer 24

No! But usually yes…

Question 25

What does genome wide significance mean?

Answer 25

It means that the significant level that we require is generally a p-value lower than *10^-8, this is done when the number of people tested is very high => or else their will be a lot of false positive