Lecture 1 - Introduction

Question 1

What is the basis for statistical genetics?

Answer 1

The allocation of genes is random by nature.

Question 2

Fathers of population genetics:

Answer 2

Sewall wright = pedigree and inbreeding analysis

Ronald Fisher = statistics

John Haldane = Mathematical models of selection

Question 3

Genotype

Answer 3

Example = Aa - the combination of allelles at a certain locus

Question 4

What is a genetic marker?

Answer 4

Genetic data at molecular level that allows for distinction of genetic differences in individuals.

Question 5

DSL

Answer 5

Disease susceptibility locus - a locus where the gene has a varuant associared with the disease

Question 6

Types of markers

Answer 6

• SNP (1 every 300 BP - 10 million)

-

Question 7

Mendel’s first law

Answer 7

One allele of each parent is randomly and independently selected, with probability 1/2 fir transmission to the offspring; the alleles unite randomly to form the offspring’s genotype.

Question 8

Mendel’s second law

Answer 8

The alleles underlying two or more different traits are transmitted to offspring independently of each other; the transmission of each trait separately follows the first law of segregation.

Question 9

What is a genetic model?

Answer 9

Describes a probabilistic relationship between an individual’s genotype, and phenotype.

Question 10

Penetrance function

Answer 10

Describes the probability of disease given a genotype. P( Y | G ). 1 if mendelian disease.

Question 11

What is a phenocopy?

Answer 11

A diseae is caused at a different locus (or by a different thing).

Question 12

Name the 5 most common modes of inheritance:

Answer 12

1) Recessive
2) Dominant
3) Additive
4) Codominant
5) Mulitplicative

Question 13

Dominant model:

Answer 13

Only one disease allele is required:

P(Y = 1, dD) = P(Y = 1, DD)

Question 14

Recessive model:

Answer 14

two disease alleles are required to elevate the risk of Y=1

Question 15

Additive

Answer 15

The risk of homozygotes is the midpoint between homozygotes:

P(Y = 1 | dD) = 0.5 * (P(Y = DD) + P(Y = dd))

Question 16

Multiplicative

Answer 16

Additive on log scale:

P(Y = 1 | dD) = sqrt(P(Y = DD) * P(Y = dd))

Question 17

Codominant model:

Answer 17

All probabilitis are are different:

P(Y = 1, dd) not = P(Y = 1, dD) not = P(Y = 1, DD)

Question 18

Expectation of phenotype Y

Answer 18

Dichotomous: E(Y|G) = P(Y = 1| G)

quantitaive = median of underlying distribution. mean for gaussian.

Question 19

Generalized linear models and expectation of Y

Answer 19

Using a GLM it is possible to calculate the E(Y | G) with other covariates (this is doen through a link fnction.)

g(E(Y|G)) = X^T * Beta

Throough correct choice of g(y) and X all modes of inheritance can be expresseen as GLM.

Question 20

Name common link functions:

Answer 20

G(y) = y = beta0 + X^T * Beta- used for quantitative traits

g(y) = logit(y) = log(y/(1-y)) = beta0 + X^T * Beta- dichotomous traits.

g(y) = log(y) = beta0 + X^T * Beta- both cont and dicho

Question 21

Haplotype

Answer 21

Set of alleles on one chromosome

Question 22

What is the recombination fraction?

Answer 22

Probability that recombination occurs between two loci.

Mathers law:

Theta = (1 - P0)/2

P0 = prob. zero crossovers.

Question 23

Difference crossover and recombination

Answer 23

Only a recombination if non-identical genetic material changes places.