DAPAB W5L1: Effect of genes

Mendelians

Genetic basis of single gene traits

Biometricians

Inheritance of polygenic traits (regression)

Modern synthesis

Integration of mendialian school and biometricians

The problem of complex traits: many genes have a tiny effect

E.g. accurate genomic prediction of human height

50.000 individuals in the data, 20.000 SNPs affected height, together they explain at most the full variance → most SNPs must have a tiny effect

Why do we want to detect genes?

Human medicine

Easier genetic improvement

• Gene based selection

• Selection across populations

• - genomic prediction works well only within populations

• Introregression (plants)

• Design optimum genotypes that nature does not provide (e.g. CRISPR)

Provides fundamental knowledge

• How do genes determine phenotype

• Consequences of selection

Marker

Observerable (known) variant in the genome

SNiP

Single Nucleotide Polymorphism, bi-allelic marker

Genotyping

Determining the markers of an individual

Sequencing

Determining hte code of all DNA of an individuals (or species)

MAP

The whole set of markers of a species, ordered by chromosome and position

Physical map

Map of the genome of a species, with distance expressed in (kilo) bases

Genetic map

Genetic map is similar to physical map (Map of the genome of a species, with distance expressed in (kilo) bases), but with distance expressed in recombination units (cM = centimorgan) (linkage or recombination map)

Mapping

Determining the location of a variant on the genome

(marker) mapping

making a mpa of the genome with markers as landmarks

QTL

Quantitative trait locus: a variant that affects a quantitative trait (true effect)

QTL - mapping

(trying to) find the loci that affect a quantitative trait (the QTL) typical term used for statistical analysis based on linkage)

GWAS

Genome wide association study

Method to find QTL based on linkage disequilibrium (association)

Typically covering the entire genome, with many markers

Manhattan plot

The result of a GWAS

Genotypic value (G)

The (true) mean phenotypic value of individuals with a certain genotype

The expectation of the phenotypic value given the genotype

Model for a single locus

Three genotypic values: Gq1q1, Gq1q2, Gq2q2; one for each genotype

Additive effect (a)

half the difference between both homozygotes

(strictly speaking, a is the additive effect of the Q2 allele, -a is the additive effect for the Q1 allele)

Dominance effect (d)

Deviation of the heterozygote from the mean of both homozygotes

Finding the genes with statistics: the one locus model

What ist he average (main) effect of the Q2 allele, the a?

LM: y = intercept + a*#Q2-alleles + e