Topic 16: Adaptive Variation

Question 1

What are putatively neutral genetic markers?

Answer 1

These are microsatellites, and other DNA non-coding regions that have no effect on phenotype
They are useful for studying demographic processes (changes in effective population size, dispersal), and for social structure (parentage and relatedness)

Question 2

How are discrete traits inherited? How do we study them?

Answer 2

Usually inherited in a Mendelian fashion and are controlled by a single gene.
Can study them through controlled crosses.

Question 3

Most traits show a continuous range of variation, why is this?

Answer 3

They are governed by many genes, each of small effect and each following the Mendelian rules of inheritance.
Hard to study genetic variation for quantitative traits when there are so many genes

Question 4

What is quantitative genetics? What is variance?

Answer 4

Quantitative genetics describes populations using trait means, variances, and co-variances among traits and relatives.
Variance is dispersion around the mean:
- individual observations usually differ from the mean
-deviation is distance from the mean
-variance is average squared deviation
Covariance is the product of deviation from the mean for two measures

Question 5

What is VA, VD and VI?

Answer 5

VA= additive genetic variation, measurement of the phenotypic variance among individuals that is due to directly inherited genetic effects
VD: non-additive source of genetic variation due to dominance
VI: non-additive source of genetic variation due to epistasis

Question 6

What is VA responsible for in terms of phenotype?

Answer 6

THe component of genetic variance that causes the resemblance between parents and offspring, and also responds to natural selection and therefore allows adaptation to happen

Question 7

How is Va often expressed?

Answer 7

As a fraction of the phenotypic variance, known as heritability (narrow sense heritability) h^2

Question 8

Phenotypic variance among relatives allows estimating of VA in a number of ways. What are some ways to estimate heritability?

Answer 8

Can be estimated as the slope of the offspring value on the average parental value, and the steeper the slope the higher the higher the VA
Can be estimated in natural populations if parentage is known and traits are measured in a large sample

Question 9

How do you determine narrow sense heritability (h^2)?

Answer 9

The slope of the line of best fit on a graph containing average phenotype of parents and average height of offspring, is used to directly measure the heritability of a trait.
Slope is narrow sense heritability.

Question 10

What else can heritability and additive genetics variance be estimated from?

Answer 10

Single parent regression, full sib/ half sib ANOVA, controlled crosses, selection experiments, more complex tests than ANOVA and regression

Question 11

What are some difficulties in estimating VA?

Answer 11

Large sample size needed for compared to most population genetic data
Good pedigree information is needed and often requires genetic markers
Phenotypic data is often harder to collect than genetic data
Parameters (heritability and variance) are population and trait specific
Extremely difficult to separate environmental and maternal effects from genetic effects

Question 12

What genetic architecture?

Answer 12

When you find out how much genetic variation there is for a particular trait but you do not know how many genes influence that trait, their relative importance, or how they influence that trait

Question 13

What are the two options for genetic dissection of complex traits?

Answer 13

Quantitative trait loci mapping
Association genetics (candidate genes and genome-wide association studies)

Question 14

What are the four steps in QTL?

Answer 14

1. Define underlying population sample (families, controlled crosses, multigenerational pedigree where trait of interest is variable and segregating)
2. Collect genetic data from markers of known chromosomal location (type each individual at a panel of markers that are mapped in the species of interest)
3. Measure phenotypes
4. Test for association (test for relationships between genotypes along chromosomes and phenotype)

Question 15

What is a linkage map?

Answer 15

An ordered array of genetic markers places at regular intervals along chromosomes.
Typically first generation maps have 250-300 markers, or one marker every 10 CM

Question 16

What is one centimorgan equal too?

Answer 16

it is equal to a 1% chance that a marker at one genetic locus on a chromosome will be seperated from a marker at a second locus due to crossing over in a single generation
Used to identify regions of chromosomes that have genes which affect trait variation
Alleles at genes that affect trait variation will co-segregate with linked marker alleles
This generates a statistical correlation between marker alleles near the gene and trait variation

Question 17

What does QTL refer to?

Answer 17

A chromosomal region that affects trait variation

Question 18

What are advantages of QTL?

Answer 18

Few markers needed (100s) to cover the genome because it relies on meiosis and physical linkage to create associations between markers and genetic variants that influence traits along wide chromosomal regions

Question 19

What are disadvantages of QTL?

Answer 19

Low resolution (wide QTL intervals), like assigning heritability to large regions of the chromosomes
Requires a pedigree or controlled crosses in which the trait and markers are segregating
Requires a genetic map (know markers)

Question 20

What are candidate genes?

Answer 20

Genes that have known function usually in a model organism, that can be hypothesized to have a similar function in another organism.
“Shortcut” to target a few genes that are likely to play a role in trait variation instead of searching across entire genome
You sequence the candidate gene in a new taxon, and look for polymorphic regions that can be used as markers and tested for association (SNPS)

Question 21

What is an example of a candidate gene in beach mice?

Answer 21

Melanocortin-1 receptor plays a role in melanin biosynthesis, good candidate gene for controlling coat colour polymorphisms, identified a SNP causing a single amino acid change in MCR-1 in beach mice therefore lighter coat
This is a large effect locus, with a quantitative trait nucleotide contributing to variation in fitness

Question 22

What are pitfalls of the candidate gene hypothesis?

Answer 22

Not always easy to identify likely candidates (genes/pathway may be unknown, or many genes in a pathway contribute to a phenotype)
Polymorphisms outside of the coding region may relate to function (expression, regulation, dosage, or other post-transcriptional modification may contribute)

Question 23

What is involved in association mapping? What does it rely on?

Answer 23

Involves detecting statistical associations between markers and phenotypes using a large sample of unrelated individuals. Relies on linkage disequilibrium between markers, and functional genes generating statistical associations
Associations extend over short distances, much denser coverage is needed compared to linkage mapping

Question 24

What is the general procedure for a GWAS study?

Answer 24

Score individuals for phenotype
Genotype them at thousands or millions of markers
Mathematically test for association between phenotype and genotype for each marker to obtain p-values (chi-square or regression)

Question 25

What is a Manhattan plot?

Answer 25

Manhattan plot: markers ordered by chromosome, and chromosomes differentially coloured

Question 26

What are the three considerations when performing GWAS tests?

Answer 26

Correction for multiple test, population structure or family relations can cause false associations, statistical power

Question 27

Why is a correction for multiple testing necessary?

Answer 27

Many individual tests conducted for each gene, therefore higher chance of error occuring, therefore perform either the Bonferroni or Benjamin-Hochberg corrections

Question 28

Why is population structure or family relations important to consider?

Answer 28

These can cause false associations, because if a population has a high incidence of some factor and is genetically structured, then the alleles that happen to be at high frequency due to the other factor will tend to associated together.
Studies conducted in a single geographical population get around this, but admixture also presents problems

Question 29

How do we deal with population structure such that we do not get false associations?

Answer 29

Assessed with a Q-Q plot (want loci to lie 1:1 on the line)
Quantified with genomic inflation factor lamba (want lamda to be 1)

Question 30

What are the differences between population genetics markers and population genomics markers?

Answer 30

Population genetics: modest number of markers (microsat, allozyme, mtDNA), selectively neutral, highly polymorphic and informative (many alleles and high heterozygosity), generally do not know the location on the chromosome)

Population genomics: many markers (SNP), cannot assume selectively neutral as some may be functional, limited polymorphism (usually 2), can be known chromosomal location or associated with specific genes)

Question 31

What are the aims of population genetics and population genomics?

Answer 31

Population genetics: describe population diversity, relatedness, and genetic structure,=, dealing with demographics of population processes, average effects over loci

Population genomics: find genes that may have adaptive significance, deal with selection/adaptation/genome structure at the molecular level, interested in individual locus and not average

Question 32

What are two other aims of population genomics?

Answer 32

1. Neutral evolutionary processes effect and shape patterns of genetic diversity across entire genome
2. Selection acts on individual genes and shapes genetic diversity only on these genes that are under selection

Question 33

What is a selective sweep?

Answer 33

Occurs when strong selection rapidly “sweeps” a favourable allele to fixation so quickly that there is little opportunity for recombination
REsults in decrease in genetic diversity in the genomic region around the allele that is under selection
loci that are physically close together are less likely to be separated during recombination in meiosis (physical linkage)
Get the adaptive mutation and all other mutations that were beside it fixed, therefore little genetic variation in that region

Question 34

How do we use molecular fingerprints of past selection?

Answer 34

Molecular fingerprints of past sweeps can be visible in a genome and can be identified by scanning the genomes for regions with unusual levels of genetic diversity
Low diversity: positive/direction selection
High diversity: balancing/stabilizing selection

Question 35

What is the Fst outlier approach to population genomics? When would we see high or low Fst?

Answer 35

If every locus is only influenced by demographic effects (drift and migration), then we expect all loci to have the same degree of differentiation measured by F-stats.
If a locus is under selection through local adaptation then we might expect that locus to show a different pattern from other loci
If the same allele is favoured in every population then we would see LOW Fst (low differentiation compared to other loci)
If different alleles are favoured in different populations then we would see HIGH Fst
If heterozygotes are favoured we would see LOW FIS