Exam 3 Flashcards
What are Mendelian traits?
discrete traits produced by a single locus or 2/3 loci
What are quantitative traits?
continuous traits produced by a large number of genes
What do selection gradients measure?
only the phenotypes relative to fitness
How can discrete entities (genes) produce traits with continuously varying values?
Because number of combos of genes are bigger and the differences between the traits are small
Complete additivity
Adding the affects of each allele
Directional selection
favors one side of extreme (linear graph)
Stabilizing selection
favors the mean value (distribution graph)
Disruptive selection
favors both extremes (parabola)
Quantitative trait loci (QTLs)
regions of the genome containing genes that influence quantitative traits
Genetic mapping
based on recombination rates
Physical mapping
based on sequencing (finds exact distance)
What is a single nucleotide polymorphism SNP?
something we can identify in the chromosome that lets us figure out the position in the chromosome (genetic marker)
How does QTL mapping work?
- Find two parents that differ in the trait of interest (easier if inbred)
- Make genetic maps of each parent ( with unique and shared markers to tell where a particular gene comes from)
- cross parents to get F1 offspring
- Self/inbreed F1s to get recombinant inbred lines (RILs)
Why are RILs important?
we end up with individuals who are homozygous at every locus but will be made up of different combinations of the parental genomes
What does QTL mapping tell us from an evolutionary context?
It can show us what part of the chromosome is important to make that specific change
What are some characteristics of GWAS (genome-wide association studies)
-QTL on a massive scale
-uses many thousands of SNPs
- with many thousands of individuals (not just two parents)
-genotypes every single individual
-Have to correct for multiple comparisons
-p-value= probability of seeing this association just by chance (below threshold are significant)
- generates Manhattan plot
GWAS vs QTL
GWAS:
-natural variation (more genetic variation)
- not only dealing with two parents (allows us to find more of the genetic variation that is affecting that species)
-many markers for fine scale mapping
-expensive
-only suitable for model organisms
QTL:
-only variation between parents is assessable
-usually uses fewer markers
- can use with more organisms
- less expensive
Candidate loci/genes
With some prior knowledge you can attempt to predict that certain genes/loci are going to be important genes.
-Not very reliable
Some possible follow up studies:
-targeted mutagenesis : to figure out which mutations affect function
-cell culture: can be used to apply evolutionary relevant treatments
-expression: expression level may be a more important evolutionary mechanism than protein changes
-Knockout: to assess effect of gene
-gene phylogeny: to assess evolutionary origin
What is the simplest molecular evolution?
a point mutation
Point mutation
change of a single nucleotide from one base to another
Types of point mutations
Transitions -> purine to purine (A, G) or pyrimidine to pyrimidine (C, T)
Transversions -> Purine to pyrimidine and vice versa
Jukes-Cantor Model
a model where everything changes at the same rate
- for every one transition there are two transversion changes ( in categories of changes)
Characteristics for rate of substitution
- It is higher for transitions than transversions
- it is slower than the mutation rate
- it is usually higher for synonymous positions in a coding region than for nonsynonymous positions
Substitution
If a mutation replaces the original base pair in the population, it is a substitution
How to measure the rate
Example:
You have sequences from 2 species. Each sequence is 50 bp long, and they differ at 7 positions. In addition, the species’ most recent common ancestor is estimated to have lived 2.5 million years ago. What is the rate of substitution per site per million years (sub/site/million years) for those sequences in those species?
7/50 = 0.14
sub rate= 0.14 sub/site / 5 million yr = 0.028 sub/site/million
General time reversible model (GTR)
get a good estimate of what the relationships are