Genetic Mapping

Goal for Genetic Testing

Looking for a specific DNA mutation (change in DNA sequence) that changes the phenotype in some way

How do we find mutations out of billions of nucleotides?

We can use molecular markers like SNPs to find where to look

Molecules markers

sequence-based

They're spread throughout the genome (across all chromosomes)

Different allles means we can have different genotypes

polymorphic means we have variety in alleles and can have lots of different varieties but in SNPs, only two

Single Nucleotide Polymorphism (SNP)

Single base change

Different alleles allow us to look at different genotypes and trace through pedigrees

Uses chromatogram

Haplotype

Group of alleles/chunk of DNA inherited together most of the time unless broken through crossing-over or recombination

Linkage

Genes/markers are physically linked on the same chromosome

Decay in linkage

recombination starts taking place between out mutation and the surrounding markers

Linkage Disequilibrium (LD)

Non-random association of alleles at different loci within a population.

What is seen when there is LD?

Certain alleles being inherited together more often than expected.

What are we looking for with LD?

Where LD exists between alleles and our phenotype of interest

Why does LD occur?

The closer the marker is to a mutation of interest, the more often see it with the mutation.

Goal of Genome-wide Association study (GWAS)

Find a region in the genome where mutation is

In GWAS, do we want genotypes to be the same or different between cases and control groups?

Different because GWAS relies on LD looking at frequency of certain alleles/genotypes existing with phenotype of interest.

Genome-wide Association Study (GWAS)

analysis of genetic variation across an entire genome, searching for associations (linkages) between variations in DNA sequence and a genome region encoding a specific phenotype

In GWAS, do all of the individuals have the specific phenotype of interest?

No, there will be cases and controls, people with the phenotype and people without.

Manhattan Plot

Visual representation of SNPs in the association analysis to a phenotype (GWAS results)

x-axis on Manhattan Plot

Location of SNPs based on chromosomes and location within chromosomes

y-axis of Manhattan Plot

Significance value, basically how much LD and how different cases are from controls.

Dots on Manhattan Plot

Represent single SNPs

Quantitative Trat Loci (QTL)

Regions that contain gene(s) that involve a quantitative trait

Expect several locations when dealing with quantitative traits with clusters of several genes.

eQTL

Expression quantitative trait loci affecting gene expression of quantitative traits.

QTL Mapping

Same set up as GWAS but a technique for identifying the regions of the genome in which quantitative trait loci occur

How is QTL mapping different from GWAS?

Now you can use a range of phenotypes looking to see where individuals fall along a scale rather than just a yes or no outcome.

Requirements for effective mapping

Must have good phenotype data

Need relatively large population size

Need lots of markers spread throughout the genome

Need markers to be informative and have variation

Need quality genome data

Phenotype Quality

Need accurate phenotypes

Affect of Phenotype Quality on GWAS graph

Well-defined phenotypes --> clean associations

Poorly-defined phenotypes --> can miss association

False Positives

Might pull up other phenotypes that are different without realizing it.

Study Size for GWAS

Need a large enough population size to see any association (difference between varying phenotypes)

Study size affect on GWAS graph

Small sample size --> not enough power and can miss associations

Large sample --> can detect LD

Number and quality of markers for GWAS

Need enough markers (SNPs) to detect linkage disequilibrium (LD)

Affect of having monomorphic phenotype

Useless

Importance of having variety in phenotypes

Allows us to detect association with SNP and things nearby

Affect of number of markers on GWAS map

more markers --> define smaller region and able to catch more associations

Less markers --> larger region need to look at so more likely to miss association

Locus Impact

larger locus impact --> easier to find

small locus impact --> harder to find

what do peaks in Manhattan plots represent

indicates a difference between your cases and controls or highs vs lows

Significance cutoff for a GWAS

p < 0.001 or log3

Genome wide significance

Threshold for identifying significant genetic associations.

Is based on the # of tests (#SNPs) and corrects for multiple testing

Hit/peak in Manhattan plot

indicates markers have genotypes/alleles that are statistically different between groups

Fine mapping

Refining the region focusing on by zooming in region

Positional candidates genes

possible genes impacting gene of interest based on position to SNP

Fine mapping goal

Find gene(s) in region where association was found that could cause the phenotype.

Large haplotype + many significant SNPs

Peak is wider and has LD decay decreasing in significance

Small haplotype + very strong signal

1st SNP has really high signal and there's a lot less underneath it

Small haplotype + weaker signal

1st SNP less high of a signal/weaker and has less SNPs backing it up beneath it

high LD affect on mapping

- when zoom in you tend to identify larger haplotype blocks

- increase chance of finding association (peaks) in initial mapping

- decrease effectiveness of fine mapping

Low LD affect on mapping

- find shorter haplotype blocks because they're broken down more

- decrease chance of finding association (peaks) in initial mapping

- increase effectiveness of fine mapping

Effect of closed breeding on chance for LD

close breeding increase chance of LD

GWAS main goal

Where do I look in the genome

Genome secondary goals (3)

What is causing my phenotype of interest

Find mutation and develop genetic test

Find predictive SNPS