AddGenotypePriorProb_Mapping2Parents: Expected Genotype Frequencies in Mapping Populations

In lvclark/polyRAD: Genotype Calling with Uncertainty from Sequencing Data in Polyploids and Diploids

Expected Genotype Frequencies in Mapping Populations

Description

EstimateParentalGenotypes estimates the most likely genotypes of two parent taxa. Using those parental genotypes, AddGenotypePriorProb_Mapping2Parents estimates expected genotype frequencies for a population of progeny, which are added to the "RADdata" object in the $priorProb slot.

Usage

AddGenotypePriorProb_Mapping2Parents(object, ...)
## S3 method for class 'RADdata'
AddGenotypePriorProb_Mapping2Parents(object,
    donorParent = GetDonorParent(object), 
    recurrentParent = GetRecurrentParent(object), 
    n.gen.backcrossing = 0, n.gen.intermating = 0, n.gen.selfing = 0,
    minLikelihoodRatio = 10, ...)
    
EstimateParentalGenotypes(object, ...)
## S3 method for class 'RADdata'
EstimateParentalGenotypes(object,
    donorParent = GetDonorParent(object), 
    recurrentParent = GetRecurrentParent(object), 
    n.gen.backcrossing = 0, n.gen.intermating = 0, n.gen.selfing = 0,
    minLikelihoodRatio = 10, ...)

Arguments

object	A "RADdata" object. Ideally this should be set up as a mapping population using SetDonorParent, SetRecurrentParent, and AddAlleleFreqMapping.
...	Additional arguments, listed below, to be passed to the method for "RADdata" objects.
donorParent	A character string indicating which taxon is the donor parent. If backcrossing was not performed, it does not matter which was the donor or recurrent parent.
recurrentParent	A character string indicating which taxon is the recurrent parent.
n.gen.backcrossing	An integer, zero or greater, indicating how many generations of backcrossing to the recurrent parent were performed.
n.gen.intermating	An integer, zero or greater, indicating how many generations of intermating within the population were performed. (Values above one should not have an effect on the genotype priors that are output, i.e. genotype probabilities after one generation of random mating are identical to genotype probabilities after >1 generation of random mating, assuming no genetic drift or selection).
n.gen.selfing	An integer, zero or greater, indicating how many generations of selfing were performed.
minLikelihoodRatio	The minimum likelihood ratio for determining parental genotypes with confidence, to be passed to GetLikelyGen for both parental taxa.

Details

AddGenotypePriorProb_Mapping2Parents examines the parental and progeny ploidies stored in object$taxaPloidy and throws an error if they do not meet expectations. In particular, all progeny must be the same ploidy, and that must be the ploidy that would be expected if the parents produced normal gametes. For example in an F1 cross, if one parent was diploid and the other tetraploid, all progeny must be triploid. If both parents are tetraploid, all progeny must be tetraploid.

The most likely genotypes for the two parents are estimated by EstimateParentalGenotypes using GetLikelyGen. If parental gentoypes don't match progeny allele frequencies, the function attempts to correct the parental genotypes to the most likely combination that matches the allele frequency.

For each ploidy being examined, F1 genotype probabilities are then calculated by AddGenotypePriorProb_Mapping2Parents. Genotype probabilities are updated for each backcrossing generation, then each intermating generation, then each selfing generation.

The default, with n.gen.backcrossing = 0, n.gen.intermating = 0 and n.gen.selfing = 0, will simulate an F1 population. A BC1F2 population, for example, would have n.gen.backcrossing = 1, n.gen.intermating = 0 and n.gen.selfing = 1. A typical F2 population would have n.gen.selfing = 1 and the other two parameters set to zero. However, in a self-incompatible species where many F1 are intermated to produce the F2, one would instead use n.gen.intermating = 1 and set the other parameters to zero.

Value

A "RADdata" object identical to that passed to the function, but with data stored in three new slots:

priorProb	A two-dimensional list of matrices, with rows corresponding to object$possiblePloidies and columns corresponding to unique values in object$taxaPloidy. Each item in the list is a matrix. For each matrix, allele copy number (from zero to the total ploidy) is in rows, and alleles are in columns. Each value is the probability of sampling an individual with that allele copy number from the population. For any taxa ploidies other than the progeny ploidy, even priors are returned.
likelyGeno_donor	A matrix of the donor parent genotypes that were used for estimating genotype prior probabilities. Formatted like the output of GetLikelyGen.
likelyGeno_recurrent	A matrix of the recurrent parent genotypes that were use for estimating gentoype prior probabilities.

Note

For the time being, in allopolyploids it is assumed that copies of an allele are distributed among as few isoloci as possible. For example, if an autotetraploid genotype had two copies of allele A and two copies of allele B, it is assumed to be AA BB rather than AB AB. This may be remedied in the future by examining distribution of genotype likelihoods.

Author(s)

Lindsay V. Clark

See Also

AddGenotypeLikelihood, AddGenotypePriorProb_HWE

Examples

# load dataset and set some parameters
data(exampleRAD_mapping)
exampleRAD_mapping <- SetDonorParent(exampleRAD_mapping, "parent1")
exampleRAD_mapping <- SetRecurrentParent(exampleRAD_mapping, "parent2")
exampleRAD_mapping <- AddAlleleFreqMapping(exampleRAD_mapping,
                                           expectedFreqs = c(0.25, 0.75),
                                           allowedDeviation = 0.08)
exampleRAD_mapping <- AddGenotypeLikelihood(exampleRAD_mapping)

# examine the dataset
exampleRAD_mapping
exampleRAD_mapping$alleleFreq

# estimate genotype priors for a BC1 population
exampleRAD_mapping <- AddGenotypePriorProb_Mapping2Parents(exampleRAD_mapping,
                                                      n.gen.backcrossing = 1)
exampleRAD_mapping$priorProb

lvclark/polyRAD documentation built on Jan. 15, 2024, 4:19 a.m.