poolHelper (version 1.1.0)

maeHet: Average absolute difference between the expected heterozygosity computed from genotypes and from Pool-seq data

Description

Calculates the average absolute difference between the expected heterozygosity computed directly from genotypes and from pooled sequencing data.

Usage

maeHet(
  nDip,
  nloci,
  pError,
  sError,
  mCov,
  vCov,
  min.minor,
  minimum = NA,
  maximum = NA,
  theta = 10
)

Value

a data.frame with columns detailing the number of diploid individuals, the pool error, the number of pools, the number of individuals per pool, the mean coverage, the variance of the coverage and the average absolute difference between the expected heterozygosity computed from genotypes and from pooled data.

Arguments

nDip

is an integer or a vector representing the total number of diploid individuals to simulate. Note that scrm::scrm() actually simulates haplotypes, so the number of simulated haplotypes is double of this. If it is a vector, then each vector entry will be simulated independently. For instance, if nDip = c(100, 200), simulations will be carried out for samples of 100 and 200 individuals.

nloci

is an integer that represents how many independent loci should be simulated.

pError

an integer or a vector representing the value of the error associated with DNA pooling. This value is related with the unequal contribution of both individuals and pools towards the total number of reads observed for a given population - the higher the value the more unequal are the individual and pool contributions. If it is a vector, then each vector entry will be simulated independently.

sError

a numeric value with error rate associated with the sequencing and mapping process. This error rate is assumed to be symmetric: error(reference -> alternative) = error(alternative -> reference). This number should be between 0 and 1.

mCov

an integer or a vector that defines the mean depth of coverage to simulate. Please note that this represents the mean coverage across all sites. If it is a vector, then each vector entry will be simulated independently.

vCov

an integer or a vector that defines the variance of the depth of coverage across all sites. If the mCov is a vector, then vCov should also be a vector, with each entry corresponding to the variance of the respective entry in the mCov vector. Thus, the first entry of the vCov vector will be the variance associated with the first entry of the mCov vector.

min.minor

is an integer representing the minimum allowed number of minor-allele reads. Sites that, across all populations, have less minor-allele reads than this threshold will be removed from the data.

minimum

an optional integer representing the minimum coverage allowed. Sites where the population has a depth of coverage below this threshold are removed from the data.

maximum

an optional integer representing the maximum coverage allowed. Sites where the population has a depth of coverage above this threshold are removed from the data.

theta

a value for the mutation rate assuming theta = 4Nu, where u is the neutral mutation rate per locus.

Details

The average absolute difference is computed with the mae function, assuming the expected heterozygosity computed directly from the genotypes as the actual input argument and the expected heterozygosity from pooled data as the predicted input argument.

Note that this functions allows for different combinations of parameters. Thus, the effect of different combinations of parameters on the average absolute difference can be tested. For instance, it is possible to check what is the effect of different coverages by including more than one value in the mCov input argument. This function will run and compute the average absolute difference for all combinations of the nDip, pError and mCov input arguments. This function assumes that a single pool of size nDip was used to sequence the population.

Examples

# a simple test with a simple combination of parameters
maeHet(nDip = 100, nloci = 10, pError = 100, sError = 0.01, mCov = 100, vCov = 200, min.minor = 1)

# effect of two different pool error values in conjugation with a fixed coverage and pool size
maeHet(nDip = 100, nloci = 10, pError = c(100, 200), sError = 0.01,
mCov = 100, vCov = 200, min.minor = 1)

# effect of two different pool error values in conjugation with a fixed pool size
# and two different coverages
maeHet(nDip = 100, nloci = 10, pError = c(100, 200), sError = 0.01,
mCov = c(100, 200), vCov = c(200, 500), min.minor = 1)