Loading Existing Data into a Simulation

In BreedingSchemeLanguage: Describe and Simulate Breeding Schemes

knitr::opts_chunk$set(fig.dpi = 96)
knitr::opts_chunk$set(echo = TRUE)

Load the BreedingSchemeLanguage.

library(BreedingSchemeLanguage)

Avoid writing over a previously existing simulation environment

if (exists("simEnv")){
  rm(list=names(simEnv), envir=simEnv)
  rm(simEnv)
}

Find the path to the HapMap file in the package

filePathName <- system.file("extdata", "exampleHapMapFile.hmp", package = "BreedingSchemeLanguage")

• This is what the file looks like

hapmap <- readLines(con=filePathName)
hapmap[11] <- ""
for (i in 1:11)
  cat(paste(substr(hapmap[i], 1, 83), "..."), "\n")

• The first row is a header
• The ninth row (locus8) has specification for a QTL
  • 17 is an unordered effect ID
  • 0 indicates additive gene action
  • -1.311 is the QTL effect

Load the historical haplotypes into the simulation environment

simEnv <- defineSpecies(loadData=NULL, importFounderHap=filePathName, nChr=5, lengthChr=100, nMarkers=300)
initializePopulation()

• Also create the population of diploid founders

Run selection on founders

phenotype()
genotype()
predictValue()
select(nSelect=50)
cross(equalContribution=T)

• Since the population was just genotyped, the default is for predictValue() to use markers to enable information sharing across observations
• The equalContribution parameter in cross() ensures that all parents contribute to the same number of progeny

Intense selection on genomic predictions

genotype()
predictValue()
select(nSelect=5)
cross()
cycleMeans <- plotData()

• The penultimate population was not phenotyped before selection
• The final cross() function serves to evaluate the response to genomic selection
• Note that this simulation only allows one replication: there will be variability in response including the possibility of a negative response to genomic selection

BreedingSchemeLanguage documentation built on May 2, 2019, 10:17 a.m.