RR-BLUP GCA Model — RRBLUP

Fits an RR-BLUP model that estimates seperate marker effects for females and males. Useful for predicting GCA of parents in single cross hybrids. Can also predict performance of specific single cross hybrids.

Usage

RRBLUP_GCA(
  pop,
  traits = 1,
  use = "pheno",
  snpChip = 1,
  useQtl = FALSE,
  maxIter = 40L,
  simParam = NULL,
  ...
)

Arguments

pop: a Pop-class to serve as the training population
traits: an integer indicating the trait to model, a trait name, or a function of the traits returning a single value.
use: train model using phenotypes "pheno", genetic values "gv", estimated breeding values "ebv", breeding values "bv", or randomly "rand"
snpChip: an integer indicating which SNP chip genotype to use
useQtl: should QTL genotypes be used instead of a SNP chip. If TRUE, snpChip specifies which trait's QTL to use, and thus these QTL may not match the QTL underlying the phenotype supplied in traits.
maxIter: maximum number of iterations for convergence.
simParam: an object of SimParam
...: additional arguments if using a function for traits

Examples

#Create founder haplotypes
founderPop = quickHaplo(nInd=10, nChr=1, segSites=20)

#Set simulation parameters
SP = SimParam$new(founderPop)
SP$addTraitA(10)
SP$setVarE(h2=0.5)
SP$addSnpChip(10)

#Create population
pop = newPop(founderPop, simParam=SP)

#Run GS model and set EBV
ans = RRBLUP_GCA(pop, simParam=SP)
#> Warning: Reached maxIter without converging
pop = setEBV(pop, ans, simParam=SP)

#Evaluate accuracy
cor(gv(pop), ebv(pop))
#>        est_GV_Trait1
#> Trait1     0.6468518