GAGBLUP (version 1.0)

gagblup: Genetic algorithm assisted genomic best liner unbiased prediction for genomic selection

Description

Performs genomic selection with genetic algorithm assisted genomic best liner unbiased prediction

Usage

gagblup(
  genotype,
  phenotype,
  fit_fun = "HAT",
  maxiter = 2000,
  nfold = 10,
  nTimes = 1,
  seed = 123,
  n_core = 1
)

Value

A list with following information is returned: $R2 the squared pearson correlation coefficient between the true value and the predicted value, $predicted_value the predicted value and the true value of the phenotype, $marker_selection a vector of the selected markers, with 1 indicates selected, 0 indicates not selected.

Arguments

genotype

a matrix for genotypes in numeric format, with individuals in rows and markers in cols.

phenotype

a vector of phenotype, missing (NA) values are not allowed.

fit_fun

the fitness function. There are four options: fitness = "AIC"/"BIC"/"FIT"/"HAT", default is "HAT"

maxiter

max number of iterations for GAGBLUP, default is 2000

nfold

the number of folds. Default is 10.

nTimes

the number of independent replicates for the cross-validation. Default is 1.

seed

the random number. Default is 123.

n_core

the number of CPU to be used, default is 1.

Examples

## load example data from GAGBLUP package
data(phenotype)
data(bins)
phenotype <- phenotype[1:200,3]
result <- gagblup(bins[1:200,],phenotype,fit_fun='HAT',maxiter=1,nfold=2,nTimes=1,seed=123,n_core=1)