Selection of specific SNPs covariates.

Select markers based on either:

1. their effect size or the variance of their effects across environments estimated by a penalized linear regression model.

2. GWAS in each environment implemented via FarmCPU

select_markers(
  METData,
  trait,
  method_marker_effects = "FarmCPU",
  method_selection_EN = "variance_across_env",
  size_subset_most_variable_markers = 200,
  size_top_markers_by_env = 50,
  plot_penalty_regression_coefficients = F,
  plot_gwas = T,
  path_save_res = NULL,
  ...
)

Arguments

method_marker_effects	character Name of the method to estimate marker effects in each environment.
method_selection_EN	character Name of the method to select markers kept for further analyses. Options are variance_across_env or effect_size_per_env. variance_across_env is the default option.
size_subset_most_variable_markers	numeric Number of markers kept if the selection is based on the variability of marker effects across environments.
size_top_markers_by_env	numeric Number of markers kept if the selection is based on marker effect size by environment.
plot_penalty_regression_coefficients	logical Whether to plot on a grid environment ~ Chromosome the results from the Elastic Net variable selection by env.
plot_gwas	logical Whether to plot on a grid environment ~ Chromosome the results from the GWAS by env.
path_save_res	character Path where the plot should be saved.
...	Arguments passed to the marker_effect_per_env_EN() or marker_effect_per_env_FarmCPU() functions.
METData.	An object created by the initial function of the package, create_METData.R()
trait.	character Name of the trait under study for which a subset of markers should be chosen.

Value

a character vector containing the names of the markers selected for further analyses

Author

Cathy C. Westhues cathy.jubin@uni-goettingen.de