impute: Impute missing data for bi-allelic markers

In jendelman/polyBreedR: Genomics-assisted breeding for polyploids (and diploids)

Impute missing data for bi-allelic markers

Description

Impute missing data for bi-allelic markers

Usage

impute(
  in.file,
  out.file,
  ploidy,
  method,
  geno,
  min.DP = 1,
  max.missing,
  params = NULL,
  n.core = 1
)

Arguments

in.file	VCF input file
out.file	VCF output file
ploidy	ploidy
method	One of the following: "pop","EM","RF"
geno	One of the following: "GT","DS"
min.DP	genotypes below this depth are set to missing (default=1)
max.missing	remove markers above this threshold, as proportion of population
params	list of method-specific parameters
n.core	multicore processing

Details

Assumes input file is sorted by position. Markers with no genetic variance are removed.

method="pop" imputes with the population mean for geno="DS" and population mode for geno="GT".

method="EM" uses parameter "tol" (default is 0.02, see rrBLUP A.mat documentation). Imputed values are truncated if needed to fall in the interval [0,ploidy].

method="RF" uses parameters "ntree" (default 100) for number of trees and "nflank" (default 100) for the number of flanking markers (on each side) to use as predictors. Because RF first uses EM to generate a complete dataset, parameter "tol" is also recognized.

jendelman/polyBreedR documentation built on Jan. 5, 2025, 12:13 a.m.