G_mat: Genomic relationship matrix
In jendelman/polyBreedR: Genomics-assisted breeding for polyploids (and diploids)

G_mat

R Documentation

Genomic relationship matrix

Description

Genomic relationship matrix

Usage

G_mat(geno, ploidy, p.ref = NULL, method = "VR1", sep = "_", n.core = 1)

Arguments

`geno`	genotype matrix or filename
`ploidy`	ploidy
`p.ref`	optional, reference population frequency for method "VR1"
`method`	"VR1" or "AM"
`sep`	character separating id from haplotype for "AM" method
`n.core`	number of cores, only used with "AM"

Details

For method="VR1", Method 1 of VanRaden (2008) is used, and its polyploid extension (Endelman et al. 2018). Missing data is replaced with the population mean for each marker. If p.ref is NULL, the current population is used as the reference population. For "VR1", geno is an allele dosage matrix for bi-allelic loci (sites x indiv).

For method="AM", the Allele Matching coefficient is calculated, which is the probability that two haplotypes sampled at random (with replacement) are identical (Weir and Goudet 2017). Missing data are not allowed. For "AM", geno is a phased genotype matrix, with alleles coded as positive integers. The name of each column is the id and haplotype concatenated with a separating character, sep. This character needs to be unique (i.e., not present in id or haplotype).