Genomic dominance relationship matrix (Vitezica)

The D of Vitezica et al. (2013), the parametrization the book adopts for its GBLUP dominance model (Mrode & Pocrnic, 4th ed., Eqn 13.6, p.231, and Example 13.3, p.233): each genotype 0, 1, 2 is coded -2p^2, 2pq, -2q^2 and D = W W' / sum((2 p q)^2), frequencies from the genotyped animals themselves. Under Hardy-Weinberg equilibrium this partition is orthogonal to the additive one, which is why the additive and the dominance term enter the model as separate covariance groups. Like g_matrix(), the result is raw and may need a ridge before entering kernel().

Usage

g_dominance(genotypes)

Arguments

genotypes

list with ids and m (animals x markers, 0/1/2 coded); an NA is imputed with the mean of the observed dominance codes of its marker, and monomorphic markers are left out

Value

genomic dominance relationship matrix, rows and columns named by ids

Examples

geno <- list(ids = c("a", "b"), m = rbind(c(0, 1, 2), c(2, 1, 0)))
g_dominance(geno)
#>   a b
#> a 1 1
#> b 1 1