WOMBAT Example 22

Illustrates post-BLUP/REML calculations to transform estimates of `breeding
values' into estimates of marker effects together with their standard errors,
test statistics and p-values;
see Gualdron Duarte et al 2014 or Aguilar et al 2019

Simulated data for a trait with heritability of 0.4; 130 animals with allele
counts for 100 SNPs  

A: GBLUP analysis (all animals have genotypes) building the GRM using
   Van Raden's Method 1 and centering markers using observed frequencies

A1: as A but using GWAS (EMMAX) via SNP Snappy (run option --snap);
   demonstrates equivalence of test statistics

B: as A but deleting 20% of genotypes at random to yield single step GWAS




