When the entire genome is partitioned in K parts, a multiple variance components model can be used to jointly estimate the variance components [33], (6)y=Xβ+∑k=1k=KWku+Cv+e,uk~N(0,σuk2I),v~N(0,συ2I),e~N(0,σe2I),where Wk is the standardized genotype matrix corresponding to the k-th part of the genome, and σuk2 is the corresponding variance component. Similarly, the proportion of variance explained by the k-th part is given by hk2=Mkσuk2∑kMkσuk2+σe2, where Mk is the number of markers in the k-th part. In this study, we used two approaches: Partitioning the genome into chromosomes.Partitioning the genome into genic and intergenic regions.
