Fig. 4
From: A new approach fits multivariate genomic prediction models efficiently
Bias of estimates of heritability for scenario 1 (a wheat dataset) and scenario 2 (b soybean dataset) for different true heritabilities (columns) and true genetic correlations (rows), based on 100 replicates of the simulation. Letters indicate Tukey’s test of multiple comparisons (\(\alpha =0.05\)). Asterisk indicates that the mean is significantly different from zero (\(\alpha =0.05\))