Fig. 3. The estimates of genomic prediction accuracy with actual datasets.

The genomic prediction accuracy is measured via the Pearson’s correlation coefficient between the genomic estimated breeding value (GEBV) and the observed phenotypic data representing the true breeding value. A repeated fivefold cross-validation approach was adopted to compute the correlation, and the experiment was replicated 100 times. The final accuracy is computed by taking an average over all the fivefolds, and 100 replicates. The SBLUP achieved the lowest accuracy among all the methods. The GBLUP achieved higher accuracy for the yield traits of wheat except for WY1. Except for BTW, the Bayesian methods performed better for the rest of the four traits. The higher accuracy of BLUP and Bayesian methods are seen to be according to their heritability estimates.