Table 5.

Accuracy and bias of DGV in the 5-fold cross-validation using single-step GBLUP and Bayes approach

Traits	Data set	Accuracy (rDGV,DEBV)		Bias (bDEBV,DGV)
		single-step GBLUP	Bayes approach	single-step GBLUP	Bayes approach
MY305	Training 1	0.326	0.296	1.707	1.056
	Training 2	0.303	0.349	1.230	1.061
	Training 3	0.332	0.345	1.613	1.208
	Training 4	0.291	0.304	1.316	0.963
	Training 5	0.330	0.380	1.62	1.624
	Average (SD)	0.316 (±0.018)	0.335 (±0.034)	1.497 (±0.210)	1.182 (±0.262)
FY305	Training 1	0.324	0.358	1.618	1.031
	Training 2	0.466	0.462	1.808	1.239
	Training 3	0.29	0.328	1.296	0.856
	Training 4	0.377	0.381	1.812	1.197
	Training 5	0.414	0.417	2.191	1.368
	Average (SD)	0.374 (±0.070)	0.389 (±0.052)	1.745 (±0.3266)	1.138 (±0.199)
PY305	Training 1	0.394	0.363	1.852	1.113
	Training 2	0.403	0.399	1.743	1.115
	Training 3	0.377	0.360	1.507	1.041
	Training 4	0.298	0.305	1.383	1.022
	Training 5	0.298	0.36	1.440	1.384
	Average (SD)	0.354 (±0.051)	0.357 (±0.033)	1.585 (±0.203)	1.135 (±0.145)

DGV, direct genomic value; GBLUP, genomic best linear unbiased prediction; MY305, adjusted 305-d milk yield; SD, standard deviation; FY305, adjusted 305-d fat yield; PY305, adjusted 305-d protein yield.