A hybrid expectation maximisation and MCMC sampling algorithm to implement Bayesian mixture model based genomic prediction and QTL mapping

Table 5 The accuracy and bias of with-breed prediction of GBLUP, BayesR(BR), emBayesR (EM), and HyB_BR (HB)

		Milk yield		Protein yield		Fat%		Fertility
		Acc.	Bias	Acc.	Bias	Acc.	Bias	Acc.	Bias
		Holstein reference to predict Holstein validation
GBLUP	+Poly^a	0.57	0.96	0.63	0.98	0.73	0.96	0.43	1.26
GBLUP	-Poly^b	0.56	0.86	0.59	0.87	0.71	1.15	0.42	1.27
BR	+Poly^a	0.63	0.91	0.64	1.01	0.79	1.06	0.43	1.19
BR	-Poly^b	0.61	1.00	0.63	1.06	0.77	1.13	0.41	1.19
EM	+Poly^a	0.62	0.79	0.63	0.85	0.77	0.98	0.42	1.15
EM	-Poly^b	0.62	0.92	0.62	0.94	0.74	1.06	0.41	1.15
HB	+Poly^a	0.63	0.93	0.63	0.97	0.79	1.09	0.43	1.19
HB	-Poly^b	0.63	1.03	0.62	1.06	0.76	1.17	0.42	1.19
		Jersey reference to predict Jersey validation
GBLUP	+Poly^a	0.59	0.93	0.65	0.91	0.54	0.71	0.15	1.05
GBLUP	-Poly^b	0.58	1.05	0.64	1.09	0.54	0.77	0.14	1.08
BR	+Poly^a	0.64	0.94	0.68	0.93	0.71	0.87	0.15	1.02
BR	-Poly^b	0.63	0.98	0.68	1.01	0.69	0.93	0.14	1.04
EM	+Poly^a	0.64	0.87	0.68	0.92	0.69	0.75	0.15	1.09
EM	-Poly^b	0.64	0.98	0.66	1.01	0.67	0.79	0.15	1.09
HB	+Poly^a	0.64	0.97	0.68	0.90	0.71	0.89	0.15	1.02
HB	-Poly^b	0.64	1.06	0.66	0.96	0.69	0.87	0.15	1.02

^ameans polygenic effects are included in the predictions; while ^bmeans the predictions do not include polygenic effects into the model

ISSN: 1471-2164