Table 7.

The averaged predictive ability across 100 replications for different methods in 4 scenes of simulation

Method	0.2				0.5
	5qtl		1000qtl		5qtl		1000qtl
	Predictive ability	p-value	Predictive ability	p-value	Predictive ability	p-value	Predictive ability	p-value
ELPGV	0.4346 ± 0.0116	—	0.3079 ± 0.0109	—	0.7034 ± 0.0064	—	0.5851 ± 0.0085	—
BayesA	0.4042 ± 0.0122	3.317E−15	0.2961 ± 0.0110	1.565E−08	0.6865 ± 0.0068	1.057E−16	0.5765 ± 0.0085	2.652E−17
BayesB	0.4257 ± 0.0118	0.001E−00	0.2842 ± 0.0112	1.096E−08	0.6985 ± 0.0065	0.002E−00	0.5756 ± 0.0086	6.910E−09
BayesCπ	0.3741 ± 0.0140	4.577E−11	0.2956 ± 0.0108	2.828E−10	0.6955 ± 0.0064	1.712E−12	0.5768 ± 0.0086	2.048E−16
GBLUP	0.2952 ± 0.0144	1.032E−22	0.2959 ± 0.0110	6.161E−07	0.5505 ± 0.0096	3.553E−34	0.5727 ± 0.0084	7.160E−13

ELPGV is the ensemble learning based on BayesA, BayesB, BayesCπ and GBLUP

— Represents no explicit result was found in this method