Table 2. Mean predictive ability of each model in each trait for three SNP densities (NSNP).

TRAIT	TYPE	NSNP	ABLUP	GBLUP	BayesB	BLUPGA#
OC	OIL	10K	0.278	0.333	0.353	0.339
OC	OIL	100K	0.278	0.360	0.375	0.362
OC	OIL	500K	0.278	0.368	0.380	0.368
MONO	OIL	10K	0.276	0.374	0.401	0.390
MONO	OIL	100K	0.276	0.398	0.417	0.403
MONO	OIL	500K	0.276	0.406	0.422	0.407
SESQ	OIL	10K	0.444	0.438	0.459	0.439
SESQ	OIL	100K	0.444	0.465	0.475	0.539
SESQ	OIL	500K	0.444	0.469	0.473	0.497
CIN	OIL	10K	0.335	0.688	0.705	0.703
CIN	OIL	100K	0.335	0.701	0.724	0.711
CIN	OIL	500K	0.335	0.706	0.727	0.716
APIN	OIL	10K	0.302	0.552	0.566	0.553
APIN	OIL	100K	0.302	0.563	0.574	0.573
APIN	OIL	500K	0.302	0.567	0.576	0.587
PCIN	OIL	10K	0.398	0.770	0.764	0.771
PCIN	OIL	100K	0.398	0.784	0.787	0.785
PCIN	OIL	500K	0.398	0.783	0.787	0.790
HT	BIOMASS	10K	0.084	0.197	0.187	0.247
HT	BIOMASS	100K	0.084	0.161	0.169	0.196
HT	BIOMASS	500K	0.084	0.160	0.163	0.172
dHT	BIOMASS	10K	0.108	0.136	0.132	0.192
dHT	BIOMASS	100K	0.108	0.134	0.133	0.182
dHT	BIOMASS	500K	0.108	0.141	0.140	0.167

^#BLUP|GA with weighting of the top 0.1% of SNPs by squared effect size as estimated with BayesB. BLUP|GA includes the case of ω=0 in each CV, so the minimum possible outcome for each CV is equal to GBLUP.