Table 2. Estimates of accuracy for genomic evaluation and individual GEBV across imputation scenarios.

Trait	Scenarioa	Imputation Accuracyb	rEBV, GEBVc	rEBVd	$\frac{r_{EBV,GEBV}}{{\overline{r}}_{EBV}}$	${\overline{r}}_{GEBV}$	HPDe
BF	1	(1, 1)	0.68101	0.8510	0.7998	0.6852	[0.5395, 0.8211]
	2	(1, 0.95)	0.67951		0.7981	0.6861	[0.5467, 0.8164]
	3	(0.88, 0.88)	0.65982		0.7749	0.7014	[0.5727, 0.8267]
	4f	(1,1)	0.7210		0.8405	0.8560	[0.8174, 0.8768]
D250	1	(1, 1)	0.66031	0.8020	0.8229	0.6575	[0.5073, 0.7948]
	2	(1, 0.95)	0.65551,2		0.8170	0.6585	[0.5187, 0.7962]
	3	(0.88, 0.88)	0.64632		0.8054	0.6750	[0.5345, 0.7985]
	4f	(1,1)	0.5354		0.6550	0.8438	[0.8048, 0.8704]
LEA	1	(1, 1)	0.65161	0.8529	0.7639	0.6859	[0.5386, 0.8325]
	2	(1, 0.95)	0.64911		0.7610	0.6868	[0.5377, 0.8214]
	3	(0.88, 0.88)	0.63642		0.7461	0.7040	[0.5667, 0.8330]
	4f	(1,1)	0.7165		0.8201	0.8549	[0.8223, 0.8787]

GEBV, genomic breeding value; EBV, estimated breeding values; HPD, highest posterior density; BF, backfat thickness; D250, number of days to 250 lb; LEA, loin muscle area.

^aScenarios 1: all observed genotypes, 2: genotypes in prediction animals imputed with large reference haplotype panel (~1800), 3: genotypes in prediction animals imputed with small haplotype reference panel (128), and 4: validation animals with at least one close relative in the reference panel.

^bAccuracy of genotype imputation R² for training and validation animals: $\left(R_T^2,R_V^2\right)$.

^cTukey honest significant difference post-hoc comparison of accuracy of genomic evaluation across imputation scenarios.

^dAverage accuracy of EBV in the validation panel.

^e95% HPD interval of GEBV accuracy across validation animals.

^fScenario with young animals in the validation panel that almost all have at least one close relative in the training panel.

^1,2Means with different superscript differ significantly according to Tukey post-hoc tests with α = 0.05.