. 2015 May 26;5(8):1575–1583. doi: 10.1534/g3.115.018119

Table 6. GEBV accuracies from across-population prediction using GBLUP (scenarios 8–11).

			N Training			N Validation			Accuracy (SD)	Slope^b
Trait	Scenario	r²	DL	LW	F1	DL	LW	F1	GBLUP	GBLUP
AFI	8	0.45	1067	—	—	—	1389	—	−0.05	−0.57
	9	0.45	711	—	—	—	1389	—	−0.04 (0.01)^a	−0.71
	10	0.45	—	1389	—	1067	—	—	−0.02	−0.27
	11	0.45	—	711	—	1067	—	—	−0.02 (0.03)^a	−0.43
TNB	8	0.45	1066	—	—	—	1383	—	0.05	1.01
	9	0.45	693	—	—	—	1383	—	0.04 (0.01)^a	1.37
	10	0.49	—	1383	—	1066	—	—	0.03	0.56
	11	0.49	—	693	—	1066	—	—	0.00 (0.02)^a	0.00
LBW	8	0.78	1070	—	—	—	1385	—	0.26	0.83
	9	0.78	708	—	—	—	1385	—	0.23 (0.04)^a	0.83
	10	0.80	—	1385	—	1070	—	—	0.22	0.73
	11	0.80	—	708	—	1070	—	—	0.16 (0.03)^a	0.65
LVR	8	0.50	1069	—	—	—	1385	—	0.17	0.70
	9	0.50	705	—	—	—	1385	—	0.15 (0.03)^a	0.75
	10	0.53	—	1385	—	1069	—	—	0.20	1.40
	11	0.53	—	705	—	1069	—	—	0.13 (0.04)^a	1.22

r² is the mean reliability of deregressed estimated breeding values from the training population. SD, standard deviation, DL, Dutch Landrace; LW, Large White; F1, cross between DL and LW; AFI, age at first insemination; TNB, total number of piglets born; LBW, litter birth weight; LVR, litter variation. Scenario 8, across-population prediction for LW; scenario 9, across-population prediction for LW with a reduced number of DL animals; scenario 10, across-population prediction for DL; scenario 11, across-population prediction with a reduced number of LW animals.

Estimate obtained by 20-random training-validation populations.

Regression coefficient of the GEBV on the DEBV.