Estimation and consequences of direct-maternal genetic and environmental covariances in models for genetic evaluation in broilers

Table 3 Observed accuracies of prediction estimated for the five models per sex

	Moda		Modam		Coram		Corepe		Coramepe
	f	m	f	m	f	m	f	m	f	m
cor(y_c, EBV_reduced)	0.216	0.248	0.205	0.241	0.202	0.239	0.199	0.235	0.192	0.230
Halfsibs cor(y_c, EBV_reducedHS)	0.197	0.203	0.198	0.205	0.197	0.205	0.196	0.203	0.195	0.202
cor(EBV_full, EBV_reduced)	0.614	0.679	0.595	0.669	0.587	0.659	0.618	0.686	0.614	0.680

Observed accuracies of prediction are defined as the correlation between the phenotype corrected for fixed effects (y_c) and the predicted breeding values (EBV_reduced), either when these values were estimated for the same individual (cor(y_c, EBV_reduced)) or for paternal half-sibs pairs (Halfsibs cor(y_c, EBV_reducedHS)). The third definition of the accuracy of prediction cor(EBV_full, EBV_reduced) is defined as the correlation between the breeding values estimated with the full dataset (EBV_full) and the one estimated using the reduced dataset (EBV_reduced). Moda is the model for which no maternal genetic effects are included and for which the direct-maternal environmental covariance is null. Modam is the model for which both the direct-maternal genetic and the direct-maternal environmental covariance are null. Coram is the model for which the direct-maternal genetic effect is considered as non-null while the direct-maternal environmental covariance is null. Corepe is the model for which the direct-maternal environmental effect is considered as non-null while the direct-maternal genetic covariance is null. Coramepe is the model for which both the direct-maternal genetic and the direct-maternal environmental covariance are non-null

ISSN: 1297-9686