Improved precision of QTL mapping using a nonlinear Bayesian method in a multi-breed population leads to greater accuracy of across-breed genomic predictions

Table 1 Number of phenotypic records for each trait in the reference and validation sets

				Total	Reference		Validation
Breed	Trait	h ²	r	records	Bulls	Cows	YOB	Bulls
Holstein	FY, F%, MY, PY, P%	0.33	0.56	11 789	3049	8478	2005	262
Holstein	STAT	0.45	1	4481	1484	2746	2003	251
Holstein	FERT	0.03	0.05	11 040	2806	7838	2004	396
Holstein	SURV	0.025	0.035	10 999	2810	7825	2004	364
Jersey	FY, F%, MY, PY, P%	0.33	0.56	4793	770	3917	2005	105
Jersey	STAT	0.45	1	2552	300	2167	2001	85
Jersey	FERT	0.03	0.05	4628	716	3830	2005	81
Jersey	SURV	0.025	0.035	4592	697	3791	2004	103
Australian Red	FY, F%, MY, PY, P%	Validation animals only, 247 cow and 114 bull records.

FY = fat yield (kg/lactation), MY = milk yield (L/lactation), PY = protein yield (kg/lactation), F% = fat percentage (%) and P% = protein percentage in milk (%); STAT = stature; FERT = fertility (calving interval, days); and SURV = daughter survival (annual probability); h ² = trait heritability, r = trait repeatability, YOB = oldest year of birth. h ² and r were assumed for each trait when calculating the weights in the reference population.

ISSN: 1297-9686