Sequence variants selected from a multi-breed GWAS can improve the reliability of genomic predictions in dairy cattle

Table 4 Scenarios with best reliability (r²) for each breed and trait

Breed	Trait	50 K^a	Best scenario	n^b	r²	∆^c
HOLDK	Milk yield	0.44	MB-50 K + MBQTL10-25/1	3130	0.53	0.09
	Fat yield	0.48	MB-50 K + MBQTL20-25/1	424	0.58	0.10
	Protein yield	0.39	MB-50 K + MBQTL10-1/1	522	0.44	0.06
HOLFR	Milk yield	0.33	MB-50 K + MBQTL14-25/1	1046	0.41	0.08
	Fat yield	0.37	MB-50 K + MBQTL20-25/1	424	0.46	0.10
	Protein yield	0.37	MB-50 K + MBQTL14-25/10	517	0.44	0.06
JER	Milk yield	0.30	WB-50 K + MBQTL20-1/10	7	0.40	0.10
	Fat yield	0.16	MB-50 K + MBQTL10-10/10	775	0.20	0.04
	Protein yield	0.22	WB-50 K + MBQTL20-1/10	12	0.27	0.05
RDC	Milk yield	0.14	WB-50 K + MBQTL20-10/2	138	0.20	0.06
	Fat yield	0.11	WB-50 K + MBQTL14-10/2	635	0.19	0.07
	Protein yield	0.09	WB-50 K + MBQTL10-10/10	911	0.14	0.05

HOLDK Danish Holstein, HOLFR French Holstein, JER Jersey, RDC Danish Red
^aReliabilities when using 50 K SNPs
^bNumber of QTL variants included in the scenario
^cDifference in reliability between the best scenario and that obtained with 50 K SNPs

ISSN: 1297-9686