Table 5.

P-values for the Hotelling-Williams’ t-statistic to test H ₀: the reliability obtained by the model selecting haploblocks is equal to the reliability obtained by the model using all the haploblocks vs. H ₁: the reliabilities are different, according to number of main SNP effects

Main	Protein			Fertility			Mastitis
SNPs †	BLUP	4 mixture		BLUP	4 mixture		BLUP	4 mixture
1000	0.00	0.00		0.15	0.01		0.00	0.00
2000	0.00	0.00		0.26	0.01		0.03	0.09
3000	0.03	0.01		0.06	0.16		0.14	0.32
4000	0.15	0.05		0.17	0.36		0.19	0.60
5000	0.34	0.10		0.34	0.49		0.42	0.69
6000	0.55	0.16		0.31	0.51		0.43	0.77
7000	0.60	0.30		0.34	0.57		0.59	0.84
8000	0.79	0.41		0.45	0.79		0.74	0.98
9000	0.74	0.54		0.45	0.79		0.82	0.93
10 000	0.91	0.60		0.54	0.96		0.83	0.85
20 000	0.87	0.93		0.82	0.76		0.98	0.82
30 000	0.85	0.91		0.81	0.93		0.99	0.63
40 000	0.88	0.91		0.76	0.99		0.96	0.72
50 000	0.65	0.98		0.86	0.88		0.87	0.75

^†number of highest (absolute) SNP effects used to select haploblocks