Genomic dissection and prediction of heading date in perennial ryegrass

BMC Genomics

Table 2 Population size and results (with SE) for all CV schemes

CV scheme	Pop.size	ρ_ӯf;ĝ ^†	ρ_g;ĝ	bias (b)
k-fold	1757	0.90 _0.01 ^a	1.04 _0.07	1.02 _0.01
pp-fold	1757	0.84 _0.01 ^b	0.98 _0.06	1.10 _0.02
k-fold (UK)	466	0.78 _0.03 ^a	0.86 _0.09	1.06 _0.04
pp-fold (UK)	466	0.52 _0.04 ^b	0.57 _0.07	1.30 _0.10
k-fold (others)	1291	0.90 _0.01 ^a	1.04 _0.07	1.02 _0.01
pp-fold (others)	1291	0.86 _0.01 ^b	0.99 _0.07	1.17 _0.02
UK - > others	466	0.78 _0.02 ^N	0.90 _0.07	1.46 _0.03
Others - > UK	1291	0.71 _0.03 ^N	0.78 _0.08	0.92 _0.04
F2s - > SYNs (GS)	1757	0.88 _0.05 ^a	0.93 _0.24	1.02 _0.06
F2s - > SYNs (GWAS)^‡	1757	0.74 _0.07 ^b	0.78 _0.21	1.33 _0.13

^†different letters indicate a significant difference between the two CV schemes (P < 0.001) based on Hotelling-Williams test. N indicates that the comparison does not apply, as models were based on different sets of data
^‡using all SNPs that were declared significant after FDR test

ISSN: 1471-2164