Table 3.

Average of prediction accuracy calculated by using the reproducing kernel Hilbert space (RKHS) model with different numbers of markers.

Number of markers	Fruit length		Fruit shape		Fruit width		Fruit weight		Pericarp thickness
	Avg a	SD	Avg	SD	Avg	SD	Avg	SD	Avg	SD
All SNPs (18,663 SNPs)	0.770	±0.010	0.751	±0.007	0.860	±0.005	0.836	±0.006	0.840	±0.002
2nd set (9,282 SNPs)	0.777	±0.011	0.756	±0.007	0.861	±0.002	0.830	±0.006	0.847	±0.003
3rd set (4,896 SNPs)	0.774	±0.010	0.739	±0.006	0.851	±0.009	0.830	±0.008	0.839	±0.005
4th set (2,758 SNPs)	0.768	±0.010	0.740	±0.008	0.843	±0.004	0.821	±0.004	0.826	±0.009
5th set (1,391 SNPs)	0.741	±0.008	0.702	±0.006	0.838	±0.003	0.814	±0.013	0.805	±0.007
6th set (711 SNPs)	0.700	±0.008	0.685	±0.013	0.801	±0.008	0.792	±0.005	0.774	±0.005

^aThe mean of 10-fold cross-validation values.