Predicting complex traits using a diffusion kernel on genetic markers with an application to dairy cattle and wheat data

Table 3 Averages of kernel elements and their predictive correlations for the wheat data

Kernel	θ	k(x_i,x_i)	k(x_i,x_j)	Cor(ŷ^test,y^train)	MSE
Diffusion	3	1	0.136	0.586	0.685
	3.25	1	0.289	0.580	0.673
	3.5	1	0.466	0.577	0.681
	4	1	0.752	0.547	0.704
	5	1	0.962	0.522	0.721
Gaussian	0.005	1	0.134	0.582	0.686
	0.003	1	0.290	0.579	0.697
	0.002	1	0.434	0.562	0.697
	0.001	1	0.655	0.558	0.703
	0.0005	1	0.809	0.556	0.673
G 1	NA	2	-0.003	0.518	0.709
G 2	NA	2	-0.003	0.521	0.708

Average of diagonal k(x_i,x_i) and off-diagonal k(x_i,x_j) kernel elements, predictive correlation, and mean-squared error of prediction (MSE) for the diffusion, Gaussian, and two additive genomic relationship kernels at different values of the bandwidth parameter θ for the wheat data. The predictive correlation and the MSE were obtained from a 10-fold cross-validation. Additive genomic relationship kernels (G 1 and G 2) do not involve bandwidth parameters. The best prediction within the same kernel is underlined.

ISSN: 1297-9686