Comparison of whole-genome prediction models for traits with contrasting genetic architecture in a diversity panel of maize inbred lines

Table 3 Prediction accuracies $(r_{(g, \hat{g})})$ and their standard deviations (s.d.) for different WGP models

Trait	h ²	RR-BLUP		LASSO		Elastic net		RKHS		BayesB
		$r_{(g, \hat{g})}$	s.d.	$r_{(g, \hat{g})}$	s.d.	$r_{(g, \hat{g})}$	s.d.	$r_{(g, \hat{g})}$	s.d.	$r_{(g, \hat{g})}$	s.d.
Dry matter yield	0.93	0.61	0.07	0.51	0.11	0.56	0.08	0.61	0.07	0.59	0.08
Plant height	0.97	0.57	0.09	0.45	0.11	0.48	0.11	0.57	0.09	0.56	0.08
Lignin content	0.88	0.69	0.07	0.60	0.08	0.60	0.10	0.68	0.07	0.58	0.09
Dopamine	0.97	0.74	0.06	0.79	0.06	0.79	0.06	0.74	0.07	0.75	0.06
Ribitol	0.95	0.49	0.12	0.61	0.10	0.63	0.10	0.50	0.10	0.50	0.11
719700-204	0.96	0.79	0.06	0.82	0.05	0.82	0.05	0.80	0.05	0.80	0.08

Results are averaged over all 100 cross-validation runs. For the agronomic traits, h²is the heritability on a line-mean basis and for the metabolites, the repeatability is shown.

ISSN: 1471-2164