Figure 4

Comparison of prediction models using the parental population. Prediction accuracy was measured as the Pearson’s correlation coefficient (r) between predicted genotypic values and phenotypic values. (A) Twelve methods were tested. RR: ridge kernel regression, GAUSS: Gaussian kernel regression. (B) Regression models were built based on the results of GWAS. The three SNPs that showed high –log10(p) values in GWAS (red points in Supplementary Fig. S5) were selected. MLR: Multiple Leaner Regression. (C) Prediction models that considered only additive or both additive and dominance effects were used.