Fig. 5

Prediction abilities of genomic selection models for FLW, GY and PH based on 5 fold cross-validation using the 1k-RiCA. Average predictive abilities across 5 fold stratified cross validations experiments (k = 5) using 353 rice lines measured for flowering time (FLW), grain yield (GY), and plant height (PH) for seven different statistical models; Pedigree BLUP (Pedigree), BayesA, BayesB, BayesC, BayesLasso (Bayes L), rrBLUP (ridge regression) and RKHS using genomic and pedigree relationship matrices (RKHS G + A)