Cross-validated standardized accuracies of prediction ( r
) for genomic selection compared to marker-assisted selection. Genomic selection (GS) was based on ridge regression best linear unbiased prediction (RR-BLUP), while marker-assisted selection (MAS) was based on QTL mapped with a limit of detection (LOD) threshold of 3.73 and the detected QTL (numbers above columns). Cross-validation was performed separately within population A and population B (CVG Within-Within-Same) for traits grain yield, plant height, starch content, and total pentosan content. Accuracies of prediction for MAS were taken from  and were based on SSR and DArT markers for population A and on SSR and SNP markers for population B.