Figure 1.
Figure 1.

Simple linear regression of genomic selection prediction accuracy (rGS) on phenotypic prediction accuracy (rP) and conventional marker-assisted selection prediction accuracy (rMAS) on rP. Training population size (NTP) for both genomic selection (GS) and marker-assisted selection (MAS) equaled 288 lines. rM, marker-based prediction accuracy.

 


Figure 2.
Figure 2.

The effect of training population size (NTP) on marker-based prediction accuracy (rM). AA, association analysis; AK, association analysis including kinship as a covariate; BC, BayesCπ; RR, ridge regression best linear unbiased prediction.

 


Figure 3.
Figure 3.

The effect of marker number (NM) on marker-based prediction accuracy (rM). AA, association analysis; AK, association analysis including kinship as a covariate; BC, BayesCπ; RR, ridge regression best linear unbiased prediction.