Figure 3From: Modelling the genetic architecture of flowering time control in barley through nested association mappingCross-validated proportion of explained genotypic variance (pG) of different applied models. The box-whisker plots depict the variation of explained genotypic variance after 100 cross-validations. The tested QTL models are (i) the single SNP locus Ppd-H1 (Mean pG = 0.36), (ii) GWAS with peak markers, representing the eight major QTL indicated in Table 1 (Mean pG = 0.64), (iii) the whole genome ridge regression best linear unbiased prediction (RR-BLUP, mean pG = 0.71), (iv) the BayesCπ prediction (Mean pG = 0.74), and (v) RR-BLUP including epistasis (Mean pG = 0.77).Back to article page