Fig. 4

Comparison of the reliability of across population genomic prediction using Bayesian variable selection or GBLUP models. Comparison of the mean reliability of genomic prediction using Bayesian variable selection or GBLUP models for the four across population scenarios with genetic correlation of 1.0, 0.8 or 0.4 across populations; a Scenario 1: reference = HF, selection candidates = GWH; b Scenario 2: reference = HF & MRY, selection candidates = GWH; c Scenario 3: reference = HF, selection candidates = MRY; d Scenario 4: reference = HF & GWH, selection candidates = MRY. The vertical line indicates the natural logarithm of the number of independent chromosome segments (M _e ). M _e is estimated by Wientjes et al. [25] as: \( {M}_e=\frac{1}{Var\left({\mathbf{G}}_{Pop{.1}_i, Pop{.2}_j}-{\mathbf{A}}_{Pop{.1}_i, Pop{.2}_j}\right)} \); where \( {\mathbf{G}}_{Pop{.1}_i, Pop{.2}_j} \) refers to the genomic relationship between individual i from population 1 and individual j from population 2, \( {\mathbf{A}}_{Pop{.1}_i, Pop{.2}_j} \) refers to the pedigree relationship between individual i from population 1 and individual j from population 2, and the variance is taken over all pair-wise relationships between the individuals in the reference population and the selection candidates