Fig. 5From: Optimizing genomic reference populations to improve crossbred performanceAverage empirical (a and c) and predicted accuracy (b and d) of estimated breeding values of purebred selection candidates for crossbred performance. The reference population consisted of purebreds (PB), two-way crossbreds (CB) derived from the same sires as used in the nucleus (2wayCB_TOP), different sires with zero (2wayCB) or one multiplication step (2wayCB_1MP), four-way crossbreds derived from the same sires as used in the nucleus population (4wayCB_TOP), different sires with zero (4wayCB), one (4wayCB_1MP), or two multiplication steps (4wayCB_2MP). The purebred-crossbred correlation was equal to 0.75 (a and b) or 0.5 (c and d). Heritability was 0.2 in the purebred and crossbred populations. Averages were calculated across 50 replicatesBack to article page