The impact of different genetic models on the optimum design of crossbreeding experiments

Abstract

ABSTRACTA method for the iterative set-up of optimum designs for crossbreeding experiments was used to study the robustness of designs to differences in the biological interpretation of two-locus epistatic interaction. Designs could be found which are efficient for the estimation of genetic models including, alternatively, seven different types of epistatic effects. Also, the design efficiency of a large-scale beef cattle crossbreeding experiment between Angus and Hereford cattle conducted at the Clay Center, Nebraska, and reported by Koch, Dickerson, Cundiff and Gregory (1985) was investigated and found to be high (proportionately 0·88 of the optimum). It was concluded that choice of the right genetic groups (i.e. types of crossbreds) seems to be more important for a good design than the exact number of observations allocated to each group.