A new mixed analytical method for genetic analysis of diallel data

Diallel is a popular mating design used for crop and tree breeding programs, but its unique feature of a single observation with two levels of the same main effect, general combining ability (GCA), makes it difficult to analyze with standard statistical programs. A new approach using the SAS PROC MIXED is developed in this study for analyzing genetic data from diallel mating. Dummy variables for GCA effects were first constructed with SAS PROC IML, then PROC MIXED procedure was used to estimate variance components and to obtain BLUE (best linear unbiased estimators) of fixed effects and BLUP (best linear unbiased predictors) of random genetic effects (GCA and specific combining ability (SCA) effects) simultaneously. The new method can also be used for predicting individual breeding values with BLUP methodology, applying SAS IML to the outputs provided by PROC MIXED to calculate breeding value for each individual in the progeny test, adjusted for the fixed effects such as test location. The accurate BLUP prediction, the ability to estimate individual breeding values, and the ease of use would make this new method especially attractive for analyzing tree breeding data.