We can use ASReml models similar to diallel models for factorial mating designs. The
difference is that male and female parents are distinct in the factorial mating design, so we would NOT use a field definition like ‘male !AS female’ for the factorial mating because a male parent coded as ‘1’ would not be the same parent as a female parent coded as ‘1’. Similarly, we can fit male, female, and male*female interaction effects in the model, but the factorial model coding is much simpler since we do not need to use the and() function to overlay the design matrices. On the contrary, we want to allow the male and female factors to be distinct and we will obtain two different estimates of GCA variance, one for male and another for female parents.