So, lets say you have a design with a factor Block (which is your incomplete blocks from your lattice design) and a factor Trt. We will assume Block to be random (beacise it is incomplete blocks) and Trt random (from where you want to get your BLUP values..
In GenStat you need to use VREML. The following is an example code:
Another thing is about the model . I consider a Trt as fixed and a block random, also I take into account the GEI (Genotype and environment interaction ) since I conduct my trial in two different environment for two year. Is this change how to assume the Model?
Just to let you know, I am not familiar with GenStat script, I use widows version.
Thenm if Trt is fixed and Block random then you need to decide how you will model Enviornment. Usually fixed and the interaction you be fixed in this case. Follow the menus and you will be fine.