Handling large groups of individuals as single "dummy individuals"

There are many reasons for using this trick. Many of these relate to speed of operation.

Typically, all members of a group of animals (all the same sex, but possibly also all on the same farm or in the same age class) to be mated have their candidate status removed (set e.g. their Maxuse value to zero in the main datafile). They are then replaced by adding a single dummy individual to the main datafile, with the same sex as the removed candidates, and an appropriately high Maxuse value, at least as high as the actual number of matings to be carried out in that group. You might also use an appropriate AbsMinuse value. You might allocate the dummy individual with average values for trait EBVs and Indices in the group.

If you want to properly reflect the relationships between the group of animals and the rest of the candidates, you can construct a relationship matrix with the dummy individual having a relationship to each other candidate that is the average of the group’s relationships to each other candidate. You can then import this new relationship matrix using Genomic Relationship Matrix (GRM).

Such a dummy individual might represent, for example, all the commercial cows in a given region or all the female pigs in a multiplier tier. In such cases, Matesel can manage both nucleus breeding and dissemination to commercial or multiplier herds in a single analysis.

This approach has been used with knowledge of the history of pedigreed sires used unpedigreed commercial herds, together with a modified geneflow approach, to give the predicted average relationship between each current candidate male and all current unpedigreed females in the commercial herd. This helps manage progeny inbreeding at the commercial level. It could also be used together with segregation analysis to reduced expression of genetic defects at the commercial level.