Statistical Considerations for Applying Genomic Prediction to Potato

Summary

Marker assisted selection has gained considerable traction in diploid species, more specifically in the field of animal breeding. As a result, there are aspects of GP explored in diploid studies that have not been explored in tetraploid studies. In this thesis I investigated some of the statistical challenges when applying genomic prediction to tetraploid potato. The most significant takeaway is the importance of the training set as it sets apart the prediction accuracies in this thesis from other similar published studies. Across and joint ploidy prediction is investigated showing the potential of the newly proposed multi ploidy model.