Breeding Designs for the Application of Genome Wide Selection in Maize.

Genome Wide Selection (GWS) aims to predict additive genetic effects of individuals using high density marker panels (e.g., SNP markers). In this work, GWS is applied in a plant breeding design in order to select double-haploid derived inbred lines with high breeding value for yield which in turn will be evaluated through the regular variety development process. Two novel concepts in this process are the use of historical phenotypic data to estimate the effect of markers and the assumption of a polygenic model for complex traits like yield. Results based on simulation showed that the best inbred lines can be identified with average to high accuracy depending on the data structure and the number of markers in the panel. The main results of the application of this process are: better estimates of QTL with small effects, savings in resources not allocated to evaluated inbreds with low breeding value, i.e., low potential to high yield, statistical correction for (a)biotic effects that create error in the traditional phenotypic selection process, and faster selection cycle and increasing genetic gain.