CropGBM: An Ultra-Efficient Machine Learning Toolbox for Genomic Selection-Assisted Breeding in Crops

Continued improvement and falling costs of DNA sequencing have accelerated the increase in genomic resources for crop plants. From these efforts, considerable genetic diversity has been found and is aiding in the identification of markers for breeding purposes. High-density molecular markers have allowed for marker-assisted selection of quantitative traits that are controlled by a small number of genes. Recently, whole genomic selection has been proposed where markers genome-wide are used to estimate the contribution of all loci to traits of interest. In this chapter we outline the steps needed to perform genomic selection using machine learning. We describe our method called Crop Genomic Breeding Machine (CropGBM) and demonstrate its use on diverse maize lines containing high-density markers.