Pearl millet [Cenchrus americanus (L.) Morrone] is a staple food for more than 90 million farmers in arid and semi-arid regions of sub-Saharan Africa, India and South Asia. We report the ∼1.79 Gb draft whole genome sequence of reference genotype Tift 23D₂B₁-P1-P5, which contains an estimated 38,579 genes. We highlight the substantial enrichment for wax biosynthesis genes, which may contribute to heat and drought tolerance in this crop. We resequenced and analyzed 994 pearl millet lines, enabling insights into population structure, genetic diversity and domestication. We use these resequencing data to establish marker trait associations for genomic selection, to define heterotic pools, and to predict hybrid performance. We believe that these resources should empower researchers and breeders to improve this important staple crop.

Global food security demands the development and delivery of new technologies to increase and secure cereal production on finite arable land without increasing water and fertilizer use. There are several options for boosting wheat yields, but most offer only small yield increases. Wheat is an inbred plant, and hybrids hold the potential to deliver a major lift in yield and will open a wide range of new breeding opportunities. A series of technological advances are needed as a base for hybrid wheat programmes. These start with major changes in floral development and architecture to separate the sexes and force outcrossing. Male sterility provides the best method to block self-fertilization, and modifying the flower structure will enhance pollen access. The recent explosion in genomic resources and technologies provides new opportunities to overcome these limitations. This review outlines the problems with existing hybrid wheat breeding systems and explores molecular-based technologies that could improve the hybrid production system to reduce hybrid seed production costs, a prerequisite for a commercial hybrid wheat system.

Modeling epistasis in genomic selection is impeded by a high computational load. The extended genomic best linear unbiased prediction (EG-BLUP) with an epistatic relationship matrix and the reproducing kernel Hilbert space regression (RKHS) are two attractive approaches that reduce the computational load. In this study, we proved the equivalence of EG-BLUP and genomic selection approaches, explicitly modeling epistatic effects. Moreover, we have shown why the RKHS model based on a Gaussian kernel captures epistatic effects among markers. Using experimental data sets in wheat and maize, we compared different genomic selection approaches and concluded that prediction accuracy can be improved by modeling epistasis for selfing species but may not for outcrossing species.