Flywheel Genomics: Simultaneous trait discovery and genetic gain in plant breeding

Abstract

Genomic mapping has yielded extensive catalogs of quantitative trait loci underlying agronomic traits, yet translating these discoveries into breeding gains remains inefficient. Here, we introduce Flywheel Genomics, a framework that integrates trait discovery directly within rapid cycling breeding populations. Using empirical data from a smallholder-oriented sorghum breeding program, we demonstrate that recurrent intermating and selection maintain genetic diversity, effective population size, and recombination while reducing confounding from plant height and maturity. Within this population, we resolve loci underlying simple adaptive and complex environmentally responsive traits and generate large segregating populations for mapping and near-isogenic lines for locus validation. We further demonstrate applicability in a public wheat breeding program, where known agronomic loci were readily detected. Simulations show that rapid cycling better preserves the population genetic properties required for Flywheel Genomics than conventional pure line development. By integrating discovery with improvement, Flywheel Genomics reframes breeding programs as engines of both crop improvement and genetic insight.