The GRAIN Platform: Identification of Gene Targets for Improving Crop Yield

Achieving step changes in the seed yield of major crops is not likely to come about with single genetic manipulations, but rather with combinations of changes that are conceived through modeling, experimental iteration, and logical reasoning. Drawing upon its many years of metabolic engineering and modeling expertise, Yield10 is developing the GRAIN (Gene Ranking Artificial Intelligence Network) platform, which encompasses metabolic and regulatory modeling, along with a robust transformation and field trial system. Combinations of genetic changes for increased yield may encompass both traditional plant breeding and molecular breeding tools such as editing, transgenics, and rearrangements of native genetic elements. The GRAIN platform seeks to optimize combinations of these changes by utilizing stoichiometric, kinetic, thermodynamic, and co-expression analyses to narrow the list of actionable targets from many thousands to relatively few, providing plant scientists with manageably small libraries of manipulations to test. The iterative nature of GRAIN is enabled by using camelina as a model crop, because of its ease of transformation and regeneration, its rapid and compact growth, and the relatively few regulatory hurdles associated with field trials. Using this approach, Yield10 is identifying and de-risking yield traits for major agricultural crops.