A small UK company has developed a platform to solve some of the biggest problems in agriculture today: improving the sustainability of agriculture and making crops more nutritious for consumers. The company was started in 2017 and is based at a world renowned plant research institute.
Artificial selection is still very much a trial-and-error exercise. With vast amounts of genetic and sequence data emerging on plants, one new approach would be ‘directed-trial-and-learn-from-errors’ with computational assistance. The company has proven that this is doable. When one combines AI, accurate single cell manipulation and gene editing tools, the number of iterations to develop a new variety is decreased significantly. Also, the precision of necessary modification of sequences grows exponentially. Within months, desirable traits are achieved that are close to theoretical limits.
The platform is split into two components:
1. A trait discovery element that utilises AI to produce new traits that can be rapidly tested and further improved.
2. A trait introduction element based on DNA free genome editing, making this technology GMO free in many countries outside of the EU. This technology is applicable over a wide variety of species.
Current projects include introducing a trait into a tomato variety to reduce waste. The trait ensures tomatoes won’t fall off the stem prematurely and won’t damage other tomatoes in transit. Another example project is tackling the cold induced sweetening in potatoes. They are using the potato's own genetics to predict and edit the regulatory elements needed to switch this process off only in the cold.
The company is looking for both breeders and producers of commercial crops and the agri industry in a broader sense who are looking for solutions conventional breeding has failed to address. The type of cooperation would be commercial agreement with technical assistance. The partner would state the problem to be solved and fund the development work. The UK company will develop the new trait, with some technical input from the partner probably. Such programmes need an exchange of genotypes and phenotypes to guide the iterations. Upon completion, the new variety will be licensed to the partner.