Much of the attention surrounding artificial intelligence (AI) technology is focused on flashy “generative AI” tools that can create content like , images, high school essays, and even “clones” of live people. But the real breakthroughs have been happening behind the scenes where these powerful AI models can analyze massive amounts of data and identify patterns and relationships that could provide new solutions to old problems. That includes agricultural chemical makers, which are using AI to find novel ways of fighting pests and pathogens that are growing increasingly resistant to available crop chemicals.
Over at Bayer, the company’s “CropKey” system utilizes AI to unlock entirely new ways of protecting crops. The approach that Bayer scientists are taking is also entirely different. Rather than searching for molecules that could provide solutions, they are designing new molecules to affect specific proteins in specific pathogens and pests.
Just like many medicines used to cure disease, crop protection works by finding and inhibiting a given protein in the target weed, fungus, or insect. By inhibiting this one protein, the pest itself can be inhibited. Bayer says the protein can be thought of as a lock and the targeted molecules as the key.
Utilizing AI is also speeding up the discovery process. Using virtual screening, Bayer says it can increase the number of molecules analyzed exponentially. Along with predictive modeling and miniaturized testing, Bayer is able to narrow down those molecules that show promise and eliminate those that do not reach pre-defined safety and sustainability standards far faster and more accurately than previously imaginable.
Before, researchers could spend years screening compounds and conducting field trials only to discover that there was no path forward for their new solution. AI modeling and other tools can shrink this down to just to a few months.
“If we take the example of early research only, we today have at least three times the number of new modes of action compared to ten years ago,” said Bob Reiter, head of Research & Development for Bayer’s Crop Science division. “As our AI tools learn alongside us, every generation of solutions will be smarter and come faster than the ones that came before.”
Bayer’s first product developed with its CropKey tool is Icafolin, the first new post-emergent herbicide mode of action for broadacre crops in 30 years when it hits the market in Brazil in 2028. Additionally, the company says it now has more than 30 potential new molecular targets under investigation in discovery and has validated more than 10 targets as new modes of action in early research leveraging the CropKey approach, a pace of breakthrough discovery that Bayer says is unprecedented in the agricultural industry.
Syngenta is utilizing AI to develop new crop protection agents as well. The company says it now uses AI in all its R&D for new active ingredients. Bayer recently partnered with AI company InstaDeep to help accelerate the development of a “new level of trait control and crop performance.” Syngenta estimates that AI will reduce the typical time it takes to develop and bring a product to market down to 10 years from the current average of around 15 years, as well as decrease lab and field tests by as much as a third. (Sources: CropLife, The Wall Street Journal, Agriculture Dive)
