DeepMind AI system pinpoints specific DNA mutation causing flower structure changes in plants

Researchers at Paris-based biology lab Living Models combined Google DeepMind's Gemma language model with BOTANIC-1, a specialized plant genetics system. According to Google DeepMind, this combination can sift through complex plant genome data in hours, replacing years of standard laboratory and field work.

Identifying the specific DNA change responsible for desirable traits, such as drought resistance or improved yields, usually requires cultivating multiple generations of plants and sifting through thousands of candidate mutations. In a test involving melons, the new pipeline pinpointed a single key mutation out of thousands of possibilities. Accelerating this discovery process could help breeders create climate resilient crops much faster.

While software can rapidly flag promising genetic targets, real-world validation remains a necessary step. It remains unclear how smoothly discoveries from melon trials will extend to major staple crops such as corn or wheat. Furthermore, identifying useful genetic variations is only part of the challenge, as regulatory frameworks and public acceptance will ultimately determine how quickly these AI-assisted crops reach farm fields.