AI designs first functional genome, validated in the lab
Sep 18, 2025
Key Points
- Profluent Bio researchers published a preprint demonstrating the first functional AI-designed complete genome, a novel bacteriophage, validated in laboratory testing.
- The breakthrough uses language models trained on genome sequences to generate novel designs, moving synthetic biology from isolated genetic circuits to functional whole biological systems.
- The work proves that neural networks can learn latent patterns in genetic organization well enough to design novel sequences that retain biological function.
Summary
Researchers at Profluent Bio designed and validated the first functional AI-generated complete genome in a laboratory setting. Samuel King presented a preprint titled "Generative Design of Novel Bacteria Phages with Genome Language Models," demonstrating that an AI system successfully designed a novel bacteriophage genome that proved functional when tested in the lab.
The work addresses a fundamental constraint in synthetic biology. Many of the most complex and useful biological functions emerge only at the scale of whole genomes, making genome-level design a long-standing technical challenge. Profluent Bio trained a language model on genome sequences and used it to generate novel designs, moving beyond isolated genetic circuits to functional, complete biological systems.
The result validates a core premise of AI-driven drug discovery and biotech: that neural networks trained on biological data can learn patterns in genetic organization well enough to propose novel sequences that retain function. Generative design now operates at one of biology's largest scales.