Learning protein function through autonomous experimental interaction

Abstract excerpt

Biological AI learns primarily from existing observations, but many questions cannot be answered from available data alone. Here we show that AI can instead acquire knowledge by acting directly on biological systems and learning from the consequences. We developed a closed-loop framework in which autonomous agents design protein variants, construct and characterize them in a robotic laboratory, learn from the resu...

Topics

Open a Topic to create a Post that cites this publication.

• Amino Acids
• Cost-Benefit Analysis
• Genes, Fungal
• Learning
• Open Reading Frames
• Probability
• Research Design
• Robotics

Identifiers and source

Literature Corpus work

245cfad2-d949-508b-8b75-9a478f35f848

DOI

10.64898/2026.08.14.744985