AI and the World's Hard Problems · Bioengineering & synthetic biology · Nature · 2026-10-21
Programmable design of functional proteins from natural language
Pinal, a 16-billion-parameter model trained on 1.7 billion synthetically annotated protein–text pairs, designs candidate proteins from natural-language descriptions of desired function. Across four design tasks—a fluorescent protein, a polyethylene terephthalate hydrolase, an alcohol dehydrogenase, and a metabolic H-protein—the intended function was experimentally observed, including catalytic activity for both designed enzymes. Without iterative experimental optimization, a Pinal-designed H-protein increased product titre 1.7-fold relative to the native E. coli H-protein in a multi-enzyme CO2 fixation pathway.