AI Transforms Enzyme Discovery

Enzymes power essential processes, from digestion to drug development, and scientists have long sought ways to engineer better ones for sustainability and medicine. Now, a team of bioengineers and synthetic biologists has developed an AI-driven workflow that designs, predicts, and tests new enzymes all on a computer. Their study, published in Nature Communications, marks a major breakthrough in enzyme engineering.  

We can now create highly active enzyme variants in days instead of months, without relying on living cells, said  Michael Jewett , professor of bioengineering at Stanford University and senior author of the study.  

A Faster Approach to Enzyme Engineering 

Traditionally, modifying enzymes required time-consuming lab experiments, testing thousands of variants to find one with the right function. Jewett’s team uses machine learning to predict the best mutations, drastically accelerating the process. Their cell-free synthesis system further speeds up development, eliminating the need for living organisms.  

Future Potential & Challenges

Beyond pharmaceuticals, AI-designed enzymes could:  
Break down toxins
Improve food production  
Make industrial processes greener and more efficient  

However, limited functional data remains a challenge. AI needs large datasets, but enzyme data is scarce, Jewett noted. His team analyzed 3,000 enzyme mutants and 10,000 reactions, but future breakthroughs will require even more data.  

This is just the beginning,Jewett said. With AI, we can unlock enzymes that revolutionize sustainability, medicine, and beyond. 

Source: https://engineering.stanford.edu/news/hunt-new-and-better-enzymes-ai-steps-fore