Artificial Intelligence (AI), Synthetic Biology, and Robotics Combine to Improve Enzymes
A user-friendly system allows scientists to drastically improve enzyme performance
The Science
A team of researchers at the Center for Advanced Bioenergy and Bioproducts Innovation (CABBI) created a fast, user-friendly process to improve industrial enzymes. The researchers combined artificial intelligence (AI), synthetic biology, and a biofoundry. Biofoundries are laboratories that integrate robotics, computer-aided design, and informatics. This combination allowed the scientists to overcome common issues with enzymes such as low efficiency and an enzyme’s inability to single out a target. The team used this new process to greatly improve the performance of two important industrial enzymes. The process is versatile enough for scientists to use it to improve many others as well. The work is a big step forward in the use of AI for enzyme engineering. In enzyme engineering, scientists predict what changes to a known protein would improve its function.
The Impact
Engineering proteins to have a desired function or improve their efficiency is a slow and expensive process. It requires a specialized research team. This new platform for enzyme engineering minimizes the need for significant human intervention and specialized expertise. This work will accelerate advancements across diverse industries supporting the bioeconomy. As a result, it will add value to agriculture, grow rural economies, and increase domestic energy production.
Summary
A team at the Center for Advanced Bioenergy and Bioproducts Innovation (CABBI) used robots and AI to move towards developing a fully self-driving laboratory. First, the researchers ask the AI tool how to improve a desired enzyme’s performance. The AI tool searches datasets of known enzyme structures and suggests sequence changes. The automated protein-building machines, or biofoundries, produce the suggested enzymes. The machinery then rapidly tests the enzymes to characterize their functions. Scientists feed data from those tests into another AI model. This model uses this information to improve the next round of suggested protein designs. In a case study with two industrially relevant enzymes, researchers increased enzyme activity by 16 and 26 times with greater speed and less human intervention than ever before.
More broadly, the work contributes to the development of a completely self-driving lab. This is an AI-powered lab that designs, builds, tests, and learns the protein, all before designing the next cycle. By reducing the time, labor, and costs involved, this research accelerates automated enzyme engineering and supports the transition to a more bio-based economy.
Contact
Huimin Zhao
Center for Advanced Bioenergy and Bioproducts Innovation
University of Illinois Urbana-Champaign
zhao5@illinois.edu
Funding
The work was supported by the Center for Advanced Bioenergy and Bioproducts Innovation, a U.S. Department of Energy (DOE) Bioenergy Research Center supported by the Biological and Environmental Research (BER) program in the Office of Science.
Publications
Singh, N., et al. “A generalized platform for artificial intelligence-powered autonomous enzyme engineering.”Nature Communications 16, 5648 (2025). [DOI: 10.1038/s41467-025-61209-7]
Related Links
Self-driving lab: AI and automated biology combine to improve enzymes, University of Illinois Urbana-Champaign News Bureau
Highlight Categories
Program: BER
Performer: University
