Vanderbilt team releases AI tool for enzyme design

Assistant Professor of Chemistry Zhongyue Yang at Vanderbilt University and a team of researchers have introduced a new artificial intelligence tool designed to accelerate enzyme engineering. The system, named MutexaGPT, processes plain-English instructions to generate design proposals for modified proteins. Yang, along with former doctoral student Qianzhen Shao, now a postdoc at Stanford, and other team members, developed the software to assist scientists in customizing enzymes that drive chemical reactions in daily life.

Enzymes act as biological catalysts that speed up specific chemical processes. Each enzyme is built from hundreds of amino acids. Changing a single amino acid, a process known as a mutation, can drastically alter the protein's function. Current industry methods often rely on extensive trial and error to predict useful mutations. This approach is expensive and time-consuming for a market that generates hundreds of billions of dollars in revenue annually.

MutexaGPT addresses this challenge by using large language model agents to interpret user requests. The system elicits missing information, constructs physics-based models, and converts the data into actionable design plans. Yang stated that the tool translates human intuition into workable plans rather than replacing the workforce. “We are not using AI to get things done more efficiently without people,” Yang said. “We are showing that with AI you can translate human intuition, which is often imaginative and incredible, to create a workable plan.”

The tool has several practical applications in food processing and manufacturing. It can help engineer cold-active enzymes that maintain high performance at lower temperatures. This capability is useful for producing lactose-free milk and preserving nutrients in food. The system also supports the design of laundry detergents that break down stains without hot water. In pharmaceutical manufacturing, the tool helps create enzymes with higher catalytic efficiency, which can reduce energy use and chemical waste.

The name Mutexa is an abbreviation for Alexa for Mutant. Yang explained that the name reflects the goal of creating an intelligent assistant similar to Amazon's Alexa. The system allows scientists to communicate through natural language while connecting their questions to computational tools. “Just as Alexa helps people accomplish everyday tasks through natural-language conversations, MutexaGPT aims to help scientists navigate the complex world of enzyme mutations,” Yang said. The GPT suffix indicates the integration of large language models.

Yang earned his bachelor of science from Nankai University in China and completed his doctorate at UCLA in 2018. He held a postdoctoral fellowship at MIT before joining the Vanderbilt Arts and Science faculty in 2020. His research is funded by the National Institute of General Medical Sciences and the National Science Foundation. He recently received an NSF Cyberinfrastructure Systems for Scientific Innovation grant to strengthen the infrastructure for Mutexa.

The Vanderbilt Center for Structural Biology, Data Science Institute, and Advanced Computing Center for Research and Education provided components for the tool's development. Yang noted that researchers often possess principles for how enzymes work but lack the methods to quantify them. The software translates these principles into specific designs. “I am excited to envision a future in which scientists can engineer enzymes for specific applications as readily as they consult an AI assistant,” Yang said.

The team continues to test the efficacy of MutexaGPT. The system is currently available for use by researchers in enzyme engineering and design. The funding supports the ongoing work to refine the AI models and expand their applications in medicine and sustainable manufacturing.