Enzymes are biological catalysts that accelerate chemical reactions. They are crucial for applications such as pharmaceutical production, health diagnostics, and sustainable materials recycling. Optimising enzymes can improve their efficiency, adaptability to various conditions (e.g. temperature), sustainability, and more. However, enzyme engineering is extremely time-consuming and costly.

Recently, scientists have started using generative AI to create new protein designs, which could boost efficiency, precision, and innovation in the field. In a new research project of the UvA Molecular and Materials Design Technology hub (MMD TechHub), scientists aim to build on this development and create a new method using generative AI for designing enzymes.

More targeted approach

The UvA researchers plan to create a method that can sample new enzyme designs and experimentally validate the results. They will use various components of existing generative AI methods and also create new methods themselves. According to Aalt-Jan van Dijk, professor of Data Analysis in the Life Sciences and lead scientist of the project, the existing generative AI methods on protein design are not very good at incorporating specific biological knowledge.

Van Dijk: ‘We want to incorporate more knowledge on what the protein should look like and the function that it should have.’ This more targeted approach could make it easier to optimise enzymes with a specific goal in mind, for example, increased efficiency in a specific chemical reaction. Biotech companies and universities could use the method to create better enzymes in a time-efficient manner.

Interdisciplinary team 

To develop this method, the UvA research group consists of scientists from computer science, life science and chemistry. One of the goals of the MMD TechHub is to bring researchers from various disciplines together. Van Dijk: ‘I'm very enthusiastic about the collaboration between these different disciplines. The opportunity to do this kind of collaboration is the reason I decided last year to work at the UvA Faculty of Science’.