Heidelberg Researchers Develop AI Model to Forecast Two Proteins for Gene Editing

A team of researchers from Heidelberg University has developed a new AI tool, called ProDomino, that allows forecasting of two proteins to engineer a functional, adjustable new protein. Led by Prof. Dr. Dominik Niopek from the Institute of Pharmacy and Molecular Biotechnology (IPMB), this research has a broad potential for biotechnology and medical applications.

Proteins are composed of interlinked subunits known as domains, which regulate various vital processes, including recognizing external signals and catalyzing chemical reactions. Inspired by this process, the researchers developed an AI model on the artificial recombination of protein domains using a tailored dataset of over 100,000 proteins to determine protein domain combinations that will generate merged proteins with new properties.

Using ProDomino, the team was able to engineer hybrid proteins, including CRISPR-Cas variants that can be switched on or off, improving the safety of gene editing techniques. “Our AI model allows for producing artificial proteins not only more easily but also in a more targeted way than before,” said Prof. Niopek. The researchers have made the software more accessible to scientists by making it available as an open-source tool.

For more information, read the press release from Heidelberg University.