A unique combination of state-of-the-art bioinformatics methods, artificial intelligence, directed evolution, protein design, biocatalysis and biocatalytic process development will be applied and further developed in this project to overcome the current bottlenecks in the industrial application of glycosyltransferases. Methods for a better understanding and prediction of sequence-structure-function relationships will be applied and improved to rationally select or predict the most suitable glycosyltransferases for specific reactions. In addition, rational methods are developed to ensure high recombinant overexpression of glycosyltransferases.
The project is funded by Eurostars, a funding program for research-driven SMEs. The project duration is three years.
c-LEcta GmbH, Leipzig, Germany
builds and characterizes a collection of glycosyltransferases.
Innophore GmbH, Graz, Austria
uses their Catalophore™ platform based on artificial intelligence to search for suitable enzyme structures.
Prof. Martin Weigt from the Sorbonne University (Paris, France)
uses statistical sequence models (Direct Coupling Analysis, DCA) to predict enzyme functional properties from sequence, and to propose optimized sequence.
ass. Prof. Per-Olof Syrén (KTH, Stockholm, Sweden)
will use ancestral sequence reconstruction to design enzymes that may have existed millions of years ago and have beneficial properties.