We utilize, design, and implement algorithms solving problems of computational molecular biology. More specifically, our goal is to contribute to a deeper understanding of proteins observed in nature. This is why we concentrate on two problems, which are rational protein design and enzyme evolution.
For rational protein design, we use state-of-the-art software suites like Rosetta, comprehensive databases like InterPro or PDB, and specific algorithms, which we implement if specific tools are not available. Using these methods, we can plan wet-lab experiments together with our collaboration partners.
We are also interested in elucidating protein evolution by means of phylogenetic methods. To study the evolution of enzymes, we reconstruct the corresponding sequence of predecessors and characterize the properties of the resurrected proteins in collaboration with Prof. Reinhard Sterner and his the group.
Our focusing on protein design and evolution is also setting the direction of software design. The tools, which we develop, are often dedicated to the characterization of functionally or structurally important residues. In many cases, we utilize knowledge-based potentials deduced from large data sets like multiple sequence alignments or protein structures to corroborate characteristic properties of proteins or residue-positions. Thus, we deduce properties that can contribute to a classification by means of support vector machines or other state-of-the-art classifiers.
In addition, we analyse large data sets to corroborate the specificity of enzymes or to predict function of so far uncharacterized proteins.
Our webserver offers some of our in silico methods.