Our group studies the genomes of entire microbial communities, based on the new and exciting metagenomics approach. Metagenomics describes a technique that encompasses the sampling of DNA directly from an environment; the DNA is then shotgun sequenced without prior knowledge of what organisms are actually present. This approach has become necessary because most naturally occurring microbes cannot be grown in the laboratory, which has lead to a situation where our molecular knowledge of microbes is vastly incomplete and quite biased. Only recently, researchers have begun to fill this gap by cultivation-independent molecular techniques such as metagenomics sequencing. Our group uses computers to analyze the vast amounts of molecular sequence data this produces. We ask questions such as: what types of organisms can be found by such an unbiased approach? What molecular functions do they encode? How can we describe microbial biodiversity, and does it follow the same ecologial principles as macroscopic life does?

Another research area that we are very interested in concerns protein-protein interactions, and the networks they form. We develop algorithms that help to score and integrate protein interaction data, and make them more easily accessible for researchers to search and browse. We also develop procedures to transfer interaction knowledge between model organisms, and to predict interactions from genomic context. We are particularly interested in comparing interaction networks across species, in order to learn how they evolve and at which organizational level selection is acting. As part of our interest in protein networks, we are participating in the STRING consortium, which maintains one of the most comprehensive public network resources to date (see here).