Our group uses computational methods based on physical principles for the study of the structure and dynamics of proteins and nucleic acids. We are especially interested in the following areas:

Membrane-bound proteins are crucial components in many biological processes. However, relatively little detailed structural and dynamical information is available, as the heterogeneous nature of biological membranes presents challenges for experiments as well as computer-based modeling efforts. We are applying new methods that represent the membrane environment implicitly, in order to study the dynamics and energetics of membrane proteins such as the vitamin B12 ATP-binding cassette (ABC) transporter shown on the right. more information

In biological systems, proteins and nucleic acids interact during gene duplication, transcription, and regulation. On a molecular level, many details of these processes are not yet well understood. In particular, we are interested in the question of how the DNA mismatch recognition protein MutS and its eukaryotic homologues recognize defects in newly replicated DNA and how DNA repair is subsequently initiated. A detailed understanding of this DNA mismatch repair system is relevant in finding cures for some types of cancer that have been linked to defective mismatch repair. We are using computer simulations to investigate structural deformations in this large complex during mismatch DNA binding. more information

The accurate prediction of protein structures from its amino acid sequence remains a serious challenge to computational methods. It has become possible, however, to generate native-like models in many cases, when structural templates are available from related proteins through sequence homology or fold recogniton. We are interested in developing new methods that allow the refinement of approximate structure predictions near the level of experimental accuracies. Our efforts, in this respect, are based on a combination of enhanced sampling methods with force-field based scoring functions. more information

An overarching theme in our group is the development and application of realistic implicit solvent models. Compared to explicit solvent representations, implicit solvent methods are computationally much more efficient. In particular, our efforts are based on generalized Born theory, which describes the solvation free energy of a set of partial charges surrounded by a continuum high-dielectric environment. We are involved in tuning and evaluation of such methods for simple aqueous solvent as well as heterogeneous dielectric environments in comparison with experimental data and explicit solvent simulations. We are also pursuing novel approaches for enhanced sampling of biomolecules by combining molecular representations of proteins at different levels of detail. more information

We are building the infrastructure for a molecular dynamics trajectory database, where simulation data is available for public access through a flexible set of analysis functions. One of the main motivations for the SimDB project is the ability to compare dynamical features between different simulations of the same or similar biomolecules. We are especially interested in using such comparison to learn more about the extent to which dynamical features are conserved in protein families with similar function and ultimately how protein dynamics are relevant for biological function. more information