Current topics include
Score functions for comparing protein structures
Score functions, potential energy functions and even the very long-titled "protein sequence to structure compatibility functions" can be used to provide something like a fingerprint for a protein. We are using them to cluster and classify proteins.
Modelling of real proteins
Our collaborators, mainly in the Bernard-Nocht Institute have been trying to teach us some real biology and we have been working on proteins from lassa virus and malaria.
Protein sequence optimisation
We would like to build better proteins than nature. Fortunately, we can define what we mean by better (more stable, with special amino acids...). We are interested in ways to sample the possible sequences for a protein structure and also in the mathematical functions which measure this.
Applications of cunning Monte Carlo and other sampling methods
Methods like Monte Carlo or classical molecular dynamics simulation are our roots. These days, we are applying new methods with special sampling properties (more efficient or in different ensembles). The intention is to apply the methods in classical physical spaces, but also to some of the more abstract spaces in other projects.
Protein threading and structure prediction
We are interested in how best to thread a sequence of unknown structure onto the coordinates of a known structure. This means both the threading algorithms and how best to build the score functions one needs.
Epidermal Systems Biology
Goal of the project is to develop systems biological models able to predicted properties of epidermal tissue in healthy and diseased state. The models are validated by their ability to predict epidermal morphology, tissue kinetics and molecular concentrations reported in literature.
Link to project.