Complementing our work in particular grant projects, we pursue strategic work in:

The development of computational methods for the analysis of quantitative experiments in the life sciences. We work on approaches of algorithmic, statistical, and probabilistic nature. Inference from heterogeneous data sources is a particular interest of ours. From a strategic perspective, we prefer Bayesian models.

The development and validation of novel laboratory protocols for calibrated, truly quantitative measurements of molecular characteristics in the life sciences. Concurrently, the corresponding necessary computational methods are developed. Multi-channel microarrays are currently the primary target of our efforts. We do, however, also collaborate on improving multi-channel 2D protein gel experiments.

The analysis of selected biological questions which provide increased focus for our methodological work. Taking a quantitative approach from the start we can, moreover, investigate classes of questions that are not yet successfully pursued in the life sciences. Our primary interest is in the analysis of phenomena that are intractable using traditional gene-by-gene approaches. The laboratory focusses on work with the model organism Drosophila, using this fruitfly, for example, to examine non-genetic sources of individuality and molecular mechanisms in ageing. In addition, we have experience in the analysis of clinical samples and experimental data from other organisms.

The integration of heterogeneous databases and analysis tools. We tackle this challenge mostly in collaboration, as a necessary prerequisite for many sophisticated bioinformatics analyses. We can contribute years of experience and use cases for applications from our current work. In house software development by our group is in this area, however, now typically restricted to specific applied analyses.

Selected domains of sequence analysis. We study sequence regions that have traditionally received less attention, such as regions with strong compostional bias, which do not permit the application of common tools like traditional sequence comparison algorithms.