Our group studies the design and application of novel data mining and machine learning techniques, motivated by specific questions in biology and medicine. To this end, our group combines the expertise of both strong analytical skills, exemplified by solid backgrounds in applied mathematics, computer science and engineering, with expertise in applied bioinformatics.
In the field of systems immunology, our group develops new computational approaches to unravel the regulatory landscape of immune cell differentiation and functioning. High-throughput methods such as microarrays, next-generation-sequencing (NGS), multiplexed flow cytometry and imaging are currently revolutionizing the field of immunology, allowing us to study cells and their interactions into unprecedented depth. While these technologies are able to generate massive amounts of data on cell behavior and functioning, interpreting these data and making sense out of it is currently the next challenge. Our group develops novel systems biology approaches to model the regulatory landscape of immune cells using module networks and computational flow cytometry.
On the more algorithmic side, our research group develops new machine learning approaches to deal with challenging extensions of the classical learning paradigms, including high-dimensional, small sample settings, semi-supervised learning, imbalanced data, and structured input and output representations. Our group has a solid expertise in the development of feature selection (biomarker selection) algorithms, and is currently exploring the potential of these techniques for new data types, such as flow cytometry and imaging. A great part of our research effort also goes to the analysis of integrated “omics” approaches, where we take a network approach to integrate various data sources, and make use of novel graph mining approaches to formulate biological questions as machine learning questions on graphs.
In addition to our methodological research, we also aim to provide the scientific community with freely available, easy-to-use webtools, databases and other publicly available resources that result from our newly developed algorithms. An overview of our tools can be found here