Our main research areas comprise some of the most interesting interdisciplinary aspects across Medical Information Sciences.
Recent advances in high-throughput technologies have led to an exponential increase in biological data (such as genomic, epigenomic and proteomic data). To find meaningful insights in such large data collections, efficient, robust and interpretable learning methods are needed.
We are interested in developing and applying new Machine Learning / statistical learning methods to solving biomedical problems and answering new biomedical questions. Our aim is to provide state-of-the-art data-driven methods for the prevention and/or treatment of infections by SARS-CoV-2, HIV, HCV, or influenza virus as well as malaria, cancer and other diseases based on clinical, genomic and epigenomic data.
All those topics can also be subsumed under the category “Machine Learning for Precision Medicine”.
To make these models possible, it is important to get access to large medical data sets. If anonymization of these data is not possible, it is hard to gather all these data in one central location. Therefore, we also develop privacy-preserving machine learning methods that keep the data private while at the same time enabling high prediction performance.