Machine Learning & Computational Biology

(Karsten Borgwardt)

Our research area was the development of intelligent algorithms for analysing complex systems in biology. It is located at the intersection of machine learning, data mining and bioinformatics, and contributes to these three fields. Our research reaches deep into statistics, algorithmics and scientific computing.

The focus of the “Machine Learning & Computational Biology” research group is algorithmic systems biology. Our research reaches deep into systems biology, bioinformatics and statistical genetics, but also into machine learning, data mining and scientific computing. We develop algorithms and statistical tests to examine the effect of single genes on a biological system. The methods we are working on are part of the field of “Machine Learning”. Machine Learning is concerned with the development of computer-based statistical procedures for finding patterns and dependencies in large volumes of data.

We develop such techniques in order to predict the function of a gene or a chemical compound. For this purpose, graph-based techniques are of utmost important. This is due to the fact that graphs can be used to model the interactions of genes and proteins, or to describe the structure of a molecule. For this reason, machine learning on graphs and networks is a central topic in our research.

Furthermore, we work on methods for Machine Learning for genome-wide association studies. Here we study whether sequence variation in the genomes of individuals leads to phenotypic variation. In a collaboration with Prof. Dr. Detlef Weigel’s department at the Max Planck Institute for Developmental Biology in Tübingen, we analyse whether genetic factors can explain the variation in flowering time in Arabidopsis thaliana. With the Max Planck Institute for Psychiatry in Munich, we explore the question whether one can predict the reaction of a depressive patient to treatment with antidepressants.