Lab Profile

Machine Learning group, ICB

Coming from a focus on network biology and mechanistic modeling, the machine learning group evolved to a focus for data driven modeling of mostly single-cell transcriptomic data. This includes manifold learning and stochastic processes for mapping high-dimensional, non-linear data. Often, these techniques are combined with neural-network-based representations of these data. At the same time, the group continues to apply and develop statistical and differential equation-based model to gain insight into the role of specific variables in these data. We collaborate with various experimental groups with applications from stem cell biology, differentiation to disease modeling.

  • Manifold learning and stochastic processes for mapping high-dimensional data
  • Neural networks for imaging sequence data and – since recently – count data
  • Scaling established data analysis algorithms to millions of observations and thousands of dimensions
  • Bridging mechanistic and data-driven models
  • Network biology for multi-omics data

Institution Name

Institute of Computational Biology, Helmholtz Zentrum München

Institution Short Name


Institution Address

Ingolstädter Landstr. 1,
85764 Neuherberg, Munich, Germany

Institution Website


Principal Investigator

Fabian Theis

Director of ICB, Head of Machine Learning Group
Professor at Technical University of Munich