Extracting rules of behavior in collective tumor cell systems

• Applicant: Privatdozent Dr. Claus Metzner
• Subject Area: Biophysics
• Term: from 2017 to 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 347962689

According to the assumption of weak emergence, all collective properties of a system can be computed, at least in principle, from the properties of the system components and their interactions. For most physical many particle systems, the single particle properties and interaction forces are known precisely. In more complex systems, such as bird flocks and fish schools, the rules of individual behavior in the context of a group are currently not well understood. Researchers have therefore constructed rules of individual behavior that could reproduce some of the observed collective system properties in simulations. However, it remains unclear whether the same set of rules would also correctly describe entirely new situations that were not present in the 'training data set'.Here, we propose an automatic machine-learning method to extract from video recordings the rules of behavior in aggregates of tumor cells on planar substrates and in 3-D tissue-like environments. The result will be a quantitative, probabilistic model of how individual tumor cells migrate and proliferate in the presence of other cells of the same type, but also in contact with other cell types. We will use the extracted model to predict the collective properties of cell colonies, in particular the shape evolution of a growing colony over time. Finally, our method will be applied to quantify the interaction between natural killer cells and tumor cells, thereby contributing to the development of an immunotherapy against cancer.

• DFG Programme: Research Grants
• Major Instrumentation: Motorized quantitative phase contrast microscope (HoloMonitor M4, Phase Holographic Imaging, Lund, Sweden)
• Instrumentation Group: 5000 Labormikroskope