Statistical Neuroscience group

Statistical Neuroscience


Nowadays electrophysiological recordings enable the observation of >100 single neurons simultaneously. Such data (provided by our collaborators) allows us to get insights into neuronal interactions and their relevant temporal and spatial scales. To do so, we develop analysis strategies and tools that uncover the concerted neuronal activity in massively parallel spike trains and local field potential recordings.

On the one hand we develop statistical methods that enable to detect spike correlations and higher-order spiking activity patterns within and across cortical areas that signify assemblies which are assumed contribute to the processing of sensory inputs and lead to behavior. The findings are contrasted to results based on measures on different time scales, and are put in context to different cortical states. Local field potential analysis supports on a mesoscopic spatial scale the interactions of groups of neurons involved in bottom-up and top-down interactions. The methods developed by the group are provided in the open source toolbox “Elephant”.

Network models and network theory help to get insight into the neuronal mechanisms which lead to the phenomena we observe in experimental data. In cooperation with collaborators we set up network models and analyze their neuronal activities to reach an understanding of brain processing.

Research Topics

higher-order correlation analysis, processing dynamics, cell assemblies, integrative loop, model validation, reproducible analysis and workflows


Prof. Dr. Sonja Grün


Building 15.22 / Room 4012

+49 2461/61-9302


Research Foci

Brain Science

  • Dynamical interactions in the brain network relevant for behavior and cognition.
  • Signatures of network processing in massively parallel experimental recordings.
  • Intense interaction with experimenters.

Data Analytics

  • Development of statistical data analysis tools for activity data from awake behaving animals.
  • Data analysis to extract and condense the relevant characteristics of the system.

Integrative Loop

  • Closing the loop between neural network models and experimental data.
  • Interpretation of system dynamics through construction of theoretical (biophysical and functional) models.

Team Data Science of Electro- and Optophysiology in Behavioural Neuroscience (DSEO)

Advancing methods to manage neuronal activity data and designing robust processes that enable reproducible data analysis and rigorous validation of network simulations, the team focuses on co-designing software, services and processes in tight interaction with scientific projects.


Last Modified: 27.06.2022