The first versions of the cerebellar neuron models can be found on the Brain Simulation Platform.

In detail:

- the code for automatic fitting of the mouse mono compartmental granule cell model (Masoli et al., 2017).

- the code for automatic fitting of the mouse multi compartmental granule cell model (Masoli et al., 2017).

These two models can already be launched from the Collaboratory, producing optimised granule cell models.

- the code of the mouse Purkinje cell model (derived from Masoli et al., 2015; Masoli and D'Angelo, 2017).

- the code of the mouse Stellate cell model (Rizza, Locatelli, Masoli, Munoz, D'Angelo, in preparation).

- the code of the mouse Golgi cell model (Rizza, Locatelli, Masoli, Munoz, D'Angelo, in preparation).

- the code of an initial mouse deep cerebellar nuclear cell model (Masoli and D'Angelo, in preparation).

- the first scaffold model of the mouse cerebellum, cell placement and connectome (Casellato, Casali, Marenzi, Medini, D'Angelo).

- a dedicated pipeline for data communication with the Neuroinformatics Platform and for the use of HPC resources in running these models (PRACE cerebellum).

You can access them via the Cerebellum Collab.

If you do not have an account to access the Brain Simulation Platform, please email bsp-support@humanbrainproject.eu



A list of recent and relevant publications:

The cerebellum is of great importance because it contains the second major cortex of the brain and has specific microcircuit organisation and plasticity dynamics. The cerebellum plays an essential role in numerous cerebrocortical loops involved in sensori-motor control and cognition. A detailed model of the cerebellum is needed, not just to understand microcircuit processing and to generalise modelling procedures, but also to implement large-scale brain simulations and robotic control systems. Nonetheless, critical experimental information is still missing, mostly on the side of microcircuit spatio-temporal dynamics, as well as on the molecular and cellular properties of neurons and on their synaptic connectivity.

What makes the cerebellum special?

The cerebellum is known to be involved in a wide set of apparently disparate functions: from motor coordination and learning to visual attention; from timing to sequence learning and generation; and from sensory prediction to Theory of Mind. The inner mechanisms connecting them all within the context of cerebellar functioning still needs to be understood.

What are the specific questions we want to address in the HBP?

The cerebellar microcircuit has been the workbench for theoretical and computational modelling since the beginning of neuroscientific research. The regular neural architecture of the cerebellum inspired different solutions to the long-standing issue of how its circuitry could control motor learning and coordination. Computational modelling in the HBP therefore plays a major role in developing new theories about the cerebellum and brain functioning.

What is our specific take?

We have developed a scaffold model of the cerebellar network along with detailed models of its neurons and synapses. These models are critical for implementing the Brain Simulation Platform and for explaining experimental data produced by the HBP Subproject on Mouse Brain Organisation. This multiscale modelling strategy aims at the generation of the first, full realistic model of the cerebellar network. The cerebellar model, in collaboration with the partnering project "cerebNEST", is also going to be simplified and used in the Neurorobotics Platform and in the Neuromorphic Computing Platform, to generate novel controllers and computational systems.



In June 2017, as part of the Brain Simulation Platform (BSP), we released on the Collaboratory several cerebellar neuron models and from the end of March (2018) released the full cerebellar scaffold model, including cell placement and connectome.

In the next period of activity, the aim is to build on the scaffold model developed in the previous phase, in order to complete:

  • the missing microcircuit components
  • to update the mechanisms at the basis of these models according to the evolving physiological and anatomical discoveries
  • to embed subcellular components (such as synaptic plasticity mechanisms) into the scaffold model
  • and to simplify the latter and embed it into large-scale brain loops.

These BSP models will be used to investigate multiscale cerebellar circuit functions and for applications to neurorobotics. Moreover, these models will serve as the first step to activate the "Cerebellum Collab", which will aggregate the cerebellum modelling community around the BSP and the HBP. 


Who is Involved

Building a model of cerebellum from experimental data will improve our understanding of the cerebellar function. It requires a critical mass of people and expertise.

The following people are driving this effort:

Egidio D'Angelo, University of Pavia, Italy 

Claudia Casellato, University of Pavia, Italy 

Simona Tritto, University of Pavia, Italy 

Stefano Masoli, University of Pavia, Italy 

Stefano Casali, University of Pavia, Italy 

Martina Rizza, University of Pavia, Italy 

Elisa Marenzi, University of Pavia, Italy 

Chaitanya Medini, University of Pavia, Italy 

Lisa Mapelli, Francesca Prestori, Francesca Locatelli, Teresa Soda, Marialuisa Tognolina, Letizia Casiraghi, Giuseppe Gagliano, Ileana Montagna and Anita Monteverd are providing the critical data required for the model reconstruction and validation.


Benefits to the Community

Model use: in the Cerebellum Collab you will find the Data Centre with detailed information about the cerebellar network, the neurons and the connectivity rules.

Results of single cell models derived from optimisation procedures can be found in:

  • the Single Cell Optimisation (Purkinje cells, Golgi cells and granule cells): currently, only the granule cell optimisation procedure code is available.
  • the Granular Cell synthesis: a Jupyter notebook with python code for connecting granule cell dendrites to mossy fibre terminals is available.
  • the Circuit Building Pipeline: a Jupyter notebook with python code for the reconstruction of a network volume is available; you can simply go to the Brain Simulation Platform and launch one of the respective use cases.
  • Alternatively, you can download data and models as indicated under Resources.

Participate to community modelling: we would be happy to hear from you, if you would like to get involved and contribute to our community effort. For all information, please contact Simona Tritto.