Investigating the organization of the mossy fibers with an automatized neuroinformatic pipeline

This work aims at investigating the organization of the mossy fibers (mf), which provide a rich excitatory drive to the cerebellar cortex. They originate from the spinal cord and from a wide range of nuclei in the brain stem. Their arrangement is partially somatotopic. They contact granule cells which are responsible for expansion recoding and representations of information in the cerebellar cortex. The synaptic organization between the mossy fibers and the granule cells form the input stage of the cerebellum and represent a large part of its vast coding capabilities. The importance of the mossy fiber to granule cell layer is highlighted by its prominent role throughout the historical understanding of the computational properties and models of the cerebellum, but a systematic description of their exact growth and innervation patterns remains unkown. We have developed a neuroinformatic pipeline which i) creates a skeleton representation of the mossy fiber nerve bundles present in experimental images of cerebellar slices; ii) defines an initial arrangement of fiber paths and glomeruli distribution within the volume of the bundle; iii) connects glomeruli with granule cells (placed in the volume), following biologically-grounded connectivity principles; iv) generates the activity of granule cells (simulated 2D map of activity) depending on the bundle stimulations; v) compares the simulated activity with the experimental map coming from Multi-electrode Array (MEA) recordings of the cerebellar slices under electrical stimulations (Local field potentials, LFP); vi) optimizes the branching model parameters to maximize the matching between simulated and recorded activity maps. Our preliminary results suggest that the electrical stimulation propagates for 40\% of the bundle and activates about 40\% of the fibers inside that area. The developed pipeline is designed to find optimal solutions for any given branching model, with different parameter sets. This will allow to shed light on the structural branching of the fibers, responsible for the functional spatio-temporal patterns of activity.