singlecell_input_mapper ❭ map_singlecell_inputs

map_singlecell_inputs

Map synapses onto a postsynaptic cell.

This module provides a full pipeline for creating dense connectome models of the rat barrel cortex, based on methods and data presented in Udvary et al. [13].

This runfile assumes you have downloaded and extracted the barrel cortex model data from https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/JZPULNa. If this is not the case, please consult installer/download_bc_model and extract.

Attention

This file is specific to the barrel cortex model data. If you want to use it for other data, you need to adapt the paths to the data accordingly. This runfile can serve as a template.

Inputs:

  • Morphology of the post-synaptic neuron

  • 3D field of synapse densities or synapse counts.

  • Number of cells per cell type in the brain area of interest.

  • Connections spreadsheet containing PST length/area constants of the post-synaptic cell for normalization.

  • Bouton locations of individual axon tracings.

Attention

This runfile has default values for the barrel cortex, and so assumes that you have downloaded and extracted the barrel cortex model data from https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/JZPULNa. If this is not the case, please consult installer/download_bc_model and extract, or adapt the paths in this file to your data.

This module then uses NetworkMapper to assign synapses to a single post-synaptic cell morphology, based on the inputs mentioned above. This happens according to the following pipeline:

  1. The bouton density field and PST density fields are converted to scalar fields with defined voxel resolution.

  2. Calculates the overlap between these voxels and the dendrites of the postsynaptic neuron morphology using Liang-Barsky clipping [14]. Only these voxels are further considered for potential synapses.

  3. Calculates a synapse density field by multiplying the bouton density field with the PST density fields at these voxels.

  4. Normalizes the previous synapse density fields using cell-type specific PST length/area constraints and the number of cells per cell type.

  5. Poisson samples synapses from this normalized synapse density field to realize synapses. These are randomly placed onto the dendritic branch within that voxel. One such sample is called an “anatomical realization”.

  6. (optional) Repeat steps 4 and 5 to create a collection of anatomical realizations.

Density meshes are accessed using ScalarField. SynapseMapper makes use of SynapseDensity for steps 2, 3 and 4, and finalizes step 5 by itself.

Outputs:

  • summary file containing information about number and presynaptic type and column of anatomical synapses

  • AmiraMesh landmark file containing 3D synapse locations of anatomical synapses of each presynaptic type and column

  • Synapse location (.syn) and connectivity (.con) file compatible with simrun.

Functions

map_singlecell_inputs(cellName, cellTypeName, nrOfSamples, numberOfCellsSpreadsheetName, connectionsSpreadsheetName, ExPSTDensityName, InhPSTDensityName, boutonDensityFolderName)

Map inputs to a single cell morphology.