singlecell_input_mapper

Create an empirically constrained dense connectome model with presynaptic activity.

This package provides classes and methods to create a dense connectome model with defined activity patterns. It can be largely divided in two parts: Connectivity and Activity.

See also

This package should not be confused with single_cell_parser.

This package is specialized to create empirically consistent network models, providing fine-grained control over the network realization process, and ability to constrain it with empirical data.

single_cell_parser provides only basic functionality to recreate such network realizations from the results generated here. The purpose of single_cell_parser is to provide an API to the NEURON simulator, and read in results from network realizations.

Connectivity

singlecell_input_mapper.singlecell_input_mapper is responsible for assigning synapses to the morphology of a postsynaptic neuron, and keeping track of the synapse type and associated presynaptic cell type. Based on this presynaptic cell type, different spike times can be generated (see section Activity below). Assigning synapses onto the postsynaptic morphology is referred to as a ‘network realization’. The network realization in ISF is based on the following inputs:

  • The morphology and location of the postsynaptic neuron

  • The 3D density of post-synaptic targets (PSTs) in the brain region of interest (cell type unspecific)

  • The 3D density of boutons in the brain region of interest (cell type specific)

  • The 1D and 2D densities of PSTs onto the postsynaptic neuron per length and area (cell type specific)

See singlecell_input_mapper.singlecell_input_mapper for more information, and singlecell_input_mapper.map_singlecell_inputs.map_singlecell_inputs() for a pipeline to create anatomical realizations.

Other ways of network realization are also possible, depending on the empirical data available. If you want to use other methods, please familiarize yourself with the data formats such that you can either:

  1. Convert the input data to the required input format for ISF and run the same network realization pipeline. Input file formats are described in more detail in singlecell_input_mapper.map_singlecell_inputs.

  2. Create your own network realization, and convert the output to the format used in ISF. These are .syn and .con files. For more info on how to generate and use these, refer to the network modeling section of the Tutorials

Activity

This section is responsible for generating activity patterns for the assigned synapses based in empirically observed PSTHs of the presynaptic neurons. ISF distinguishes two kinds of activity:

  1. Ongoing activity: the baseline synaptic activity patterns in the absence of the in vivo condition of interest. The ongoing activity is defined in tandem with the network parameters in a Network parameters file.

  2. Evoked activity: the activity patterns in response to a specific in vivo condition. Its file format is described in Activity data.

The general workflow is as follows:

  1. Read in individual spike times of presynaptic neurons.

  2. Create PSTHs for each cell type for the ongoing and evoked activity. Such files are present in getting_started/example_data/functional_constraints

  3. Create a network parameter file from the PSTHs.

Modules

singlecell_input_mapper

Calculate the connectivity of individual neuron morphologies.

evoked_PSTH_from_spike_times

Create PSTHs from spike times.

evoked_network_param_from_template

Runfile to create a Network parameters file that captures the population activity of a rat barrel cortex during passive whisker touch in anasthesized animals.

map_singlecell_inputs

Map synapses onto a postsynaptic cell.

ongoing_network_param_from_template

Create a network parameter template file.