region-grower

Region Grower

Synthesize neuronal morphologies in a given spatial context provided by an atlas.

Introduction

This package provides a general tools to synthesize cells in a given spatial context (a brain Atlas) with embarrassingly parallel computation (with distributed computing via dask-mpi). This package is difficult to use on its own, as it requires several specific inputs from atlas and synthesis. It is advised to use it via synthesis-workflow, which is a workflow generating most of the inputs of region-grower.

Installation

This package should be installed using pip:

pip install region-grower

Usage

This package provides one main command which provides several tools. Two tools to generate input parameters and input distributions and another to synthesize the cells.

Generate distributions

Generate the TMD distribution file:

region-grower generate-distributions --help

Input data

The command region-grower generate-distributions needs the following inputs:

• a folder containing the cells which are used to generate the parameters (see the input_folder and --ext parameters).
• a dat file (see the dat_file parameter) with 3 columns:
  • morphology names,
  • any integer value (this column is not used by region-grower)
  • mtypes.
• optionally a JSON file containing the specific parameters used for diametrization (see the --diametrizer-config parameter).

Output data

The command region-grower generate-distributions will create the following outputs:

• a JSON file containing the formatted distributions (see the --parameter-filename parameter).

Generate parameters

Generate the TMD parameter file:

region-grower generate-parameters --help

Input data

The command region-grower generate-parameters needs the following inputs:

• a folder containing the cells which are used to generate the parameters (see the input_folder and --ext parameters).
• a dat file (see the dat_file parameter) with 3 columns:
  • morphology names,
  • any integer value (this column is not used by region-grower)
  • mtypes.
• optionally a JSON file containing the specific parameters used for diametrization (see the --diametrizer-config parameter).

Output data

The command region-grower generate-parameters will create the following outputs:

• a JSON file containing the formatted parameters (see the --parameter-filename parameter).

Synthesize cells

Synthesize morphologies into an given atlas according to the given TMD parameters and distributions:

region-grower synthesize-morphologies --help

Input data

The command region-grower synthesize-morphologies needs the following inputs:

• a sonata file containing the positions of the cells that must be synthesized.
• a JSON file containing the parameters used to synthesize the cells (see the --tmd-parameters parameter). This file should follow the schema given in :ref:Parameters.
• a JSON file containing the distributions used to synthesize the cells (see the --tmd-distributions parameter). This file should follow the schema given in :ref:Parameters.
• a TSV file giving which morphology should be used for axon grafting and the optional scaling factor (see the --morph-axon parameter). The morphologies referenced in this file should exist in the directory given with the --base-morph-dir parameter.
• a directory containing an Atlas.

Output data

The command region-grower synthesize-morphologies will create the following outputs:

• a sonata file containing all the positions and orientations of the synthesized cells (see --out-cells parameter).
• a directory containing all the synthesized morphologies (see --out-morph-dir and --out-morph-ext parameters).
• a YAML file containing the apical point positions (see --out-apical parameter).
• a YAML file containing the Neuron IDs of the sections containing the apical points (see --out-apical-nrn-sections parameter).

Funding & Acknowledgment

The development of this software was supported by funding to the Blue Brain Project, a research center of the École polytechnique fédérale de Lausanne (EPFL), from the Swiss government's ETH Board of the Swiss Federal Institutes of Technology.

Copyright (c) 2022-2024 Blue Brain Project/EPFL