Bioname: ngv circuit build recipes¶

Bioname¶

bioname is the folder where all the circuit building datasets are stored.

An ngv circuit bioname includes the following files:

MANIFEST.yaml
cell_placemement.yaml
gliovascular_connectivity.yaml
endfeet_area.yaml
microdomains.yaml
synthesis.yaml
tns_context.json
tns_distributions.json
tns_parameters.json
astrocyte_gap_junction_recipe.xml

MANIFEST.yaml¶

This is the main configuration for the circuit build defining:

atlas space
vasculature morphology and mesh
base neuronal circuit
individual task parameters

It starts with a common section:

common:
  parallel: true
  log_level: 'WARNING'
  seed: 0
  atlas: '/gpfs/bbp.cscs.ch/project/proj62/NGV/atlases/O1-230/20190517'
  vasculature: '/gpfs/bbp.cscs.ch/project/proj62/NGV/ARCHNGV/Data/vasculature_datasets/raw_pruned_cap_circuit_coo_inscribed_spec_atlas_translated_new_spec.h5'
  vasculature_mesh: '/gpfs/bbp.cscs.ch/project/proj62/NGV/ARCHNGV/Data/vasculature_datasets/raw_pruned_cap_circuit_coo_inscribed_spec_atlas_translated_triangulated_zup_fixed.obj'
  base_circuit: '/gpfs/bbp.cscs.ch/project/proj62/Circuits/O1/20190912_spines/CircuitConfig'
  base_circuit_sonata: '/gpfs/bbp.cscs.ch/project/proj62/Circuits/O1/20190912_spines/sonata/circuit_config.json'
  base_circuit_cells: '/gpfs/bbp.cscs.ch/project/proj62/Circuits/O1/20190912_spines/sonata/networks/nodes/All/nodes.h5'
  base_circuit_connectome: '/gpfs/bbp.cscs.ch/project/proj62/Circuits/O1/20190912_spines/sonata/networks/edges/functional/All/edges.h5'

and follows with separate sections for each phase.

We’ll provide a short description for eac of the common values here.

parallel: Flag for enabling parallel running.
log_level: Logging level to be used for the build.
seed: Random seed to be used for the build.
atlas: Folder path of atlas
vasculature: Path to vasculature morphology.
vasculature_mesh: Path to vasculature watertight surface mesh.
base_circuit: Path to the CircuitConfig file of the neuronal circuit that will be used for the ngv building.
base_circuit_sonata: Path to the sonata config of the neuronal circuit that will be used for the ngv building.
base_circuit_cells: Path to the sonata node population file of neurons.
base_circuit_connectome: Path to the sonata edge population file of the synaptic connectivity.

assign_emodels¶

Assign emodels parameters determine the electrical models that are going to be used in the building.

assign_emodels:
  templates_dir: '/gpfs/bbp.cscs.ch/project/proj62/NGV/emodels'
  hoc_template: 'astrocyte'

templated_dir: Directory for the electrical models
hoc_template: Template that will be used for astrocytes

cell_placement¶

Cell placement parameters determine the spatial density and disperion of astrocytic somata.

density: '[density]astrocytes'

soma_radius: [5.6, 0.74, 0.1, 20]

Energy:
    potentials:
        spring: [32.0, 1.0]

MetropolisHastings:
    n_initial: 10
    beta: 0.01
    ntrials: 3
    cutoff_radius: 60.0

density: The nrrd density dataset to choose from the atlas folder.
soma_radius: Normal distribution of soma size (mean, std, min, max)
Energy: The repulsion potential to use for evenly spacing astrocytic somata
MetropolisHastings: Parameters for the Metropolis-Hastings algorithm.

microdomains¶

Microdomains are the bounding polygons that determine the extent of astrocytes.

overlap_distribution:
  type: normal
  values: [0.1, 0.001]

overlap_distribution: Distribution of the overlapping volume of the microdomains.

synthesis¶

Morphology synthesis of astrocytes. tns_parameters, tns_distributions, tns_context will be explained below. Here only the perimeter statistic model is determined.

perimeter_distribution:
    enabled: true
    statistical_model:
        slope: 2.060005867796768
        intercept: 1.0219733661696733
        standard_deviation: 1.1161359624857308
    smoothing:
        window: [1.0, 1.0, 1.0, 1.0, 1.0]

perimeter_distribution: The statistical model to predict the perimeters on the synthesized morphologies based on their diameters.

gliovascular_connectivity¶

Gliovascular connectivity establises connections between astrocytes and the vasculature based on the microdomain of each astrocyte. Morphology has not been generated in this step.

graph_targeting:
    linear_density: 0.17

connection:
    reachout_strategy: 'maximum_reachout'
    endfeet_distribution: [2, 2, 0, 15]

graph_targeting*: During graph targeting points are evenly distributed on the skeleton of the vasculature, with a linear density specified above.
connection: To establish connection between astrocytes and the vasculature, the reachout strategy needs to be specified as well as the distribution of endfeet per astrocyte.

endfeet_area¶

Parameters for the surface growing of astrocytic endfeet.

fmm_cutoff_radius: 1000.
area_distribution: [200.0, 10.0, 0.0, 1000.0]
thickness_distribution: [1.0, 0.1, 0.01, 2.0]

fmm_cutoff_radius: The maximum radius that the growing of an endfoot with stop.
area_distribution: Distribution of the endfeet areas for pruning (mean, sdev, min, max).
thickness_distribution: Distribution of the thickness of each astrocyte endfoot (mean, sdev, min, max).

tns_context.json¶

The context specifies the spatial parameters that will be used in morphology synthesis.

{
  "field": {
    "type": "logit",
    "slope": 0.11832134,
    "intercept": 0.36720545
  },
  "space_colonization": {
    "kill_distance_factor": 15.0,
    "influence_distance_factor": 45.0
  }
}

The space colonization algorithm uses an attraction field for each endfoot target, so that processes grow towards the vascular sites. Furthermore, the kill_distance_factor and influence_distance_factor after multiplied with the segment length, determine the distance for removal and attraction of resources respectively.

tns_distributions.json¶

Distributions include the barcode sampling and statistical distributions required for morphology synthesis.

tns_parameters.json¶

The parameters for morphology synthesis.

{
  "basal": {
    "randomness": 0.3,
    "targeting": 0.01,
    "radius": 0.3,
    "orientation": null,
    "growth_method": "tmd_space_colonization",
    "branching_method": "bio_oriented",
    "modify": null,
    "step_size": {
      "norm": {
        "mean": 0.1,
        "std": 0.001
      }
    },
    "metric": "path_distances",
    "tree_type": 3,
    "filtration_metric": "path_distance",
    "barcode_scaling": false
  },
  "apical": {},
  "axon": {
    "randomness": 0.25,
    "targeting": 0.07,
    "radius": 0.3,
    "orientation": null,
    "growth_method": "tmd_space_colonization_target",
    "branching_method": "bio_oriented",
    "modify": null,
    "step_size": {
      "norm": {
        "mean": 0.1,
        "std": 0.001
      }
    },
    "metric": "path_distances",
    "tree_type": 2,
    "filtration_metric": "path_distance",
    "bias": 0.9,
    "barcode_scaling": false
  },
  "origin": [
    0.0,
    0.0,
    0.0
  ],
  "grow_types": [
    "basal",
    "axon"
  ],
  "diameter_params": {
    "mtypes_file": null,
    "models": [
      "astrocyte"
    ],
    "neurite_types": [
      "basal",
      "axon"
    ],
    "terminal_threshold": 2.0,
    "taper": {
      "max": 1.1,
      "min": -1.1
    },
    "asymmetry_threshold": {
      "axon": 0.05,
      "basal": 0.05
    },
    "n_samples": 1,
    "seed": 0,
    "trunk_max_tries": 100,
    "n_cpu": 1,
    "method": "external"
  }
}

astrocyte_gap_junction_recipe.xml¶

This is the touchdetector’s recipe for establishing gap junctional touches between astrocytes and neighboring astrocytes.