SONATA Technical description

The following is the definition of the file format used in the SONATA file.

Sonata Files

A Sonata circuit possesses multiple files which contain the different network data. This includes at least a node file and an edge file. Each file can contain multiple populations (see below) of nodes and edges.

At BBP, we decided to use the file extension .h5 for all the sonata files in HDF5 format. That is the files that contain scientific data: the node, edge, spike report and frame report files. This follows the official recommendations. The configuration files are in JSON format and have the .json extension. That is: the circuit configuration, simulation configuration and nodesets definition.

Warning

One cannot make assumption about the file name. As opposed to legacy CircuitConfig and BlueConfig, SONATA specification do not enforce a name for circuit and simulation configuration json files.

Even if technically possible, we decided not to mix the node and edge populations inside the same file.

The name convention for the files outputted from circuit-build is :

• nodes.h5 for the node file

• edges.h5 for the edge file

Node File

Fields for Nodes

The fields for the nodes depend on the model_type of the node, which should be stored in the sonata config file in as per Sonata Node Types

Some properties can be stored in a @library group for more efficient storage use. These attributes /<population>/<group>/X are stored as an integer that maps to a corresponding attribute in /<population>/<group>/@library/X.

Fields for biophysically detailed neurons (model_type: biophysical)

Group column represents the HDF group where the dataset is located under /<population>. “/” means it is directly under /<population>.

Group	Field	Type	Requirement	Description
/0	x, y, z	float32	Mandatory	The position of the center of the soma in the local world in µm.
/0	rotation_angle_[x|y|z]axis	float32	Optional	Euler angle representation of the rotation around the given axis of the morphology around the soma in radians.
/0	orientation_[w|x|y|z]	float32	Mandatory	Preferred way to define the rotation as quaternions. Note: the quaternions are not guaranteed to be normalized
/0	morphology	utf8	Mandatory	Morphology file relative path, without file extension. Example “mymorphology” or “mypath/mymorphology”. The file format may be different depending on the consumer.
/0	layer	utf8	Optional?	Layer for the neuron.
/0	model_template	utf8	Mandatory	See details below.
/0	model_type	utf8	Mandatory	biophysical
/0	morph_class	utf8	Mandatory	Used to define the morphology classification. Whether they are “PYR” or “INT”, for example.
/0	etype	utf8	Mandatory	Defines the electrical type of the node.
/0	mtype	utf8	Mandatory	Defines the morphological type of the node.
/0	me_combo	utf8	Optional	Deprecated Replaced by dynamics_params. Foreign key to me combo tsv file.
/0	synapse_class	utf8	Mandatory	Defines the synapse type of the node; whether the neuron is inhibitory or excitatory. “EXC” or “INH”.
/0	region	utf8	Optional	Attribute assigning a brain region to the associated cell.
/0/dynamics_params	threshold_current	float32	Mandatory	The minimal amplitude (in nA) of a step current clamp injection that triggers an action potential.
/0/dynamics_params	holding_current	float32	Mandatory	The current clamp amplitude (in nA) necessary to hold the cell at a predefined holding voltage (typically around -85 mV for BBP).
/0/dynamics_params	AIS_scaler	float32	Optional	Multiplicative factor to scale AIS (morphology dependent and optional, used only for synthesis).
/0/dynamics_params	input_resistance	float32	Optional	Input resistance of the cell in MOhm.
/0	exc-mini_frequency, inh-mini_frequency	float32	Optional	Mini-frequencies are associated with incoming connections of a cell, and depend on the incoming connection’s synapse type, and the receiving cell’s layer. If the synapse type is excitatory then exc-mini_frequency is used, otherwise inh-mini_frequency is used. Default to the one provided by the circuit config if not present (Unit is Hz).
/0	hemisphere	utf8	Optional	“left” or “right” hemisphere.
/	node_type_id	int64	Mandatory	Set to -1. Foreign key to node type csv file not used at BBP..

Fields for astrocyte population (model_type: astrocyte)

Group	Field	Type	Requirement	Description
/0	x, y, z	float32	Mandatory	The position of the center of the soma in the local world in µm.
/0	radius	float32	Mandatory	The radius of the astrocytic soma in µm.
/0	mtype	utf8	Mandatory	The morphological type of the astrocyte.
/0	morphology	utf8	Mandatory	The name of the astrocytic morphology without extension.
/0	model_type	utf8	Mandatory	astrocyte JDC: seems to be incorrect in generated files..
/0	model_template	utf8	Mandatory	See details below.
/0	layer	utf8	Optional?	Layer for the neuron.
/0	region	utf8	Optional	Attribute assigning a brain region to the associated cell.
/0	hemisphere	utf8	Optional	“left” or “right” hemisphere.
/	node_type_id	int64	Mandatory	Set to -1.

This type of population requires extra datasets. These datasets are represented in the sonata config file as:

“microdomains_file”: path/to/the/microdomain/tessellation

where “microdomains_file” correspond to the microdomain tessellation (See here for full description of the microdomains dataset).

Fields for vasculature population (model_type: vasculature)

The edges/segment of the vasculature morphology are stored as a population of nodes.

Thus, the id of the vasculature population node corresponds to the i-th edge in the vascular morphology.

This representation is essential for the endfoot connection edges that require a target node which is here a particular edge of the vascular morphology.

In order, to maintain the mapping between the edge identification and the section/segment one, the section and segment ids for each edge of the vascular morphology are stored so that the section representation can be uniquely reconstructed if needed. An edge_id is unique to a vasculature morphology, a segment_id is local to a section of a vasculature morphology.

Group	Field	Type	Requirement	Description
/0	start_[x|y|z]	float32	Mandatory	The coordinate of the starting point of vasculature’s segment in the local world in µm.
/0	end_[x|y|z]	float32	Mandatory	The coordinate of the ending point of vasculature’s segment in the local world in µm.
/0	start_diameter	float32	Mandatory	The diameter of the starting point of vasculature’s segment in µm.
/0	end_diameter	float32	Mandatory	The diameter of the ending point of vasculature’s segment in µm.
/0	start_node	uint64	Mandatory	Morphology node id of the starting point of the vasculature segment.
/0	end_node	uint64	Mandatory	Morphology node id of the ending point of the vasculature segment.
/0	type	int32	Mandatory	The edge type is an integer in the list [1, 2, 3, 4, 5, 6, 7] corresponding to vein, artery, venule, arteriole, venous-capillary, arterial-capillary, transitional.
/0	section_id	uint32	Mandatory	The corresponding section id in the vasculature morphology.
/0	segment_id	uint32	Mandatory	The corresponding segment id in the vasculature morphology.
/0	model_type	utf8	Mandatory	vasculature
/	node_type_id	int64	Mandatory	Set to -1.

This type of population requires extra datasets. These datasets are represented in the sonata config file as:

“vasculature_mesh”: path/to/the/mesh “vasculature_morphology”: path/to/the/skeleton

The vasculature_mesh is the watertight representation of the vasculure produced using Ultralizer or any other tool that can produce such surface meshes. The surface mesh’s extension is .obj.

Fields for point neuron population (model_type: point_neuron)

This is for AdEx point neuron models, see Brette R. and Gerstner W. (2005) [1].

The equivalent NEST model is: https://nest-simulator.readthedocs.io/en/v3.0/models/aeif_cond_beta_multisynapse.html.

Group	Field	Type	Requirement	Description
/0/dynamics_params	C_m	float32	Mandatory	Membrane capacitance. (pF)
/0/dynamics_params	Delta_T	float32	Mandatory	Slope factor. (mV).
/0/dynamics_params	e_L	float32	Mandatory	Leak reversal potential. (mV).
/0/dynamics_params	i_e	float32	Mandatory	Constant external input current (pA).
/0/dynamics_params	v_peak	float32	Mandatory	Spike detection threshold. (mV).
/0/dynamics_params	v_reset	float32	Mandatory	Reset value for V_m after a spike. (mV).
/0/dynamics_params	v_th	float32	Mandatory	Spike initial threshol. (mV).
/0/dynamics_params	a	float32	Mandatory	Subthreshold adaptation. (nS).
/0/dynamics_params	b	float32	Mandatory	Spike triggered adaption. (pA).
/0/dynamics_params	g_L	float32	Mandatory	Leak conductance. (nS).
/0/dynamics_params	t_ref	float32	Mandatory	Duration of refractory period. (ms).
/0/dynamics_params	tau_w	float32	Mandatory	tau_w (ms). Time constant for adaptation current.
/0/dynamics_params	nb_receptors	uint32	Mandatory	Number of receptors in the postsynaptic membrane
/0/dynamics_params	e_rev	float32	Mandatory	Synaptic reverse potential. (mV). nb_receptors elements. Reverse potential.
/0/dynamics_params	tau_decay	float32	Mandatory	Synaptic decay time constant (for afferent synapses). (ms) cardinality is nb_receptors elements.
/0/dynamics_params	tau_rise	float32	Mandatory	Rise time of the synaptic conductance (for afferent synapses). (ms) cardinality is nb_receptors elements.
/0	x, y, z	float32	Mandatory	The position of the center of the soma in the local world in µm.
/0	etype	utf8	Mandatory	Electrical type of the node.
/0	mtype	utf8	Mandatory	Morphological type of the node.
/0	synapse_class	utf8	Mandatory	Defines the synapse type of the node; whether the neuron is inhibitory or excitatory. “EXC” or “INH”.
/0	region	utf8	Optional	Brain region of the cell.
/0	hemisphere	utf8	Optional	“left” or “right” hemisphere.
/0	model_type	utf8	Mandatory	point_neuron
/	node_type_id	int64	Mandatory	Set to -1. Foreign key to node type csv file not used at BBP..

References

Fields for virtual population (model_type: virtual)

This is for virtual nodes (i.e., source nodes of projections).

Group	Field	Type	Requirement	Description
/0	model_type	utf8	Mandatory	virtual
/0	model_template	utf8	Mandatory	See details below.
/	node_type_id	int64	Mandatory	Set to -1. Foreign key to node type csv file not used at BBP..

model_template

The model_template is used to reference a template or class describing the electrophysical properties and mechanisms of the node(s). Its value and interpretation is context-dependent on the corresponding ‘model_type’. When there is no applicable model template for a given model type (i.e. model_type=virtual) it is assigned a value of NULL. Otherwise, within BBP, it uses a colon-separated string-pair with the following syntax: hoc:resource where resource is a reference to the template file-name or class.

Edge File

Fields for Edges

Fields for chemical connection type edges

Connection type is chemical. Group column represents the HDF group where the dataset is located under /<population>. “/” means it is directly under /<population>.

Group	Field	Type	Requirement	Description
/0	afferent_center_[x|y|z]	float32	Mandatory	Position on the axis of the cell’s section/segment in µm.
/0	afferent_surface_[x|y|z]	float32	Mandatory	Position on the surface of a cylindrical cell segment, radially outward from the center position in the direction of the other cell in µm.
/0	afferent_section_id	uint32	Mandatory	The specific section on the target node where a synapse is placed.
/0	afferent_section_pos	float32	Mandatory	Fractional position along the length of the section (normalized to the range [0, 1], where 0 is at the start of the section and 1 is at the end of the section).
/0	afferent_section_type	uint32	Mandatory	Neurite or soma type of the afferent as in morphIO: soma=1, axon=2, basal_dendrite=3, apical_dendrite=4.
/0	afferent_segment_id	uint32	Mandatory	Numerical index of the section of the cell (soma is index 0).
/0	afferent_segment_offset	float32	Mandatory	If triple synapse addressing is being used, the offset within the segment in um. See Frequently Asked Questions.
/0	efferent_center_[x|y|z]	float32	Mandatory	Same as afferent_center_[x|y|z], but for the synapse position at the axon of the presynaptic cell.
/0	efferent_surface_[x|y|z]	float32	Mandatory	Same as efferent_center_[x|y|z], but the for the synapse location on the axon surface.
/0	efferent_section_id	uint32	Mandatory	Same as afferent_section_id, but for source node.
/0	efferent_section_pos	float32	Mandatory	Same as afferent_section_pos, but for source node.
/0	efferent_section_type	uint32	Mandatory	Neurite or soma type of the afferent.
/0	efferent_segment_id	uint32	Mandatory	Numerical index of the section of the cell (soma is index 0).
/0	efferent_segment_offset	float32	Mandatory	If triple synapse addressing is being used, the offset within the segment in µm. See Frequently Asked Questions.
/0	conductance	float32	Mandatory	The conductance of the synapse (nanosiemens); also referred to as g_syn
/0	decay_time	float32	Mandatory	The decay time of the synapse (milliseconds).
/0	depression_time	float32	Mandatory	The depression time constant of the synapse (milliseconds), also referred to as d_syn.
/0	facilitation_time	float32	Mandatory	The facilitation time constant of the synapse (milliseconds), also referred to as f_syn.
/0	u_syn	float32	Mandatory	The u parameter in the Tsodyks Markram Model.
/0	n_rrp_vesicles	uint32	Mandatory	Number of readily releasable pool of vesicles.
/0	spine_length	float32	Mandatory	Distance between the two surface positions in µm.
/0	spine_morphology	uint32	Optional	Integer that maps to a corresponding attribute in /0/@library/spine_morphology
/0	spine_psd_id	int64	Optional	The id of the spine postsynaptic density. Set to -1 if synapse forms on the shaft. A list containing the postsynaptic densities is defined in the spine morphology file, which is set in the spine_morphology field. This id refers to the list index where that psd is stored.
/0	spine_sharing_id	int64	Optional	The sharing id of the spine. This field value is the same for all the synapses that share the same spine. These synapses must be on different postsynaptic densities. Set to -1 if synapse forms on the shaft.
/0	conductance_scale_factor	float32	Optional	The scale factor for the conductance (no unit).If no value or negative, no change is applied.
/0	u_hill_coefficient	float32	Optional	A coefficient describing the scaling of u to be done by the simulator. If no value, no change is applied. \[u_\text{final} = u \cdot y \cdot \frac{ca^4}{u_\text{Hill}^4 + ca^4}\] where \(ca\) denotes the simulated calcium concentration in millimolar and \(y\) a scalar such that at \(ca = 2.0:\ u_\text{final} = u\). (Markram et al., 2015)
/0	syn_type_id	uint32	Mandatory	Takes the value 0 for inhibitory synapses, 100 for excitatory synapses (numerical values due to historic reasons)
/0	syn_property_rule	uint32	Mandatory	The position of the rule that leads to the synapse in the recipe
/0	delay	float32	Mandatory	The axonal delay (in ms, NaN for dendro-dendritic synapses).
/0/@library	spine_morphology	utf8	Optional	Spine morphology file relative path, without the file extension. Example “mymorphology” or “mypath/mymorphology”. Set to an empty string if synapse forms on the shaft.
/	edge_type_id	int64	Mandatory	Links an edge to the underlying CSV file; not used at BBP.
/	source_node_id	uint64	Mandatory	The id of the presynaptic neuron.
/	target_node_id	uint64	Mandatory	The id of the postsynaptic neuron.

source_node_id and target_node_id datasets have an HDF5 attribute of type string named node_population defining the source and target node population name respectively.

Extra fields for plasticity

Plasticity (the activity dependent change in synaptic release probability and conductance) has 8 extra parameters on top of the above ones. These fields are needed by the simulator when there is a “Glusynapse” provided in a modoverride.

Group	Field	Type	Requirement	Description
/0	volume_CR	float32	Mandatory	The volume of the spine (not simulated per se, just used for converting currents to concentrations in GluSynapse.mod).In µm³.
/0	rho0_GB	int64	Mandatory	Initial value of the efficacy, rho parameter of the Graptner and Brunel 2012 model (0: depressed, 1:potentiated).
/0	Use_d_TM	float32	Mandatory	Most depressed value (lower bound) of the release probability, u parameter of the Tsodyks-Markram model.
/0	Use_p_TM	float32	Mandatory	Most potentiated value (upper bound) of the release probability, u parameter of the Tsodyks-Markram model.
/0	gmax_d_AMPA	float32	Mandatory	Most depressed value (lower bound) of the conductance. In nanosiemens.
/0	gmax_p_AMPA	float32	Mandatory	Most potentiated value (upper bound) of the conductance. In nanosiemens.
/0	theta_d	float32	Mandatory	Lower threshold of the efficacy, rho parameter of the Graptner and Brunel 2012 model. Once crossed, it triggers depression.In µ / liter.
/0	theta_p	float32	Mandatory	Higher threshold of the efficacy, rho parameter of the Graptner and Brunel 2012 model. Once crossed, it triggers potentiation.In µ / liter.

GlialGlial connectivity glialglial connection type

Connection type is glialglial. This type of connectivity happens between astrocytes. The properties are similar to Gap Junctions but still chemical connections.

Group	Field	Type	Requirement	Description
/0	afferent_center_[x|y|z]	float32	Mandatory	Position on the axis of the cell’s section/segment in µm on the target cell..
/0	afferent_surface_[x|y|z]	float32	Mandatory	Position on the surface of a cylindrical cell segment, radially outward from the center position in the direction of the other cell in µm on the target cell.
/0	afferent_section_id	uint32	Mandatory	The specific section on the target node where a synapse is placed.
/0	afferent_section_pos	float32	Mandatory	Fractional position along the length of the section (normalized to the range [0, 1], where 0 is at the start of the section and 1 is at the end of the section).
/0	afferent_section_type	uint32	Mandatory	Neurite or soma type of the afferent as in morphIO: soma=1 glia_perivascular_process=2 glia_process=3.
/0	afferent_segment_id	uint32	Mandatory	Numerical index of the section of the cell (soma is index 0).
/0	afferent_segment_offset	float32	Mandatory	If triple synapse addressing is being used, the offset within the segment in um. See Frequently Asked Questions.
/0	efferent_center_[x|y|z]	float32	Mandatory	Same as afferent_center_[x|y|z], but for the source cell.
/0	efferent_surface_[x|y|z]	float32	Mandatory	Same as efferent_center_[x|y|z], but for the connection on the surface of the source cell.
/0	efferent_section_id	uint32	Mandatory	Same as afferent_section_id, but for source node.
/0	efferent_section_pos	float32	Mandatory	Same as afferent_section_pos, but for source node.
/0	efferent_section_type	uint32	Mandatory	Neurite or soma type of the afferent as in morphIO: soma=1 glia_perivascular_process=2 glia_process=3.
/0	efferent_segment_id	uint32	Mandatory	Numerical index of the section of the cell (soma is index 0).
/0	efferent_segment_offset	float32	Mandatory	If triple synapse addressing is being used, the offset within the segment in µm. See Frequently Asked Questions.
/0	spine_length	float32	Mandatory	Distance between the two surface positions in µm.
/	edge_type_id	int64	Mandatory	Links an edge to the underlying CSV file; not used at BBP.
/	source_node_id	uint64	Mandatory	The id of the presynaptic cell.
/	target_node_id	uint64	Mandatory	The id of the postsynaptic cell.

source_node_id and target_node_id datasets have an HDF5 attribute of type string named node_population defining the source and target node population name respectively.

Fields for electrical_synapse connection type edges

Connection type is electrical_synapse. Used for gap junctions between neurons

Group	Field	Type	Requirement	Description
/0	afferent_center_[x|y|z]	float32	Mandatory	Position on the axis of the cell’s section/segment in µm on the target cell..
/0	afferent_surface_[x|y|z]	float32	Mandatory	Position on the surface of a cylindrical cell segment, radially outward from the center position in the direction of the other cell in µm on the target cell.
/0	afferent_section_id	uint32	Mandatory	The specific section on the target node where a synapse is placed.
/0	afferent_section_pos	float32	Mandatory	Fractional position along the length of the section (normalized to the range [0, 1], where 0 is at the start of the section and 1 is at the end of the section).
/0	afferent_section_type	uint32	Mandatory	Neurite or soma type of the afferent as in morphIO: soma=1, axon=2, basal_dendrite=3, apical_dendrite=4.
/0	afferent_segment_id	uint32	Mandatory	Numerical index of the section of the cell (soma is index 0).
/0	afferent_segment_offset	float32	Mandatory	If triple synapse addressing is being used, the offset within the segment in um. See Frequently Asked Questions.
/0	afferent_junction_id	uint32	Mandatory	An internal identifier for the simulator to perform the electrical coupling.
/0	efferent_center_[x|y|z]	float32	Mandatory	Same as afferent_center_[x|y|z], but for the source cell.
/0	efferent_surface_[x|y|z]	float32	Mandatory	Same as efferent_center_[x|y|z], but for the connection on the surface of the source cell.
/0	efferent_section_id	uint32	Mandatory	Same as afferent_section_id, but for source node.
/0	efferent_section_pos	float32	Mandatory	Same as afferent_section_pos, but for source node.
/0	efferent_section_type	uint32	Mandatory	Neurite or soma type of the afferent as in morphIO: soma=1, axon=2, basal_dendrite=3, apical_dendrite=4.
/0	efferent_segment_id	uint32	Mandatory	Numerical index of the section of the cell (soma is index 0).
/0	efferent_segment_offset	float32	Mandatory	If triple synapse addressing is being used, the offset within the segment in µm. See Frequently Asked Questions.
/0	efferent_junction_id	uint32	Mandatory	Same as afferent_junction_id, but for source node.
/0	spine_length	float32	Mandatory	Distance between the two surface positions in µm.
/0	conductance	float32	Mandatory	The conductance of the gap junction (nanosiemens); also referred to as g_syn
/	edge_type_id	int64	Mandatory	Links an edge to the underlying CSV file; not used at BBP.
/	source_node_id	uint64	Mandatory	The id of the presynaptic cell.
/	target_node_id	uint64	Mandatory	The id of the postsynaptic cell.

Fields for synapse_astrocyte connection type edges

Connection type is synapse_astrocyte. Neuroglial connectivity. Astrocytes establish tripartite connections with synapses, wrapping around them forming a three-way ensemble. There are four distrinct elements in a tripartite connection, the astrocyte, the synapse, and the pre and post synaptic neurons. In sonata an edge corresponds to a connection between an astrocyte, a synapse and its post-synaptic neuron. The pre-synaptic neuron, although not directly accessible, it can be retrieved via the synaptic connectivity using the synapse id. The synapse id is stored as a property on the edges.

Group	Field	Type	Requirement	Description
/0	astrocyte_section_id	uint32	Mandatory	The id of the closest astrocyte morphology’s section to the connected neuron-neuron synapse.
/0	astrocyte_segment_id	uint32	Mandatory	The id of the closest astrocyte morphology’s segment to the connected neuron-neuron synapse.
/0	astrocyte_segment_offset	float32	Mandatory	The offset on the respective segment corresponding to the closest point to the respective synapse. In µm.
/0	astrocyte_section_pos	float32	Mandatory	Given the section of where a synapse is closest on the astrocyte the position along the length of that section normalized to the range [0.1] where 0 is at the start of the section and 1 is at the end of the section.
/0	astrocyte_center_[x|y|z]	float32	Mandatory	Position on the axis of the cell’s section/segment in µm on the target cell. This is equivalent to calculating the branch connection point from the morphology’s section_id and section_pos.
/0	synapse_id	uint64	Mandatory	Edge id of the chemical connection to which the astrocyte connects. JDC: issue which edge population… ? source and target are not enough here.
/0	synapse_population	utf8	Mandatory	Edge population of the neuron-neuron synapse to which the astrocyte connects.
/	edge_type_id	int64	Mandatory	Links an edge to the underlying CSV file; not used at BBP.
/	source_node_id	uint64	Mandatory	The node id of the astrocyte.
/	target_node_id	uint64	Mandatory	The node id of the post-synaptic neuron of the tripartite connection.

source_node_id and target_node_id datasets have an HDF5 attribute of type string named node_population defining the source and target node population name respectively.

synapse_id has an HDF5 attribute of type string named edge_population defining the target edge population name for this synapse.

Fields for endfoot connection type edges

Connection type is endfoot. Gliovascular connectivity. Connection between the vasculature and astrocytes. Each edge corresponds to a perivascular endfoot that links an astrocyte with a vasculature segment.

Group	Field	Type	Requirement	Description
/0	endfoot_id	uint64	Mandatory	The identifier of the end foot geometry to lookup in the end_feet_area file.
/0	endfoot_surface_[x|y|z]	float32	Mandatory	The coordinate of the endfoot connection on the vasculature’s surface. In µm.
/0	vasculature_section_id	uint32	Mandatory	The section id in the vasculature morphology.
/0	vasculature_segment_id	uint32	Mandatory	The segment id in the vasculature morphology.
/0	astrocyte_section_id	uint32	Mandatory	The section id of the astrocytic section that connects to the endfoot.
/0	endfoot_compartment_length	float32	Mandatory	The length of the virtual endfoot compartment to construct in the simulator. In µm.
/0	endfoot_compartment_diameter	float32	Mandatory	The diameter of the virtual endfoot compartment to construct in the simulator. In µm.
/0	endfoot_compartment_perimeter	float32	Mandatory	The perimeter of the virtual endfoot compartment to construct in the simulator. In µm.
/	edge_type_id	int64	Mandatory	Links an edge to the underlying CSV file; not used at BBP.
/	source_node_id	uint64	Mandatory	The node id of the vasculature.
/	target_node_id	uint64	Mandatory	The node id of the astrocyte.

source_node_id and target_node_id datasets have an HDF5 attribute of type string named node_population defining the source and target node population name respectively.

This type of population requires extra datasets:

“endfeet_meshes”: path/to/endfeet/meshes

The endfeet_meshes dataset stores the endfeet surface mesh geometry of the perivascular endfeet on the surface of the vasculature. (For a description of the dataset contents see here)

Fields for neuromodulatory connection type edges

Connection type is neuromodulatory. Used for neuromodulatory projections between virtual neurons and existing connectome synapses.

Group	Field	Type	Requirement	Description
/0	afferent_section_id	int16	Mandatory	The specific section on the target node (same of target synapse).
/0	afferent_section_pos	float32	Mandatory	The specific section position (same of target synapse).
/0	afferent_segment_id	int16	Mandatory	The specific segment on the target node (same of target synapse).
/0	delay	float32	Mandatory	The axonal delay. In ms.
/0	neuromod_dtc	float32	Mandatory	Decay time constant (in ms) for the running value of neuromodulator concentration at the synapse.
/0	neuromod_strength	float32	Mandatory	Amount of the increase in neuromodulator concentration at the synapse when a neuromodulatory event is transmitted. In µm.
/	edge_type_id	int64	Mandatory	Links an edge to the underlying CSV file; not used at BBP.
/	source_node_id	uint64	Mandatory	The id of the presynaptic virtual neuron.
/	target_node_id	uint64	Mandatory	The id of the postsynaptic cell (same of target synapse).

Fields for “point neuron connectivity”

Connection type is TM_synapse.

This is a Tsodyks Markram model.

The NEST equivalent implementation can be found here: https://nest-simulator.readthedocs.io/en/v3.0/models/tsodyks2_synapse.html

Group	Field	Type	Requirement	Description
/0	U	float32	Mandatory	Parameter determining the increase of the probability of release with each spike [0,1].
/0	delay	float32	Mandatory	Axonal delay. (ms).
/0	receptor	float32	Mandatory	Receptor type. Indicates which port of the postsynaptic neuron should be used.
/0	tau_fac	float32	Mandatory	Facilitation time constant. (ms).
/0	tau_rec	float32	Mandatory	Time constant for depression. (ms).
/0	weight	float32	Mandatory	Synaptic weight.
/	edge_type_id	int64	Mandatory	Links an edge to the underlying CSV file; not used at BBP.
/	source_node_id	uint64	Mandatory	The id of the presynaptic neuron.
/	target_node_id	uint64	Mandatory	The id of the postsynaptic neuron.

source_node_id and target_node_id datasets have an HDF5 attribute of type string named node_population defining the source and target node population name respectively.

Fields for projections (virtual chemical connection type edges)

Connection type is chemical. The difference to the normal chemical type is that some fields are missing as the source nodes are virtual.

Group	Field	Type	Requirement	Description
/0	afferent_center_[x|y|z]	float32	Mandatory	Position on the axis of the cell’s section/segment in µm.
/0	afferent_section_id	uint32	Mandatory	The specific section on the target node where a synapse is placed.
/0	afferent_section_pos	float32	Mandatory	Fractional position along the length of the section (normalized to the range [0, 1], where 0 is at the start of the section and 1 is at the end of the section).
/0	afferent_section_type	uint32	Mandatory	Neurite or soma type of the afferent as in morphIO: soma=1, axon=2, basal_dendrite=3, apical_dendrite=4.
/0	afferent_segment_id	uint32	Mandatory	Numerical index of the section of the cell (soma is index 0).
/0	afferent_segment_offset	float32	Mandatory	If triple synapse addressing is being used, the offset within the segment in um. See Frequently Asked Questions.
/0	efferent_section_type	uint32	Mandatory	Neurite or soma type of the afferent.
/0	conductance	float32	Mandatory	The conductance of the synapse (nanosiemens); also referred to as g_syn
/0	decay_time	float32	Mandatory	The decay time of the synapse (milliseconds).
/0	depression_time	float32	Mandatory	The depression time constant of the synapse (milliseconds), also referred to as f_syn.
/0	facilitation_time	float32	Mandatory	The facilitation time constant (milliseconds) of the synapse.
/0	u_syn	float32	Mandatory	The u parameter in the Tsodyks Markram Model.
/0	n_rrp_vesicles	uint32	Mandatory	Number of readily releasable pool of vesicles.
/0	conductance_scale_factor	float32	Optional	The scale factor for the conductance (no unit).If no value or negative, no change is applied.
/0	u_hill_coefficient	float32	Optional	A coefficient describing the scaling of u to be done by the simulator. If no value, no change is applied. \[u_\text{final} = u \cdot y \cdot \frac{ca^4}{u_\text{Hill}^4 + ca^4}\] where \(ca\) denotes the simulated calcium concentration in millimolar and \(y\) a scalar such that at \(ca = 2.0:\ u_\text{final} = u\). (Markram et al., 2015)
/0	syn_type_id	uint32	Mandatory	The position of the rule that leads to the synapse in the recipe + 100 if it is an excitatory synapse
/0	delay	float32	Mandatory	The axonal delay (in ms, NaN for dendro-dendritic synapses).
/	edge_type_id	int64	Mandatory	Links an edge to the underlying CSV file; not used at BBP.
/	source_node_id	uint64	Mandatory	The id of the presynaptic neuron.
/	target_node_id	uint64	Mandatory	The id of the postsynaptic neuron.

source_node_id and target_node_id datasets have an HDF5 attribute of type string named node_population defining the source and target node population name respectively.

Consumers

Consumers use the sonata .h5 files, and depending on the tool the required fields are different.

TouchDetector

Required fields for TouchDetector:

• x, y, z

• orientation_w, orientation_x, orientation_y, orientation_z

• morphology

• region

• mtype

Spykfunc

Required fields for Spykfunc:

• morphology

• etype

• mtype

• synapse_class

Format of the electrodes_file

The format of the weights file is described below:

Group	Field	Type	Shape	Requirement	Description
/electrodes/{electrodename}	layer	utf8	1	Optional	Layer of the circuit in which {electrodename} is located. If the electrode is in a region without cortical layers, then “NA”. If the electrode is outside the brain, then “Outside”
/electrodes/{electrodename}	position	float32	3	Mandatory	Position of {electrodename} in microns, in cartesian coordinates
/electrodes/{electrodename}	region	utf8	1	Optional	Region in which {electrodename} is located
/electrodes/{electrodename}	type	utf8	1	Optional	Either EEG or LFP
/electrodes/{electrodename}/{population_name}	index	uint64	1	Mandatory	Index of the column corresponding to this electrode in /electrodes/{population_name}/scaling_factors
/electrodes/{population_name}	scaling_factors	float64	Total_comp x N_elec	Mandatory	Scaling factor for each compartment in the corresponding neuron, in mV/nA
/{population_name}	node_ids	uint64	N_nodes	Mandatory	List of node ids. Node ids not listed here are to be ignored
/{population_name}	offsets	uint64	N_nodes + 1	Mandatory	The offset for each node in the scaling_factors field