model of DCN pyramidal neuron
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 

619 lines
23 KiB

from __future__ import print_function
from neuron import h
import numpy as np
# import neuron as nrn
from .cell import Cell
from .. import synapses
from ..util import nstomho
from ..util import Params
from .. import data
__all__ = ["Tuberculoventral"]
class Tuberculoventral(Cell):
type = "tuberculoventral"
@classmethod
def create(cls, model="TVmouse", **kwds):
if model in ["TVmouse", "I"]:
return Tuberculoventral(**kwds)
elif model == "dummy":
return DummyTuberculoventral(**kwds)
else:
raise ValueError("Tuberculoventral type %s is unknown", model)
def __init__(self):
Cell.__init__(self)
self.spike_source = (
None
) # used by DummyTuberculoventral to connect VecStim to terminal
def make_psd(self, terminal, psd_type, **kwds):
"""
Connect a presynaptic terminal to one post section at the specified location, with the fraction
of the "standard" conductance determined by gbar.
The default condition is to try to pass the default unit test (loc=0.5)
Parameters
----------
terminal : Presynaptic terminal (NEURON object)
psd_type : either simple or multisite PSD for bushy cell
kwds: dict of options. Two are currently handled:
postsize : expect a list consisting of [sectionno, location (float)]
AMPAScale : float to scale the ampa currents
"""
if (
"postsite" in kwds
): # use a defined location instead of the default (soma(0.5)
postsite = kwds["postsite"]
loc = postsite[1] # where on the section?
uname = (
"sections[%d]" % postsite[0]
) # make a name to look up the neuron section object
post_sec = self.hr.get_section(uname) # Tell us where to put the synapse.
else:
loc = 0.5
post_sec = self.soma
if psd_type == "simple":
if terminal.cell.type in ["sgc", "dstellate", "tuberculoventral"]:
weight = data.get(
"%s_synapse" % terminal.cell.type,
species=self.species,
post_type=self.type,
field="weight",
)
tau1 = data.get(
"%s_synapse" % terminal.cell.type,
species=self.species,
post_type=self.type,
field="tau1",
)
tau2 = data.get(
"%s_synapse" % terminal.cell.type,
species=self.species,
post_type=self.type,
field="tau2",
)
erev = data.get(
"%s_synapse" % terminal.cell.type,
species=self.species,
post_type=self.type,
field="erev",
)
return self.make_exp2_psd(
post_sec,
terminal,
weight=weight,
loc=loc,
tau1=tau1,
tau2=tau2,
erev=erev,
)
else:
raise TypeError(
"Cannot make simple PSD for %s => %s"
% (terminal.cell.type, self.type)
)
elif psd_type == "multisite":
if terminal.cell.type == "sgc":
# Max conductances for the glu mechanisms are calibrated by
# running `synapses/tests/test_psd.py`. The test should fail
# if these values are incorrect
self.AMPAR_gmax = (
data.get(
"sgc_synapse",
species=self.species,
post_type=self.type,
field="AMPAR_gmax",
)
* 1e3
)
self.NMDAR_gmax = (
data.get(
"sgc_synapse",
species=self.species,
post_type=self.type,
field="NMDAR_gmax",
)
* 1e3
)
self.Pr = data.get(
"sgc_synapse", species=self.species, post_type=self.type, field="Pr"
)
# adjust gmax to correct for initial Pr
self.AMPAR_gmax = self.AMPAR_gmax / self.Pr
self.NMDAR_gmax = self.NMDAR_gmax / self.Pr
if "AMPAScale" in kwds:
self.AMPA_gmax = (
self.AMPA_gmax * kwds["AMPAScale"]
) # allow scaling of AMPA conductances
if "NMDAScale" in kwds:
self.NMDA_gmax = self.NMDA_gmax * kwds["NMDAScale"]
return self.make_glu_psd(
post_sec, terminal, self.AMPAR_gmax, self.NMDAR_gmax, loc=loc
)
elif terminal.cell.type == "dstellate": # WBI input -Voigt, Nelken, Young
return self.make_gly_psd(post_sec, terminal, psdtype="glyfast", loc=loc)
elif (
terminal.cell.type == "tuberculoventral"
): # TV cells talk to each other-Kuo et al.
return self.make_gly_psd(post_sec, terminal, psdtype="glyfast", loc=loc)
else:
raise TypeError(
"Cannot make PSD for %s => %s" % (terminal.cell.type, self.type)
)
else:
raise ValueError("Unsupported psd type %s" % psd_type)
def make_terminal(self, post_cell, term_type, **kwds):
pre_sec = self.soma
if term_type == "simple":
return synapses.SimpleTerminal(
pre_sec, post_cell, spike_source=self.spike_source, **kwds
)
elif term_type == "multisite":
if post_cell.type in [
"dstellate",
"tuberculoventral",
"pyramidal",
"bushy",
"tstellate",
]:
nzones = data.get(
"tuberculoventral_synapse",
species=self.species,
post_type=post_cell.type,
field="n_rsites",
)
delay = data.get(
"tuberculoventral_synapse",
species=self.species,
post_type=post_cell.type,
field="delay",
)
else:
raise NotImplementedError(
"No knowledge as to how to connect tuberculoventral cell to cell type %s"
% type(post_cell)
)
pre_sec = self.soma
return synapses.StochasticTerminal(
pre_sec,
post_cell,
nzones=nzones,
spike_source=self.spike_source,
delay=delay,
**kwds
)
else:
raise ValueError("Unsupported terminal type %s" % term_type)
class Tuberculoventral(Tuberculoventral):
"""
Tuberculoventral Neuron (DCN) base model
Adapted from T-stellate model, using target parameters from Kuo et al. J. Neurophys. 2012
"""
def __init__(
self,
morphology=None,
decorator=None,
nach=None,
ttx=False,
species="mouse",
modelType=None,
debug=False,
):
"""
Initialize a DCN Tuberculoventral cell, using the default parameters for guinea pig from
R&M2003, as a type I cell.
Modifications to the cell can be made by calling methods below. These include:
Converting to a type IA model (add transient K current) (species: guineapig-TypeIA).
Changing "species" to mouse or cat (scales conductances)
Parameters
----------
morphology : string (default: None)
a file name to read the cell morphology from. If a valid file is found, a cell is constructed
as a cable model from the hoc file.
If None (default), the only a point model is made, exactly according to RM03.
decorator : Python function (default: None)
decorator is a function that "decorates" the morphology with ion channels according
to a set of rules.
If None, a default set of channels aer inserted into the first soma section, and the
rest of the structure is "bare".
nach : string (default: 'na')
nach selects the type of sodium channel that will be used in the model. A channel mechanims
by that name must exist.
ttx : Boolean (default: False)
If ttx is True, then the sodium channel conductance is set to 0 everywhere in the cell.
Currently, this is not implemented.
species: string (default 'guineapig')
species defines the channel density that will be inserted for different models. Note that
if a decorator function is specified, this argument is ignored.
modelType: string (default: None)
modelType specifies the type of the model that will be used (e.g., "II", "II-I", etc).
modelType is passed to the decorator, or to species_scaling to adjust point models.
debug: boolean (default: False)
debug is a boolean flag. When set, there will be multiple printouts of progress and parameters.
Returns
-------
Nothing
"""
super(Tuberculoventral, self).__init__()
if modelType == None:
modelType = "TVmouse"
if nach == None:
nach = "nacncoop"
self.debug = debug
self.status = {
"soma": True,
"axon": False,
"dendrites": False,
"pumps": False,
"na": nach,
"species": species,
"modelType": modelType,
"ttx": ttx,
"name": "Tuberculoventral",
"morphology": morphology,
"decorator": decorator,
"temperature": None,
}
self.i_test_range = {"pulse": [(-0.35, 1.0, 0.05), (-0.04, 0.01, 0.01)]}
self.vrange = [-80.0, -60.0] # set a default vrange for searching for rmp
if morphology is None:
"""
instantiate a basic soma-only ("point") model
"""
if self.debug:
print("<< Tuberculoventral model: Creating point cell >>")
soma = h.Section(
name="Tuberculoventral_Soma_%x" % id(self)
) # one compartment of about 29000 um2
soma.nseg = 1
self.add_section(soma, "soma")
else:
"""
instantiate a structured model with the morphology as specified by
the morphology file
"""
if self.debug:
print("<< Tuberculoventral model: Creating structured cell >>")
self.set_morphology(morphology_file=morphology)
# decorate the morphology with ion channels
if decorator is None: # basic model, only on the soma
self.mechanisms = ["kht", "ka", "ihvcn", "leak", nach]
for mech in self.mechanisms:
self.soma.insert(mech)
self.species_scaling(
silent=True, species=species, modelType=modelType
) # adjust the default parameters
else: # decorate according to a defined set of rules on all cell compartments
self.decorate()
self.save_all_mechs() # save all mechanisms inserted, location and gbar values...
self.get_mechs(self.soma)
if self.debug:
print("<< Tuberculoventral cell model created >>")
def get_cellpars(self, dataset, species="mouse", celltype="TVmouse"):
cellcap = data.get(
dataset, species=species, cell_type=celltype, field="soma_Cap"
)
chtype = data.get(
dataset, species=species, cell_type=celltype, field="soma_na_type"
)
pars = Params(soma_cap=cellcap, soma_na_type=chtype)
for g in [
"soma_nacncoop_gbar",
"soma_kht_gbar",
"soma_ka_gbar",
"soma_ihvcn_gbar",
"soma_ihvcn_eh",
"soma_leak_gbar",
"soma_leak_erev",
"soma_e_k",
"soma_e_na",
]:
pars.additem(
g, data.get(dataset, species=species, cell_type=celltype, field=g)
)
return pars
def species_scaling(self, species="guineapig", modelType="TVmouse", silent=True):
"""
Adjust all of the conductances and the cell size according to the species requested.
Used ONLY for point models.
Parameters
----------
species : string (default: 'guineapig')
name of the species to use for scaling the conductances in the base point model
Must be one of mouse, cat, guineapig
modelType: string (default: 'I-c')
definition of model type from RM03 models, type I-c or type I-t
silent : boolean (default: True)
run silently (True) or verbosely (False)
"""
soma = self.soma
if self.debug:
print("modelType: ", modelType)
if modelType in ["TVmouse", "I"]:
celltype = "TVmouse" # modelType
modelType = "TVmouse"
else:
raise ValueError(
"Tuberuloventral: Model type %s not recognized" % modelType
)
if species == "mouse" and modelType in ["TVmouse", "I"]:
"""#From Kuo 150 Mohm, 10 msec tau
Firing at 600 pA about 400 Hz
These values from brute_force runs, getting 380 Hz at 600 pA at 35C
Input resistance and vm is ok, time constnat is short
*** Rin: 168 tau: 7.8 v: -68.4
Attempts to get longer time constant - cannot keep rate up.
"""
# Adapted from TStellate model type I-c'
self.vrange = [-80.0, -58.0]
self._valid_temperatures = (34.0,)
if self.status["temperature"] is None:
self.set_temperature(34.0)
pars = self.get_cellpars("TV_channels", species="mouse", celltype=modelType)
self.set_soma_size_from_Cm(pars.soma_cap)
self.status["na"] = pars.soma_na_type
self.adjust_na_chans(soma, gbar=pars.soma_nacncoop_gbar, debug=self.debug)
soma().kht.gbar = nstomho(pars.soma_kht_gbar, self.somaarea)
soma().ka.gbar = nstomho(pars.soma_ka_gbar, self.somaarea)
soma().ihvcn.gbar = nstomho(pars.soma_ihvcn_gbar, self.somaarea)
soma().ihvcn.eh = pars.soma_ihvcn_eh
soma().leak.gbar = nstomho(pars.soma_leak_gbar, self.somaarea)
soma().leak.erev = pars.soma_leak_erev
self.e_leak = pars.soma_leak_erev
self.soma.ek = self.e_k = pars.soma_e_k
self.soma.ena = self.e_na = pars.soma_e_na
self.axonsf = 0.5
else:
raise ValueError(
"Species %s or species-type %s is not recognized for Tuberculoventralcells"
% (species, type)
)
self.status["species"] = species
self.status["modelType"] = modelType
self.check_temperature()
def channel_manager(self, modelType="TVmouse"):
"""
This routine defines channel density maps and distance map patterns
for each type of compartment in the cell. The maps
are used by the ChannelDecorator class (specifically, it's private
_biophys function) to decorate the cell membrane.
Parameters
----------
modelType : string (default: 'RM03')
A string that defines the type of the model. Currently, 3 types are implemented:
RM03: Rothman and Manis, 2003 somatic densities for guinea pig
XM13: Xie and Manis, 2013, somatic densities for mouse
XM13PasDend: XM13, but with only passive dendrites, no channels.
Returns
-------
Nothing
Notes
-----
This routine defines the following variables for the class:
- conductances (gBar)
- a channelMap (dictonary of channel densities in defined anatomical compartments)
- a current injection range for IV's (when testing)
- a distance map, which defines how selected conductances in selected compartments
will change with distance. This includes both linear and exponential gradients,
the minimum conductance at the end of the gradient, and the space constant or
slope for the gradient.
"""
if modelType == "TVmouse":
print("decorate as tvmouse")
# totcap = 95.0E-12 # Tuberculoventral cell (type I), based on stellate, adjusted for Kuo et al. TV firing
self.set_soma_size_from_Section(self.soma)
totcap = self.totcap
refarea = self.somaarea # totcap / self.c_m # see above for units
self.gBar = Params(
nabar=1520.0e-9 / refarea,
khtbar=160.0e-9 / refarea,
kltbar=0.0e-9 / refarea,
kabar=65.0 / refarea,
ihbar=1.25e-9 / refarea,
leakbar=5.5e-9 / refarea,
)
self.channelMap = {
"axon": {
"nacn": 0.0,
"klt": 0.0,
"kht": self.gBar.khtbar,
"ihvcn": 0.0,
"leak": self.gBar.leakbar / 4.0,
},
"hillock": {
"nacn": self.gBar.nabar,
"klt": 0.0,
"kht": self.gBar.khtbar,
"ihvcn": 0.0,
"leak": self.gBar.leakbar,
},
"initseg": {
"nacn": self.gBar.nabar,
"klt": 0.0,
"kht": self.gBar.khtbar,
"ihvcn": self.gBar.ihbar / 2.0,
"leak": self.gBar.leakbar,
},
"soma": {
"nacn": self.gBar.nabar,
"klt": self.gBar.kltbar,
"kht": self.gBar.khtbar,
"ihvcn": self.gBar.ihbar,
"leak": self.gBar.leakbar,
},
"dend": {
"nacn": self.gBar.nabar / 2.0,
"klt": 0.0,
"kht": self.gBar.khtbar * 0.5,
"ihvcn": self.gBar.ihbar / 3.0,
"leak": self.gBar.leakbar * 0.5,
},
"apic": {
"nacn": 0.0,
"klt": 0.0,
"kht": self.gBar.khtbar * 0.2,
"ihvcn": self.gBar.ihbar / 4.0,
"leak": self.gBar.leakbar * 0.2,
},
}
self.irange = np.linspace(-0.3, 0.6, 10)
self.distMap = {
"dend": {
"klt": {"gradient": "linear", "gminf": 0.0, "lambda": 100.0},
"kht": {"gradient": "linear", "gminf": 0.0, "lambda": 100.0},
}, # linear with distance, gminf (factor) is multiplied by gbar
"apic": {
"klt": {"gradient": "linear", "gminf": 0.0, "lambda": 100.0},
"kht": {"gradient": "linear", "gminf": 0.0, "lambda": 100.0},
}, # gradients are: flat, linear, exponential
}
else:
raise ValueError("model type %s is not implemented" % modelType)
def adjust_na_chans(self, soma, gbar=1000.0, debug=False):
"""
Adjust the sodium channel conductance, depending on the type of conductance
Parameters
----------
soma : NEURON section object (required)
This identifies the soma object whose sodium channel complement will have it's
conductances adjusted depending on the sodium channel type
gbar : float (default: 1000.)
The "maximal" conductance to be set in the model.
debug : boolean (default: False)
A flag the prints out messages to confirm the operations applied.
Returns
-------
Nothing
"""
if self.status["ttx"]:
gnabar = 0.0
else:
gnabar = nstomho(gbar, self.somaarea)
nach = self.status["na"]
if nach == "nacncoop":
soma().nacncoop.gbar = gnabar
soma().nacncoop.KJ = 2000.0
soma().nacncoop.p = 0.25
soma.ena = self.e_na
if debug:
print("nacncoop gbar: ", soma().nacncoop.gbar)
elif nach == "jsrna":
soma().jsrna.gbar = gnabar
soma.ena = self.e_na
if debug:
print("jsrna gbar: ", soma().jsrna.gbar)
elif nach == "nav11":
soma().nav11.gbar = gnabar * 0.5
soma.ena = self.e_na
soma().nav11.vsna = 4.3
if debug:
print("Tuberculoventral using inva11")
print("nav11 gbar: ", soma().nav11.gbar)
elif nach == "na":
soma().na.gbar = gnabar
soma.ena = self.e_na
if debug:
print("na gbar: ", soma().na.gbar)
elif nach == "nacn":
soma().nacn.gbar = gnabar
soma.ena = self.e_na
if debug:
print("nacn gbar: ", soma().nacn.gbar)
else:
raise ValueError(
"Tuberculoventral setting Na channels: channel %s not known" % nach
)
class DummyTuberculoventral(Tuberculoventral):
""" Tuberculoventral cell class with no cell body; this cell only replays a predetermined
spike train. Useful for testing, or replacing spike trains to determine
the importance of spike structures within a network.
"""
def __init__(self, cf=None, species="mouse"):
"""
Parameters
----------
cf : float (default: None)
Required: the characteristic frequency for the TV cell
Really just for reference.
"""
Tuberculoventral.__init__(self)
self.vecstim = h.VecStim()
# this causes the terminal to receive events from the VecStim:
self.spike_source = self.vecstim
# just an empty section for holding the terminal
self.add_section(h.Section(), "soma")
self.status = {
"soma": True,
"axon": False,
"dendrites": False,
"pumps": False,
"na": None,
"species": species,
"modelType": "Dummy",
"modelName": "DummyTuberculoventral",
"ttx": None,
"name": "DummyTuberculoventral",
"morphology": None,
"decorator": None,
"temperature": None,
}
print("<< Tuberculoventral: Dummy Tuberculoventral Cell created >>")
def set_spiketrain(self, times):
""" Set the times of spikes (in seconds) to be replayed by the cell.
"""
self._spiketrain = times
self._stvec = h.Vector(times)
self.vecstim.play(self._stvec)