Python and NEURON scripts for running single-cell simulations of layer V pyramidal cells (Hay model).
Tuomo Maki-Marttunen, 2015-2018
CC BY 4.0
HOC-commands for simulations based on (Hay et al. 2011, "Models of neocortical layer 5b
pyramidal cells capturing a wide range of dendritic and
perisomatic active properties", PLoS Comp Biol. 2011
Jul;7(7):e1002107)
HOC-commands for in-vivo-like synaptic inputs based on (Hay & Segev 2015, "Dendritic excitability and gain control
in recurrent cortical microcircuits", Cerebral Cortex
25(10): 3561-3571)
Library of variants based on (Maki-Marttunen et al. 2016, "Functional Effects of Schizophrenia-Linked Genetic Variants on Intrinsic
Single-Neuron Excitability: A Modeling Study", Biol Psychiatry Cogn Neurosci Neuroimaging. 2016 Jan
1;1(1):49-59.
Mod Files included:
CaDynamics_E2.mod #mod file for Intracellular Ca dynamics
Ca_HVA.mod #mod file for HVA Ca currents
Ca_LVAst.mod #mod file for LVA Ca currents
Ih.mod #mod file for HCN current
Im.mod #mod file for M-current
K_Pst.mod #mod file for persistent K current
K_Tst.mod #mod file for transient K current
NaTa_t.mod #mod file for transient Na current
Nap_Et2.mod #mod file for persistent Na current
ProbAMPANMDA2.mod #mod file for probabilistic glutamatergic synapse
ProbUDFsyn2.mod #mod file for probabilistic GABAergic synapse
SK_E2.mod #mod file for SK current
SKv3_1.mod #mod file for Kv3.1 potassium current
epsp.mod #mod file for EPSP-like current
Python files included:
calcapicalthresholds_control.py #Determine the threshold amplitude for a 200-ms stimuli at the apical dendrite (different distances from soma) for eliciting a spike
calcapicalthresholds_epsp_control.py #Determine the threshold amplitude for a alpha-shaped synaptic currents (0.5 ms rise time, 5 ms decay time) at the apical dendrite (different distances from soma) for eliciting a spike
calcifcurves.py #Calculate the fI curves for the variants
calcifcurves_comb.py #Calculate the fI curves for the variant combinations
calcnspikesperburst2.py #Calculate the numbers of bursts given different parameter changes and different input current amplitudes. Needed for Fig. 1.
calcsteadystate.py #Calculate the steady-state behaviour of different variants. Needed for Fig. 2 time courses
calcupdown2responses.py #Calculate the responses to combinations of apical and perisomatic stimuli during simplistic up and down states. Needed for Fig. 4
calcupdownresponses_noisydown.py #Calculate the responses to combinations of apical and perisomatic stimuli during in-vivo-like background synaptic firing (down state). Needed for Fig. 4
calcupdownresponses_noisyup.py #Calculate the responses to combinations of apical and perisomatic stimuli during in-vivo-like background synaptic firing (up state). Needed for Fig. 4
coding.py #Determine the neuron response (variants) to different combinations of inputs. Needed for Fig. 7A
coding_comb.py #Determine the neuron response (variant combinations) to different combinations of inputs.
coding_nonprop_comb_somaticI.py #Determine the neuron response (variant combinations) to different combinations of inputs, amplitudes of inputs not proportional to threshold conductances. Needed for Fig. S4
coding_nonprop_somaticI.py #Determine the neuron response (variants) to different combinations of inputs, amplitudes of inputs not proportional to threshold conductances. Needed for Fig. 7
collectupdownresponses_noisy.py #Collect responses to noisy up- and down state stimuli
drawfigcomb.py #Draw Fig. S4 (panels A, B, and C)
drawnspikesperburst2.py #Draw Fig. 1 (left panel) based on the output of calcnspikesperburst2.py
drawupdownresponses_noisy.py #Draw Fig. 5 data based on collectupdownresponses_noisy.py
findppicoeffs.py #Determine the PPI coefficient for different ISIs (variants)
findppicoeffs_comb.py #Determine the PPI coefficient for different ISIs (variant combinations)
findppicoeffs_complement.py #Determine the PPI coefficient for different ISIs, dense resolution (variants)
findthresholdbasalamps_coding.py #Determine the threshold conductance for spike initiation in the basal dendritic region in the coding experiment (control)
findthresholddistalamps.py #Determine the threshold conductance for spike initiation in the apical region from a given distance on (variants)
findthresholddistalamps_coding.py #Determine the threshold conductance for spike initiation in the 6 apical dendritic regions in the coding experiment (control)
findthresholddistalamps_comb.py #Determine the threshold conductance for spike initiation in the apical region from a given distance on (variant combinations)
mutation_stuff.py #Python library for the effects of SNP-like SCZ-associated variants
mytools.py #Python library, general tools
protocol.py #Python library needed by setparams.py for setting the default values for certain parameters
runcontrols_cs.py #Calculate the control neuron's firing behaviour
savebasalsynapselocations_coding.py #Determine locations for 1000 synapses in the basal dendritic region
savesynapselocations.py #Determine locations for 3000 synapses in the apical dendritic region, from a given distance on
savesynapselocations_coding.py #Determine locations for 1000 synapses in one of the six apical dendritic regions
scalemutations_cs.py #Determine the scaling parameters for the variants
setparams.py #Set parameters to get the altered Hay models from the unaltered (original) Hay model
Pre-saved data
control_cs.sav #Control neuron data (needed by scalemutations_cs.py)
controlamps_cs0.sav #Control neuron data (needed by scalemutations_cs.py)
controlamps_cs1.sav #Control neuron data (needed by scalemutations_cs.py)
controlamps_cs2.sav #Control neuron data (needed by scalemutations_cs.py)
controlamps_cs3.sav #Control neuron data (needed by scalemutations_cs.py)
controlamps_cs4.sav #Control neuron data (needed by scalemutations_cs.py)
controlamps_cs5.sav #Control neuron data (needed by scalemutations_cs.py)
controlamps_cs6.sav #Control neuron data (needed by scalemutations_cs.py)
scalings_cs.sav #Scaling coefficients (needed by all scripts running variant simulations)
synlocs300.sav #Locations of the synapses (needed by findthresholddistalamps*.py and findppicoeffs*.py)
To run NEURON simulation and plot the results (Figures 1 and A4), run the following commands:
################ Initialization ################
nrnivmodl #Compile the mechanisms
########## Simulations for Fig 1 (left panel) ##########
python calcnspikesperburst2.py #Calculates many DC stimulation experiments (44 different parameter sets, 11 current amplitudes for each).
#Saves files spikesperburst2_*.sav that contain the data on how many spikes there were per butsts. Takes 2-3 hours on standard PC.
#The script also saves figures nspikesperbursts_cs0_${iParam}_${iI}.eps that illustrate the membrane potential time course of each experiment.
python drawnspikesperburst2.py #Analyzes the data produced by calcnspikesperburst2.py and saves nSpikesPerBurstsHay.eps (the left panel of Fig. 1)
########## Simulations for Fig 5A ##########
python calcapicalthresholds_control.py #This script calculates the thresholds for apical DC input to induce a somatic spike. Saves data needed by calcupdownresponses_noisyup.py and calcupdownresponses_noisydown.py
python calcapicalthresholds_epsp_control.py #This script calculates the thresholds for apical EPSP-like input to induce a somatic spike. Saves data needed by calcupdownresponses_noisyup.py and calcupdownresponses_noisydown.py
########## Simulations for Fig A4 (panels A, B, and C) ##########
### f-I curve and PPI experiments:
python calcifcurves_comb.py #Calculates the f-I curves for the combination of variants (epsilon = 1/4, 1/2, -1/4 and -1/2 scalings) - the combination consists of the
#variants (max. one per gene) that had a maximal excitatory effect on the f-I curves in Fig. 2. This is a heavy operation, takes around 15 hours.
python findthresholddistalamps_comb.py #Calculates the threshold conductance for 3000 synapses distributed along the apical dendrite, starting from 300 um (i.e., synlocs300.sav is needed).
#Saves files thresholddistalamp300_cs*_comb0.sav. This is relatively light script, takes approximately 5 min per parameter set (only first one
#needed really) on a standard PC.
python findppicoeffs_comb.py 0 1 0 #Calculates the PPI curves using the data saved by findthresholddistalamps_comb.py. Here, three arguments are given to constrain the process:
#Only combination "0" (the same as in above scripts, i.e. the one that gave an excitatory effect in f-I experiments) is used, only one
#distribution of synapses (starting from 300 um), and only for one model (no parameter changes). This is a heavy operation, takes approximately
#4 hours.
# We also need the above data for control case (blue curves). This can be obtained by running calcifcurves.py, findthresholddistalamps.py and findppicoeffs.py. Note that
# these scripts can be used to run any of the >100 variants, but that the control simulations are included only for the first variant. Thus we run them with argument "0" indicating
# we only want the first variant. If you want further speed up, change the code that only iter=-1 will be run (but then include four or five empty lists before these data in the pickle array)
python calcifcurves.py 0 #Calculates the f-I curves for the variant "0" (epsilon = 1/4, 1/2, -1/4 and -1/2 scalings) and the control neuron. This is quite heavy, takes around 15 hours.
python findthresholddistalamps.py 0 #Calculates the threshold conductance for 3000 synapses distributed along the apical dendrite, starting from 300 um, for variant "0" and the control neuron
python findppicoeffs.py 0 1 0 #Calculates the PPI curves using the data saved by findthresholddistalamps.py. A heavy operation, 4-5 hours on standard PC.
### Coding experiment:
for i in 0 1 2 3 4 5 # In each of these six simulations,
do # we distribute the synapses
python savesynapselocations_coding.py $i # onto one of the six distal regions
done # (so that they are relatively uniformly distributed)
python savebasalsynapselocations_coding.py # Here the same for basal dendrites
#Next, we calculate the response of the neuron with the combination of variants to the 128 inputs with different synaptic conductances and a small somatic input
#This is a lot of simulations, so if you can parallelize them that's commendable. One simulation takes 0.5 - 1 hour on a standard computer.
for syncond in 0.0000025 0.000005 0.0000075 0.00001 0.0000125 0.000015 0.0000175 0.00002 0.0000225 0.000025 0.0000275 0.00003 0.0000325 0.000035 0.0000375 0.00004 0.0000425 0.000045 0.0000475 0.00005 0.0000525 0.000055
do
for somaticI in -0.1 -0.09 -0.08 -0.07 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0.0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
do
python coding_nonprop_comb_somaticI.py 0 $syncond 0 $somaticI
done
done