I'm having a problem recording a range variable into a Vector. Specifically, I'm using the Mainen et al. 1995 spike initiation model (http://senselab.med.yale.edu/modeldb/Sh ... model=8210) and want to look at the different state variables, for example of the sodium channels.

Just to demonstrate what I mean I put together this little example script:

Code: Select all

```
import neuron
import numpy as np
import matplotlib.pyplot as plt
h = neuron.h
Rm = 15000.0
cm = 1.0
Ra = 150.0
ePas = -65.0
soma = h.Section()
soma.L = 15.0
soma.diam = 15.0
soma.nseg = 1
soma.cm = cm
soma.Ra = Ra
soma.insert('pas')
for seg in soma:
seg.pas.g = 1/Rm
seg.pas.e = ePas
soma.insert('na3')
soma.insert('kd3')
rec = {}
for lbl in 't','vSoma':
rec[lbl] = h.Vector()
rec['t'].record(h._ref_t)
rec['vSoma'].record(soma(0.5)._ref_v)
stateVars = {}
for var in 'm', 'h', 'minf', 'hinf', 'mtau', 'htau':
stateVars[var] = h.Vector()
stateVars['m'].record(soma(0.5).na3._ref_m, sec=soma)
stateVars['h'].record(soma(0.5).na3._ref_h, sec=soma)
stateVars['mtau'].record(soma(0.5).na3._ref_mtau, sec=soma)
stateVars['htau'].record(soma(0.5).na3._ref_htau, sec=soma)
stateVars['minf'].record(soma(0.5).na3._ref_minf, sec=soma)
stateVars['hinf'].record(soma(0.5).na3._ref_hinf, sec=soma)
somaIC = h.IClamp(0.5, sec=soma)
somaIC.delay = 10.0
somaIC.dur = 90.0
somaIC.amp = 0.01
h.load_file('stdrun.hoc')
h.dt = 0.025
h.celsius = 37.0
h.finitialize(-65.0)
h.tstop = 100.0
h.run()
tArray = np.array(rec['t'])
vArray = np.array(rec['vSoma'])
plt.figure(1)
plt.subplot(221)
plt.plot(tArray, vArray, 'b')
plt.ylabel('Voltage [mV]')
plt.xlabel('Time [ms]')
plt.subplot(222)
plt.plot(tArray, stateVars['m'], 'b', label='m')
plt.plot(tArray, stateVars['h'], 'r', label='h')
plt.ylabel('Na state variables m/h')
plt.xlabel('Time [ms]')
plt.legend()
plt.subplot(223)
plt.plot(tArray, stateVars['mtau'], 'b', label='$\\tau_{m}$')
plt.plot(tArray, stateVars['htau'], 'r', label='$\\tau_{h}$')
plt.ylabel('Na state variables $\\tau_m$/$\\tau_h$')
plt.xlabel('Time [ms]')
plt.legend()
plt.subplot(224)
plt.plot(tArray, stateVars['minf'], 'b', label='$m_{\infty}$')
plt.plot(tArray, stateVars['hinf'], 'r', label='$h_{\infty}$')
plt.ylabel('Na state variables $m_{\infty}$/$h_{\infty}$')
plt.xlabel('Time [ms]')
plt.legend()
plt.show()
```

Here's the mod file:

Code: Select all

```
COMMENT
na3h5.mod
Sodium channel, Hodgkin-Huxley style kinetics.
Kinetics were fit to data from Huguenard et al. (1988) and Hamill et
al. (1991)
qi is not well constrained by the data, since there are no points
between -80 and -55. So this was fixed at 5 while the thi1,thi2,Rg,Rd
were optimized using a simplex least square proc
voltage dependencies are shifted approximately +5mV from the best
fit to give higher threshold
use with kd3h5.mod
Author: Zach Mainen, Salk Institute, 1994, zach@salk.edu
ENDCOMMENT
INDEPENDENT {t FROM 0 TO 1 WITH 1 (ms)}
NEURON {
SUFFIX na3
USEION na READ ena WRITE ina
RANGE m, h, gna, gbar, vshift
GLOBAL tha, thi1, thi2, qa, qi, qinf, thinf
RANGE minf, hinf, mtau, htau
GLOBAL Ra, Rb, Rd, Rg
GLOBAL q10, temp, tadj, vmin, vmax
}
PARAMETER {
gbar = 1200 (pS/um2) : 0.12 mho/cm2
vshift = 0 (mV) : voltage shift (affects all)
tha = -35 (mV) : v 1/2 for act (-42)
qa = 9 (mV) : act slope
Ra = 0.182 (/ms) : open (v)
Rb = 0.124 (/ms) : close (v)
thi1 = -50 (mV) : v 1/2 for inact
thi2 = -75 (mV) : v 1/2 for inact
qi = 5 (mV) : inact tau slope
thinf = -65 (mV) : inact inf slope
qinf = 6.2 (mV) : inact inf slope
Rg = 0.0091 (/ms) : inact (v)
Rd = 0.024 (/ms) : inact recov (v)
temp = 23 (degC) : original temp
q10 = 2.3 : temperature sensitivity
v (mV)
dt (ms)
celsius (degC)
vmin = -120 (mV)
vmax = 100 (mV)
}
UNITS {
(mA) = (milliamp)
(mV) = (millivolt)
(pS) = (picosiemens)
(um) = (micron)
}
ASSIGNED {
ina (mA/cm2)
gna (pS/um2)
ena (mV)
minf hinf
mtau (ms) htau (ms)
tadj
}
STATE { m h }
INITIAL {
trates(v+vshift)
m = minf
h = hinf
}
BREAKPOINT {
SOLVE states
gna = gbar*m*m*m*h
ina = (1e-4) * gna * (v - ena)
}
LOCAL mexp, hexp
PROCEDURE states() { :Computes state variables m, h, and n
trates(v+vshift) : at the current v and dt.
m = m + mexp*(minf-m)
h = h + hexp*(hinf-h)
VERBATIM
return 0;
ENDVERBATIM
}
PROCEDURE trates(v) {
LOCAL tinc
TABLE minf, mexp, hinf, hexp
DEPEND dt, celsius, temp, Ra, Rb, Rd, Rg, tha, thi1, thi2, qa, qi, qinf
FROM vmin TO vmax WITH 199
rates(v): not consistently executed from here if usetable == 1
tadj = q10^((celsius - temp)/10)
tinc = -dt * tadj
mexp = 1 - exp(tinc/mtau)
hexp = 1 - exp(tinc/htau)
}
PROCEDURE rates(vm) {
LOCAL a, b
a = trap0(vm,tha,Ra,qa)
b = trap0(-vm,-tha,Rb,qa)
mtau = 1/(a+b)
minf = a*mtau
:"h" inactivation
a = trap0(vm,thi1,Rd,qi)
b = trap0(-vm,-thi2,Rg,qi)
htau = 1/(a+b)
hinf = 1/(1+exp((vm-thinf)/qinf))
}
FUNCTION trap0(v,th,a,q) {
if (fabs(v/th) > 1e-6) {
trap0 = a * (v - th) / (1 - exp(-(v - th)/q))
} else {
trap0 = a * q
}
}
```

Code: Select all

```
PROCEDURE trates(v) {
LOCAL tinc
TABLE minf, mexp, hinf, hexp, mtau, htau
DEPEND dt, celsius, temp, Ra, Rb, Rd, Rg, tha, thi1, thi2, qa, qi, qinf
FROM vmin TO vmax WITH 199
rates(v): not consistently executed from here if usetable == 1
tadj = q10^((celsius - temp)/10)
tinc = -dt * tadj
mexp = 1 - exp(tinc/mtau)
hexp = 1 - exp(tinc/htau)
}
```

Thanks,

Robert