Synaptic scaling balances learning in a spiking model of neocortex (Rowan & Neymotin 2013)

 Download zip file   Auto-launch 
Help downloading and running models
Accession:147141
Learning in the brain requires complementary mechanisms: potentiation and activity-dependent homeostatic scaling. We introduce synaptic scaling to a biologically-realistic spiking model of neocortex which can learn changes in oscillatory rhythms using STDP, and show that scaling is necessary to balance both positive and negative changes in input from potentiation and atrophy. We discuss some of the issues that arise when considering synaptic scaling in such a model, and show that scaling regulates activity whilst allowing learning to remain unaltered.
Reference:
1 . Rowan MS,Neymotin SA (2013) Synaptic Scaling Balances Learning in a Spiking Model of Neocortex Adaptive and Natural Computing Algorithms, Tomassini M, Antonioni A, Daolio F, Buesser P, ed. pp.20
Model Information (Click on a link to find other models with that property)
Model Type: Realistic Network;
Brain Region(s)/Organism: Neocortex;
Cell Type(s): Neocortex V1 pyramidal corticothalamic L6 cell; Neocortex V1 pyramidal intratelencephalic L2-5 cell; Neocortex V1 interneuron basket PV cell; Neocortex fast spiking (FS) interneuron; Neocortex spiny stellate cell; Neocortex spiking regular (RS) neuron; Neocortex spiking low threshold (LTS) neuron; Abstract integrate-and-fire adaptive exponential (AdEx) neuron;
Channel(s):
Gap Junctions:
Receptor(s): GabaA; AMPA; NMDA;
Gene(s):
Transmitter(s): Gaba; Glutamate;
Simulation Environment: NEURON; Python;
Model Concept(s): Synaptic Plasticity; Long-term Synaptic Plasticity; Learning; STDP; Homeostasis;
Implementer(s): Lytton, William [billl at neurosim.downstate.edu]; Neymotin, Sam [samn at neurosim.downstate.edu]; Rowan, Mark [m.s.rowan at cs.bham.ac.uk];
Search NeuronDB for information about:  Neocortex V1 pyramidal corticothalamic L6 cell; Neocortex V1 pyramidal intratelencephalic L2-5 cell; Neocortex V1 interneuron basket PV cell; GabaA; AMPA; NMDA; Gaba; Glutamate;
/
stdpscalingpaper
batchscripts
mod
README
alz.hoc
autotune.hoc *
basestdp.hoc *
batch.hoc *
batch2.hoc *
batchcommon
checkirreg.hoc *
clusterrun.sh
col.dot *
col.hoc *
comppowspec.hoc *
condisconcellfig.hoc *
condisconpowfig.hoc *
declist.hoc *
decmat.hoc *
decnqs.hoc *
decvec.hoc *
default.hoc *
drline.hoc *
e2hubsdisconpow.hoc *
e2incconpow.hoc *
filtutils.hoc *
geom.hoc *
graphplug.hoc *
grvec.hoc *
init.hoc *
labels.hoc *
load.hoc *
local.hoc *
makepopspikenq.hoc *
matfftpowplug.hoc *
matpmtmplug.hoc *
matpmtmsubpopplug.hoc *
matspecplug.hoc *
network.hoc *
nload.hoc *
nqpplug.hoc *
nqs.hoc *
nqsnet.hoc *
nrnoc.hoc *
params.hoc
plot.py
plotbatch.sh
plotbatchcluster.sh
powchgtest.hoc *
python.hoc *
pywrap.hoc *
redE2.hoc *
run.hoc
runsim.sh
setup.hoc *
shufmua.hoc *
sim.hoc
simctrl.hoc *
spkts.hoc *
stats.hoc *
stdpscaling.hoc
syncode.hoc *
vsampenplug.hoc *
writedata.hoc
xgetargs.hoc *
                            
// $Id: nrnoc.hoc,v 1.74 2007/11/20 07:51:52 billl Exp $

proc nrnoc () {}

// Users should not edit nrnoc.hoc or default.hoc.  Any local 
// changes to these files should be made in local.hoc.

// key '*&*' is picked up by to indicate command for emacs
proc elisp () { printf("*&* %s\n",$s1) }
// if (not exists(simname)) { strdef simname, output_file, datestr, comment }

// Simctrl.hoc will automatically load stdgraph.hoc which automatically
// loads stdrun.hoc
strdef temp_string_, user_string_  // needed for simctrl
/* Global variable default values.  NOTE that stdrun.hoc, stdgraph.hoc
and simctrl.hoc all contain variable definitions and thus default.hoc
should be loaded after these files */
load_file("default.hoc")      /* Load default.hoc */

/* Allows arrays of strings */
objref hoc_obj_[2]
load_file("stdgui.hoc") // don't want to encounter other String tempate defs
load_file("simctrl.hoc")

proc run () {
  running_ = 1
  stdinit()
  continueRun(tstop)
  finish()
}

proc continueRun () { local rt, rtstart, ts
  if (numarg()==1) ts=$1 else ts=t+1e3
  realtime = 0  rt = screen_update_invl  rtstart = startsw()
  eventcount=0
  eventslow=1
  stoprun = 0
  if (using_cvode_) {
    if (cvode.use_local_dt || (cvode.current_method()%10) == 0) {
      cvode.solve(ts)
      flushPlot()
      realtime = startsw() - rtstart
      return
    }
  } else {
    ts -= dt/2
  }
  while (t<ts && stoprun==0) {
    step()
    realtime = startsw() - rtstart
    if (realtime >= rt) {
      //                        if (!stdrun_quiet) fastflushPlot()
      screen_update()
      //really compute for at least screen_update_invl
      realtime = startsw() - rtstart
      rt = realtime + screen_update_invl
    }
  }
  if (using_cvode_ && stoprun == 0) { // handle the "tstop" event
    step() // so all recordings take place at tstop
  }
  flushPlot()
  realtime = startsw() - rtstart
}

proc stdinit() {
        cvode_simgraph()
        realtime = 0
        setdt()
        init()
        initPlot()
}

proc init () {
  cvode_simgraph()
  initMech()
  initMisc1()

  // Initialize state vars then calculate currents
  // If user hand-set v in initMisc1() then v_init should be > 1000,
  // else all compartments will be set to v_init
  if (v_init < 1000) {
    finitialize(v_init)
  } else {
    finitialize()
  }

  // Set ca pump and leak channel for steady state
  setMemb()

  initMisc2()
  if (cvode_active()) cvode.re_init() else fcurrent()
  frecord_init()
}

// Initialization of mechanism variables
// NOTE: if any changes are made to the NEURON block of any local mod
// file, the user must add the necessary inits to initMisc1()
proc initMech () { 
  forall {
    if ((!ismembrane("pas")) && (!ismembrane("Passive"))) { 
      // Allow for either pas or Passive mod file usage
      // errorMsg("passive not inserted") 
    }

    if (ismembrane("na_ion")) { 
      nai = na_init
      nai0_na_ion = na_init
    }
    
    if (ismembrane("k_ion")) {
      ki = k_init
      ki0_k_ion = k_init
    }
    
    if (ismembrane("ca_ion")) { 
      cai = ca_init
      cai0_ca_ion = ca_init
    }
  }
}

//* setMemb complex -- multiple names for passive mech
//** declarations
iterator scase() { local i
  for i = 1, numarg() { temp_string_ = $si iterator_statement }}
objref paslist,pasvars[3],XO
double pasvals[2],x[1]
paslist = new List()
for ii=0,2 pasvars[ii]= new String()
for scase("fastpas","pas","Pass","Passive") paslist.append(new String(temp_string_))

//** getval(),setval() -- return/set the hoc value of a string
func retval () { return getval($s1) }
func getval () { 
  sprint(temp_string2_,"x=%s",$s1)
  execute(temp_string2_)
  return x
}
proc setval () { 
  sprint(temp_string2_,"%s=%g",$s1,$2)
  execute(temp_string2_)
}

//** findpas()
// assumes that we are starting in a live section since looks for pass mech there
qx_=0
proc findpas () {
  for ii=0,paslist.count-1 {
    XO=paslist.object(ii)
    if (ismembrane(XO.s)) {
      // print XO.s,"found"
      pasvars[2].s=XO.s
      sprint(pasvars[0].s,"g_%s(qx_)",XO.s)
      for scase("e","erev","XXXX") {  // look for the proper prefix
        sprint(temp_string_,"%s_%s",temp_string_,XO.s)
        if (name_declared(temp_string_)==1) break
      }
      if (name_declared(temp_string_)==0) { // not found
        printf("SetMemb() in nrnoc.hoc: Can't find proper 'erev' prefix for %s\n",XO.s)
      } else {
        sprint(pasvars[1].s,"%s(qx_)",temp_string_)
      }
    }
  }
}

proc setMemb () {
  if (!secp()) return
  findpas() // assume that passive name is the same in all sections
  forall for (qx_,0) {  // will eventually want 'for (x)' to handle all the segments
    if (ismembrane(pasvars[2].s)) {
        for ii=0,1 pasvals[ii]=getval(pasvars[ii].s)
        setmemb2()
        for ii=0,1 setval(pasvars[ii].s,pasvals[ii])
    }
  }
}

// secp() determine whether any sections exist
func secp () { local n
  n=0
  forall n+=1
  if (n>0) return 1 else return 0
}

func setother () {return 0} // callback stub
proc setmemb2 () { local iSum, ii, epas, gpas
  if (!secp()) return
  gpas=pasvals[0] epas=pasvals[1]
  // Setup steady state voltage using leak channel
  iSum = 0.0
  if (ismembrane("na_ion")) { iSum += ina(qx_) }
  if (ismembrane("k_ion"))  { iSum += ik(qx_)  }
  if (ismembrane("ca_ion")) { iSum += ica(qx_) }
  iSum += setother()

  if (iSum == 0) {        // Passive cmp so set e_pas = v
    epas = v
  } else {
    if (gpas > 0) {    // Assume g set by user, calc e
      epas = v + iSum/gpas

    } else {            // Assume e set by user, calc g
      if (epas != v) {
        gpas = iSum/(epas - v)
      } else { gpas=0 }
    }
    if (gpas < 0) errorMsg("bad g", gpas)
    if (epas < -100 || epas > 0) {
      printf(".")
      // printf("%s erev: %g %g %g\n",secname(),e_pas,ina,ik)
    }
  }
  pasvals[0]=gpas pasvals[1]=epas
}

proc finish () {
  /* Called following completion of continueRun() */

finishMisc()

if (graph_flag == 1) {
  if (iv_flag == 1) {
    flushPlot()
    doEvents()
  } else {
    graphmode(-1)
    plt(-1)
  }
}

if (print_flag == 1) {
  wopen("")
}
}

/*------------------------------------------------------------
User definable GRAPHICS and PRINTING routines
------------------------------------------------------------*/

proc outputData() {
  // Default procedure - if outputData() doesn't exist in the run file

  if (graph_flag == 1) {
    if (iv_flag == 1) {
      Plot()
      rt = stopsw()
      if (rt > realtime) {
        realtime = rt
        fastflushPlot()
        doNotify()
        if (realtime == 2 && eventcount > 50) {
          eventslow = int(eventcount/50) + 1
        }
        eventcount = 0
      }else{
        eventcount = eventcount + 1
        if ((eventcount%eventslow) == 0) {
          doEvents()
        }
      }

    } else {
      graph(t)
    }
  }

  if (print_flag == 1) { 
    if (t%printStep <= printStep) { printOut() }
  }
}

proc printOut() {
  /* Default procedure - if printOut() doesn't exist in the run file */
}

proc initGraph() {
  /* Default procedure - if initGraph() doesn't exist in the run file */

graph()
}

proc initPrint() {
  /* Default procedure - if initPrint() doesn't exist in the run file */

wopen(output_file)
}

/*------------------------------------------------------------
User definable BATCH RUN routines
------------------------------------------------------------*/

proc nextrun() {
  // Called from finishmisc() following completion of batch in an autorun
  wopen("")   
  runnum = runnum + 1
  sprint(output_file,"data/b%s.%02d", datestr, runnum)
}                       

// commands for emacs
proc update_runnum() { 
  runnum = $1
  sprint(output_file,"data/%s.%02d", datestr, runnum)
  print "^&^ (progn (sim-index-revert)(setq sim-runnum ",runnum,"))" }
proc nrn_write_index() { printf("&INDEX& %s\n",$s1) }
proc nrn_update () { elisp("nrn-update") }
proc nrn_message () { printf("!&! %s\n",$s1) } 

/*------------------------------------------------------------
User definable INITIALIZATION and FINISH routines
------------------------------------------------------------*/

// Default procedure - if initMisc1() doesn't exist in the run file 
// Initializations performed prior to finitialize() 
// This should contain point process inits and inits for any changes 
//        made to the NEURON block of any local mod file 
proc initMisc1() { }

// Default procedure - if initMisc2() doesn't exist in the run file 
// Initializations performed after finitialize() 
proc initMisc2() { }

// Default procedure - if finishMisc() doesn't exist in the run file 
proc finishMisc() { }

/*------------------------------------------------------------
Miscellaneous routines
------------------------------------------------------------*/

proc errorMsg() {
  /* Print warning, assumes arg1 is string and arg2 if present is a
  variable value */

sectionname(section)

if (numarg() == 0) {
  printf("ERROR in errorMsg(): Needs at least 1 argument.\n")
} else if (numarg() == 1) {
  printf("ERROR: %s in section %s.\n", $s1, section)
} else {
  printf("ERROR: %s in section %s (var=%g).\n", $s1, section, $2)
}
}

proc clear() {
  /* Clear non-interviews plot window */
plt(-3)
}

func mod() { local x, y
  /* Mod function for non-integers */

x=$1
y=$2

return (x/y - int(x/y))
}

proc whatSection() { print secname() }

proc print_pp_location() { local x //arg1 must be a point process
   x = $o1.get_loc()
   sectionname(temp_string_)
   printf("%s located at %s(%g)\n", $o1, temp_string_, x)
   pop_section()
}

//* set method with method()
proc method () { local prc
  if (numarg()==0) {
    if (cvode_active() && cvode_local()) { printf("\tlocal atol=%g\n",cvode.atol)
    } else if (cvode_active()) { printf("\tglobal atol=%g\n",cvode.atol)
    } else if (secondorder==2) { printf("\tCrank-Nicholson dt=%g\n",dt)
    } else if (secondorder==0) { printf("\timplicit dt=%g\n",dt)
    } else { printf("\tMethod unrecognized\n") }
    return
  }
  if (numarg()==2) prc = $2 else prc=0
  finitialize()
  if (strcmp($s1,"global")==0) {
    cvode_active(1)
    cvode.condition_order(2)
    if (prc) cvode.atol(prc)
  } else if (strcmp($s1,"local")==0) {
    cvode_local(1)
    cvode.condition_order(2)
    if (prc) cvode.atol(prc)
  } else if (strcmp($s1,"implicit")==0) {
    secondorder=0
    cvode_active(1)
    cvode_active(0)
    if (prc) dt=prc
  } else if (strcmp($s1,"CN")==0) {
    secondorder=2
    cvode_active(1) // this turns off local
    cvode_active(0)
    if (prc) dt=prc
  } else {
    printf("Integration method %s not recognized\n",$s1)
  }
}

//* Load local modifications to nrnoc.hoc and default.hoc
load_file("local.hoc")

if (xwindows && graph_flag) { nrnmainmenu() } // pwman_place(50,50)

print "Init complete.\n"

Rowan MS,Neymotin SA (2013) Synaptic Scaling Balances Learning in a Spiking Model of Neocortex Adaptive and Natural Computing Algorithms, Tomassini M, Antonioni A, Daolio F, Buesser P, ed. pp.20

References and models cited by this paper

References and models that cite this paper

Binzegger T, Douglas RJ, Martin KA (2004) A quantitative map of the circuit of cat primary visual cortex. J Neurosci 24:8441-53 [PubMed]

Busche MA, Eichhoff G, Adelsberger H, Abramowski D, Wiederhold KH, Haass C, Staufenbiel M, Ko (2008) Clusters of hyperactive neurons near amyloid plaques in a mouse model of Alzheimer's disease. Science 321:1686-9

Carnevale NT, Hines ML (2006) The NEURON Book

Chandler B, Grossberg S (2012) Joining distributed pattern processing and homeostatic plasticity in recurrent on-center off-surround shunting networks: noise, saturation, short-term memory, synaptic scaling, and BDNF. Neural Netw 25:21-9

Dan Y, Poo MM (2004) Spike timing-dependent plasticity of neural circuits. Neuron 44:23-30 [PubMed]

Frohlich F, Bazhenov M, Sejnowski TJ (2008) Pathological effect of homeostatic synaptic scaling on network dynamics in diseases of the cortex. J Neurosci 28:1709-20 [PubMed]

Lamsa KP, Kullmann DM, Woodin MA (2010) Spike-timing dependent plasticity in inhibitory circuits. Front Synaptic Neurosci 2:8

Lefort S, Tomm C, Floyd Sarria JC, Petersen CC (2009) The excitatory neuronal network of the C2 barrel column in mouse primary somatosensory cortex. Neuron 61:301-16 [PubMed]

Lytton WW, Omurtag A, Neymotin SA, Hines ML (2008) Just in time connectivity for large spiking networks Neural Comput 20(11):2745-56 [Journal] [PubMed]

   JitCon: Just in time connectivity for large spiking networks (Lytton et al. 2008) [Model]

Lytton WW, Stewart M (2006) Rule-based firing for network simulations. Neurocomputing 69:1160-1164

McClelland JL, McNaughton BL, O'Reilly RC (1995) Why there are complementary learning systems in the hippocampus and neocortex: insights from the successes and failures of connectionist models of learning and memory. Psychol Rev 102:419-57 [PubMed]

Neymotin S, Kerr C, Francis J, Lytton W (2011) Training oscillatory dynamics with spike-timing-dependent plasticity in a computer model of neocortex Signal Processing in Medicine and Biology Symposium (SPMB), IEEE :1-6

Neymotin SA, Lee H, Park E, Fenton AA, Lytton WW (2011) Emergence of physiological oscillation frequencies in a computer model of neocortex. Front Comput Neurosci 5:19-75 [Journal] [PubMed]

   Emergence of physiological oscillation frequencies in neocortex simulations (Neymotin et al. 2011) [Model]

Prieto G, Parker R, Vernon_iii F (2009) A Fortran 90 library for multitaper spectrum analysis Computers Geosciences 35:1701-1710

Rowan M (2011) Effects of Compensation, Connectivity and Tau in a Computational Model of Alzheimer's Disease Proc. IJCNN, IEEE :543-550

Rowan M (2012) Information-selectivity of Beta-amyloid pathology in an associative memory model. Front Comput Neurosci 6:2-23 [PubMed]

Rutherford LC, Nelson SB, Turrigiano GG (1998) BDNF has opposite effects on the quantal amplitude of pyramidal neuron and interneuron excitatory synapses. Neuron 21:521-30 [PubMed]

Small DH (2008) Network dysfunction in Alzheimer's disease: does synaptic scaling drive disease progression? Trends Mol Med 14:103-8 [PubMed]

Trasande CA, Ramirez JM (2007) Activity deprivation leads to seizures in hippocampal slice cultures: is epilepsy the consequence of homeostatic plasticity? J Clin Neurophysiol 24:154-64 [PubMed]

Turrigiano G (2011) Too many cooks? Intrinsic and synaptic homeostatic mechanisms in cortical circuit refinement. Annu Rev Neurosci 34:89-103 [PubMed]

Turrigiano GG (2008) The self-tuning neuron: synaptic scaling of excitatory synapses. Cell 135:422-35 [PubMed]

van Rossum MC, Bi GQ, Turrigiano GG (2000) Stable Hebbian learning from spike timing-dependent plasticity. J Neurosci 20:8812-21 [PubMed]

Zhang LI, Tao HW, Holt CE, Harris WA, Poo M (1998) A critical window for cooperation and competition among developing retinotectal synapses. Nature 395:37-44 [PubMed]

Eguchi A, Neymotin SA and Stringer SM (2014) Color opponent receptive fields self-organize in a biophysical model of visual cortex via spike-timing dependent plasticity 8:16. doi: Front. Neural Circuits 8:16 [Journal]

   Simulated cortical color opponent receptive fields self-organize via STDP (Eguchi et al., 2014) [Model]

Neymotin SA, Chadderdon GL, Kerr CC, Francis JT, Lytton WW (2013) Reinforcement learning of 2-joint virtual arm reaching in a computer model of sensorimotor cortex Neural Computation 25(12):3263-93 [Journal] [PubMed]

   Sensorimotor cortex reinforcement learning of 2-joint virtual arm reaching (Neymotin et al. 2013) [Model]

Neymotin SA, McDougal RA, Sherif MA, Fall CP, Hines ML, Lytton WW (2015) Neuronal calcium wave propagation varies with changes in endoplasmic reticulum parameters: a computer model Neural Computation 27(4):898-924 [Journal] [PubMed]

   Neuronal dendrite calcium wave model (Neymotin et al, 2015) [Model]

Rowan MS, Neymotin SA, Lytton WW (2014) Electrostimulation to reduce synaptic scaling driven progression of Alzheimer's disease. Front Comput Neurosci 8:39 [Journal] [PubMed]

   Electrostimulation to reduce synaptic scaling driven progression of Alzheimers (Rowan et al. 2014) [Model]

(27 refs)