// $Id: network.hoc,v 1.126 2010/12/29 03:14:37 cliffk Exp $
print "Loading network.hoc..."
// Parameters for controlling where connectivity comes from
loadsaveconnectivity=0 // Whether to make (0), load (1), or save (2) the connectivity table
//* Numbers and connectivity params
{declare("scale",1)}
{declare("colW",1,"colH",1,"torus",0)}
{declare("numcols",colW*colH)}
{declare("dbgcols",0)} // whether to debug columns by making them have the same wiring and inputs
{declare("colr",2)} // maximal trans-column projection distance; 0 within col; 1 next col etc
{declare("colnq","o[5]","lcol",new List())}
{declare("thalamusOn", 0)} // 1 = thalamic cells connected, 0 = completely disconnects all thalamic cells
{declare("autotune",0)} // From stccol
{declare("useSTDP",0)} // From stccol -- turn STDP on by default
{sprint(tstr,"o[%d]",numcols) declare("col",tstr)}
{sprint(tstr,"o[%d][%d]",colH,colW) declare("gcol",tstr)} // 2D column grid
double div[CTYPi][CTYPi][colr+1]//div[i][j]==# of outputs from type i->j
double wmat[CTYPi][CTYPi][STYPi][colr+1] // wmat[i][j][k]==weight from type i->j for synapse k
double delm[CTYPi][CTYPi]//avg. delay from type i->j
double deld[CTYPi][CTYPi]//delay variance from type i->j
double conv[CTYPi][CTYPi][colr+1]
dosetpmat=name_declared("pmat")==0
{sprint(tstr,"d[%d][%d][%d]",CTYPi,CTYPi,colr+1) declare("pmat",tstr)}
double prumat[CTYPi][CTYPi] //pruning matrix:prumat[i][j] specifies ratio (0-1) of synapses to prune
double sprmat[CTYPi][CTYPi] //sprouting matrix:sprmat[i][j] specifies ratio (0-1) to sprout i->j pathway with
double synloc[CTYPi][CTYPi]//location of synapses
//* swire & related column variables -- from stccol
declare("colside",400*sqrt(scale)) // column diameter in micrometers
declare("swire",0) // whether to use 'spatial' wiring, 2 means use swirecut
declare("checkers",0) // whether to arrange cells in checkerboard pattern
declare("EEsq",(colside/2)^2,"EIsq",(colside/2)^2,"IEsq",(colside/1)^2,"IIsq",(colside/2)^2) // params for swirecut
declare("slambda",15)
double dels[CTYPi][CTYPi]//[STYPi] //stdev of delays
double delv[CTYPi][CTYPi]//variance of delays
double vcond[CTYPi][CTYPi]//[STYPi] //conduction velocities
declare("layerzvar",500) // range of z location of cells in a layer (in micrometers); CK: was 25
declare("colxydist",50) // distance between adjacent columns (x,y)
declare("vlayerz",new Vector(CTYPi))
declare("pseed",4321)
declare("dvcond","d[2]") //intra layer conduction delays
declare("maxsfall",0.001) // max fall-off in prob of connection @ opposite side of column (used by swire)
declare("axonalvelocity",10000) // axonal velocity in um/ms -- this is 10 mm/s
dvcond[0] = 1 //if presynaptic cell is excit, conduction delay == 1.0 m/s
dvcond[1] = 0.4 //if presynaptic cell is inhib, conduction delay == 0.4 m/s
// <-- stccol
//* other variables
declare("mknetnqss",1) // whether COLUMN should make connsnq,cellsnq
declare("EEGain",4*15/11*eemod,"EIGain",15*eimod,"IEGain",4*15/11*iemod,"IIGain",4*15/11*iimod)
declare("NMAMR",0.1,"EENMGain",1,"EIGainInterC",0.125,"EEGainInterC",0.325*0.5)
declare("TCNM",0) // whether TC->NM2 on
declare("E2I4",0) // whether E2 hits I4
declare("E5IRE",0) // whether E5R,E5B hits IRE in neighboring columns
declare("SMFCTR",10) // multiplier for # of SM cells
if(autotune) {
seadsetting_INTF6 = 2
wsetting_INTF6 = 1 // use sywv during sim
} else if(useSTDP) {
seadsetting_INTF6 = 3 // use plasticity
} else {
seadsetting_INTF6 = 2 // fixed weights
}
wsetting_INTF6 = 1 // use sywv during sim (needed z-dependent efferent weights)
declare("wnqstr","") // intracol wiring NQS (single column)
if (loadsaveconnectivity==1) wnqstr="connectivity.nqs" // CK: Load NQS from file
// stccol -->
//* prdiv() - print div
proc prdiv () { local ii,jj
for ii=0,CTYPi-1 for jj=0,CTYPi-1 if(div[ii][jj][0]) {
printf("div[%s][%s][0]=%g\n",CTYP.o(ii).s,CTYP.o(jj).s,div[ii][jj][0])
}
}
// %con (con/pre) = %div (div/post)
DEAD_DIV_INTF6=0
{declare("jcn",1)}
{declare("disinhib",0)} //iff==1 , turn off inhibition, by setting wmat[I%d][...]==0 in inhiboff()
{declare("scale",1)}
declare("dstr",datestr,"setdviPT",NORM)
batch_flag=1 // CK: yes, use batch (=1); was:
{declare("params","not batch","ofile",output_file)}
{declare("dvseed",534023)} // seed for wiring
dosetcpercol=name_declared("cpercol")==0 // whether to set values in cpercol or use user-supplied values
{sprint(tstr,"d[%d]",CTYPi) declare("cpercol",tstr)} // cells of a specific type per column
{declare("vcpercol",new Vector(CTYPi))}
// <-- stccol
sprint(tstr,"d[%d][%d]",CTYPi,CTYPi)
if(!(i=name_declared("delmscale"))) {
declare("delmscale",tstr) // scale delm values by this #
}
if(!i) {
for i=0,CTYPi-1 for j=0,CTYPi-1 delmscale[i][j]=1
}
//** mklayerz -- get average z location of a layer, based on frana and tononi
// returns value in micrometers. 0 is at top (layer 1), max val
// is at layer 6 ... actual #s don't matter, only distances between
// the layers matters...
// Layers are defined in nqsnet.hoc, "layer" procedure
proc mkvlayerz () { local i,vl, cortthaldist
cortthaldist=50000 // CK: WARNING, KLUDGY -- Distance from relay nucleus to cortex -- ~5 cm = 50,000 um
vlayerz.indgen(0,CTYPi-1,1)
vlayerz.apply("layer")
for vtr(&vl,vlayerz,&i) {
if(int(vl) == 2 || int(vl) == 3) {vlayerz.x(i)=1740+cortthaldist // average of 1540+1940 from frana
} else if(int(vl) == 4) {vlayerz.x(i) = (1740+1130)/2+cortthaldist // Halfway between 2/3 and 5
} else if(int(vl) == 5) {vlayerz.x(i) = 1130+cortthaldist
} else if(int(vl) == 6) {vlayerz.x(i) = 488+cortthaldist
} else if(int(vl) == 7) {vlayerz.x(i) = 300// WARNING, KLUDGY: guess distance from relay to reticular nuclei
} else if(int(vl) == 8) {vlayerz.x(i) = 0
} else {vlayerz.x(i) = -1000} // This shold never happen
}
}
// stccol -->
//* setcpercol - set # of cells per column
proc setcpercol () { local i // (/u/samn/vcsim/notebook.dol_1:24562)(notebook.dol_1:24492)
if(dosetcpercol) { // if user didn't supply values (default), set # of cells of a type per column
cpercol[E2] = 150 * scale
cpercol[I2] = 25 * scale
cpercol[I2L] = 25 * scale
cpercol[E5R] = 167 * scale
cpercol[E5B] = 72 * scale
cpercol[I5] = 40 * scale
cpercol[I5L] = 40 * scale
cpercol[E6] = 192 * scale
cpercol[I6] = 32 * scale
cpercol[I6L] = 32 * scale
if (thalamusOn == 1) {
cpercol[TC] = 10 * scale
cpercol[IRE] = 10 * scale
}
/* cpercol[SM] = SMFCTR * cpercol[TC] * scale // CK: Let's just ignore somatosensory for now*/
}
{vcpercol.resize(CTYPi) vcpercol.fill(0)} // store the values in a vector
for i=0,CTYPi-1 vcpercol.x(i)=cpercol[i]
}
//* setpmat()
proc setpmat () { local pre,po,ii,jj,kk
if(!dosetpmat) return // if pmat setup by user (in notebook), then don't reset its values
for ii=0,CTYPi-1 for jj=0,CTYPi-1 for kk=0,1 pmat[ii][jj][kk]=0
// Add connections to/from the thalamus
if (thalamusOn == 1) {
// SM = skin. SM -> TC
pmat[SM][TC][0] = 0.5 / 6
pmat[SM][TC][1] = 0.5 / 6
pmat[SM][TC][2] = 0.5 / 6
pmat[E6][TC][0] = 0.1
pmat[E6][IRE][0] = 0.1
pmat[TC][IRE][0] = 0.4
pmat[IRE][IRE][0] = 0.1
pmat[IRE][TC][0] = 0.3
pmat[TC][E2][0] = 0.1
pmat[TC][E4][0] = 0.2
pmat[TC][E5B][0] = 0.1
pmat[TC][E5R][0] = 0.1
pmat[TC][E6][0] = 0.1
pmat[TC][I2][0] = 0.1
pmat[TC][I4][0] = 0.1
pmat[TC][I5][0] = 0.1
pmat[TC][I6][0] = 0.1
}
pmat[E2][E2][0]=0.2 // weak by wiring matrix in (Weiler et al., 2008)
pmat[E2][E2][1]=0//0.14
pmat[E2][E5B][0]=0.8 // strong by wiring matrix in (Weiler et al., 2008)
pmat[E2][E5R][0]=0.8 // strong by wiring matrix in (Weiler et al., 2008)
pmat[E2][I5L][0]=0.51 // L2/3 E -> L5 LTS I (justified by (Apicella et al., 2012)
pmat[E2][E6][0]=0 // none by wiring matrix in (Weiler et al., 2008)
pmat[E2][I2L][0]=0.51
pmat[E2][I2][0]=0.43
pmat[E2][I2][1]=0.14
pmat[E5B][E2][0]=0 // none by wiring matrix in (Weiler et al., 2008)
pmat[E5B][E2][1]=0.25
pmat[E5B][E2][2]=0.1
pmat[E5B][E5B][0]=0.04 * 4 // set using (Kiritani et al., 2012) Fig. 6D, Table 1, value x 4
pmat[E5B][E5B][1]=0.25
pmat[E5B][E5B][2]=0.1
pmat[E5B][E5R][0]=0 // set using (Kiritani et al., 2012) Fig. 6D, Table 1
pmat[E5B][E5R][1]=0.25
pmat[E5B][E5R][2]=0.1
pmat[E5B][E6][0]=0 // none by suggestion of Ben and Gordon over phone
pmat[E5B][I5L][0]=0 // ruled out by (Apicella et al., 2012) Fig. 7
pmat[E5B][I5L][1]=0.14
pmat[E5B][I5L][2]=0.07
pmat[E5B][I5][0]=0.43 // CK: was 0.43 but perhaps the cause of the blow-up?
pmat[E5B][I5][1]=0.14
pmat[E5B][I5][2]=0.07
pmat[E5R][E2][0]=0.2 // weak by wiring matrix in (Weiler et al., 2008)
pmat[E5R][E5B][0]=0.21 * 4 // set using (Kiritani et al., 2012) Fig. 6D, Table 1, value x 4
pmat[E5R][E5B][1]=0.25
pmat[E5R][E5R][0]=0.11 * 4 // set using (Kiritani et al., 2012) Fig. 6D, Table 1, value x 4
pmat[E5R][E5R][1]=0.14
pmat[E5R][E6][0]=0.2 // weak by wiring matrix in (Weiler et al., 2008)
pmat[E5R][I5L][0]=0 // ruled out by (Apicella et al., 2012) Fig. 7
pmat[E5R][I5][0]=0.43
pmat[E5R][I5][1]=0.14
pmat[E6][E2][0]=0 // none by wiring matrix in (Weiler et al., 2008)
pmat[E6][E5B][0]=0.2 // weak by wiring matrix in (Weiler et al., 2008)
pmat[E6][E5R][0]=0.2 // weak by wiring matrix in (Weiler et al., 2008)
pmat[E6][E6][0]=0.2 // weak by wiring matrix in (Weiler et al., 2008)
pmat[E6][I6L][0]=0.51
pmat[E6][I6][0]=0.43
pmat[E6][I6][1]=0.14
pmat[I2L][E2][0]=0.35
pmat[I2L][I2L][0]=0.09
pmat[I2L][I2][0]=0.53
pmat[I2][E2][0]=0.44
pmat[I2][I2L][0]=0.34
pmat[I2][I2][0]=0.62
pmat[I5L][E5B][0]=0.35
pmat[I5L][E5R][0]=0.35
pmat[I5L][I5L][0]=0.09
pmat[I5L][I5][0]=0.53
pmat[I5][E5B][0]=0.44
pmat[I5][E5R][0]=0.44
pmat[I5][I5L][0]=0.34
pmat[I5][I5][0]=0.62
pmat[I6L][E6][0]=0.35
pmat[I6L][I6L][0]=0.09
pmat[I6L][I6][0]=0.53
pmat[I6][E6][0]=0.44
pmat[I6][I6L][0]=0.34
pmat[I6][I6][0]=0.62
}
//* scalepmat(fctr) - multiply values in pmat by fctr
proc scalepmat () { local fctr,from,to,cl
fctr=$1
for from=0,CTYPi-1 for to=0,CTYPi-1 for cl=0,1 pmat[from][to][cl] *= fctr
}
//* pmat2nq - return an NQS with info in pmat
obfunc pmat2nq () { local i,j,k localobj nqpmat
nqpmat=new NQS("froms","tos","from","to","cold","pij")
{nqpmat.strdec("froms") nqpmat.strdec("tos")}
for i=0,CTYPi-1 for j=0,CTYPi-1 for k=0,colr if(pmat[i][j][k]) {
nqpmat.append(CTYP.o(i).s,CTYP.o(j).s,i,j,k,pmat[i][j][k])
}
return nqpmat
}
//* nq2pmat - load NQS ($o1) into pmat
proc nq2pmat () { local i,j,k localobj nq,vf,vto,vc,vpij
{nq=$o1 nq.tog("DB") vf=nq.getcol("from") vto=nq.getcol("to") vc=nq.getcol("cold") vpij=nq.getcol("pij")}
for i=0,CTYPi-1 for j=0,CTYPi-1 for k=0,colr pmat[i][j][k]=0
for i=0,vf.size-1 pmat[vf.x(i)][vto.x(i)][vc.x(i)]=vpij.x(i)
print "loaded " , nq , " into pmat"
}
//* synapse locations DEND SOMA AXON
proc setsynloc () { local from,to
for from=0,CTYPi-1 for to=0,CTYPi-1 {
if(ice(from)) {
if(IsLTS(from)) {
synloc[from][to]=DEND // distal [GA2] - from LTS
} else {
synloc[from][to]=SOMA // proximal [GA] - from FS
}
} else {
synloc[from][to]=DEND // E always distal. use AM2,NM2
}
}
}
//* setdelmats -- setup delm,deld
proc setdelmats () { local from,to,ii,jj
for from=0,CTYPi-1 for to=0,CTYPi-1 {
if(synloc[from][to]==DEND) {
delm[from][to]=4 * delmscale[from][to] // stccol
deld[from][to]=1
} else {
delm[from][to]=2.0 * delmscale[from][to] // stccol
deld[from][to]=0.2
}
}
// snum=0
// for ii=0,CTYPi-1 for jj=0,CTYPi-1 snum+=int(pmat[ii][jj][0]*numc[ii]*numc[jj]+1)
}
//* weight params
//** delay all 2+/-0.02 within column for now
proc setwmat () { local from,to,sy,gn,c
for from=0,CTYPi-1 for to=0,CTYPi-1 for sy=0,STYPi-1 for c=0,colr wmat[from][to][sy][c]=0
// <-- stccol
// Thalamic connection weights
if (thalamusOn == 1) {
//** skin -> thalamus
wmat[SM][TC][AM2][0] = 10
wmat[E6][TC][AM2][0] = 0.75
wmat[E6][TC][NM][0] = 0.075
wmat[E6][IRE][AM2][0] = 0.75
wmat[E6][IRE][NM][0] = 0.075
wmat[TC][E2][AM2][0] = 0.5
wmat[TC][E2][NM][0] = 0.05
wmat[TC][E4][AM2][0] = 1.0
wmat[TC][E4][NM][0] = 0.1
wmat[TC][E5B][AM2][0] = 1.5
wmat[TC][E5B][NM][0] = 0.15
wmat[TC][E5R][AM2][0] = 1.5
wmat[TC][E5R][NM][0] = 0.15
wmat[TC][E6][AM2][0] = 1.0
wmat[TC][E6][NM][0] = 0.1
wmat[TC][I2][AM2][0] = 1.5
wmat[TC][I4][AM2][0] = 1.5
wmat[TC][I5][AM2][0] = 1.5
wmat[TC][I6][AM2][0] = 1.5
wmat[TC][IRE][AM2][0] = 0.75
wmat[TC][IRE][NM][0] = 0.15
wmat[IRE][TC][GA2][0] = 1.0
wmat[IRE][IRE][GA2][0] = 0.3
}
//*** neocx -> neocx
wmat[E2][E2][AM2][0]=0.66
wmat[E2][E2][AM2][1]=0.47 * EEGainInterC
// wmat[E2][E4][AM2][0]=0.36
wmat[E2][E5B][AM2][0]=0.36
wmat[E2][E5R][AM2][0]=0.93
wmat[E2][I5L][AM2][0]=0.36
wmat[E2][E6][AM2][0]=0
wmat[E2][I2L][AM2][0]=0.23
wmat[E2][I2][AM2][0] = 0.23
wmat[E2][I2][AM2][1] = 1.5 * EIGainInterC
/* if(E2I4) {
// wmat[E2][I4][AM2][0]=0.18
wmat[E2][I4][AM2][1]=0.18 // stccol
// wmat[E2][I4L][AM2][0]=0.09
// wmat[I2][E4][AM2][0]=0.25
} */
/* no layer 4 since M1 model
wmat[E4][E2][AM2][0]=0.58*e4e2wt
wmat[E4][E4][AM2][0]=0.95
wmat[E4][E5B][AM2][0]=1.01
wmat[E4][E5R][AM2][0]=0.54
wmat[E4][E6][AM2][0]=2.27
wmat[E4][I4L][AM2][0]=0.23
wmat[E4][I4][AM2][0] = 0.23
wmat[E4][I4][AM2][1] = 1.5 * EIGainInterC
*/
wmat[E5B][E2][AM2][0]=0 // ruled out based on Ben and Gordon conversation
wmat[E5B][E2][AM2][1]=0.47 * EEGainInterC
wmat[E5B][E2][AM2][2]=0.47 * EEGainInterC
// wmat[E5B][E4][AM2][0]=0.17
wmat[E5B][E5B][AM2][0]=0.66
wmat[E5B][E5B][AM2][1]=0.47 * EEGainInterC
wmat[E5B][E5B][AM2][2]=0.47 * EEGainInterC
wmat[E5B][E5R][AM2][0]=0 // pulled from Fig. 6D, Table 1 of (Kiritani et al., 2012)
wmat[E5B][E5R][AM2][1]=0.47 * EEGainInterC
wmat[E5B][E5R][AM2][2]=0.47 * EEGainInterC
wmat[E5B][E6][AM2][0]=0 // ruled out based on Ben and Gordon conversation
wmat[E5B][I2L][AM2][1]=1.5 * EIGainInterC
wmat[E5B][I2L][AM2][2]=1.5 * EIGainInterC
wmat[E5B][I5L][AM2][0]=0 // ruled out by (Apicella et al., 2012) Fig. 7
wmat[E5B][I5L][AM2][1]=1.5 * EIGainInterC
wmat[E5B][I5L][AM2][2]=1.5 * EIGainInterC
wmat[E5B][I5][AM2][0]=0.23 //(Apicella et al., 2012) Fig. 7F (weight = 1/2 x weight for E5R->I5)
wmat[E5B][I5][AM2][1]=1.5 * EIGainInterC
wmat[E5B][I5][AM2][2]=1.5 * EIGainInterC
wmat[E5R][E2][AM2][0]=0.66
// wmat[E5R][E4][AM2][0]=0.48
wmat[E5R][E5B][AM2][0]=0.66
wmat[E5R][E5B][AM2][1]=0.47 * EEGainInterC
wmat[E5R][E5R][AM2][0]=0.66
wmat[E5R][E5R][AM2][1]=0.47 * EEGainInterC
wmat[E5R][E6][AM2][0]=0.66
wmat[E5R][I5L][AM2][0]=0 // ruled out by (Apicella et al., 2012) Fig. 7
wmat[E5R][I5][AM2][0]=0.46 // (Apicella et al., 2012) Fig. 7E (weight = 2 x weight for E5B->I5)
wmat[E5R][I5][AM2][1]=1.5 * EIGainInterC
// <-- stccol
/* if(E5IRE) {
wmat[E5B][IRE][AM2][1]=0.5
wmat[E5R][IRE][AM2][1]=0.5
wmat[E5B][IRE][AM2][2]=0.25
wmat[E5R][IRE][AM2][2]=0.25
} */
// stccol -->
wmat[E6][E2][AM2][0]=0
// wmat[E6][E4][AM2][0]=0
wmat[E6][E5B][AM2][0]=0.66
wmat[E6][E5R][AM2][0]=0.66
wmat[E6][E6][AM2][0]=0.66
wmat[E6][I6L][AM2][0]=0.23
wmat[E6][I6][AM2][0]=0.23
wmat[E6][I6][AM2][1]=1.5 * EIGainInterC
wmat[I2L][E2][GA2][0]=0.83
/* wmat[I2L][E5B][GA2][0]=0.83
wmat[I2L][E5R][GA2][0]=0.83
wmat[I2L][E6][GA2][0]=0.83 */
wmat[I2L][I2L][GA2][0]=1.5
wmat[I2L][I2][GA2][0]=1.5
/* wmat[I2L][I5][GA2][0]=0.83
wmat[I2L][I6][GA2][0]=0.83 */
wmat[I2][E2][GA][0]=1.5
wmat[I2][I2L][GA][0]=1.5
wmat[I2][I2][GA][0]=1.5
/* no layer 4 since M1 model
wmat[I4L][E4][GA2][0]=0.83
wmat[I4L][I4L][GA2][0]=1.5
wmat[I4L][I4][GA2][0]=1.5
wmat[I4][E4][GA][0]=1.5
wmat[I4][I4L][GA][0]=1.5
wmat[I4][I4][GA][0]=1.5
*/
// wmat[I5L][E2][GA2][0]=0.83
wmat[I5L][E5B][GA2][0]=0.83
wmat[I5L][E5R][GA2][0]=0.83
/* wmat[I5L][E6][GA2][0]=0.83
wmat[I5L][I2][GA2][0]=0.83 */
wmat[I5L][I5L][GA2][0]=1.5
wmat[I5L][I5][GA2][0]=1.5
// wmat[I5L][I6][GA2][0]=0.83
wmat[I5][E5B][GA][0]=1.5
wmat[I5][E5R][GA][0]=1.5
wmat[I5][I5L][GA][0]=1.5
wmat[I5][I5][GA][0]=1.5
/* wmat[I6L][E2][GA2][0]=0.83
wmat[I6L][E5B][GA2][0]=0.83
wmat[I6L][E5R][GA2][0]=0.83 */
wmat[I6L][E6][GA2][0]=0.83
/* wmat[I6L][I2][GA2][0]=0.83
wmat[I6L][I5][GA2][0]=0.83 */
wmat[I6L][I6L][GA2][0]=1.5
wmat[I6L][I6][GA2][0]=1.5
wmat[I6][E6][GA][0]=1.5
wmat[I6][I6L][GA][0]=1.5
wmat[I6][I6][GA][0]=1.5
//set NMDA weights
for from=0,CTYPi-1 for to=0,CTYPi-1 for c=0,colr wmat[from][to][NM2][c]=NMAMR*wmat[from][to][AM2][c]
//gain control
for from=0,CTYPi-1 for to=0,CTYPi-1 for sy=AM,GA2 for c=0,colr if(wmat[from][to][sy][c] > 0) {
if(ice(from)) {
if(ice(to)) {
gn = IIGain
} else {
gn = IEGain
}
if(IsLTS(from) && !IsLTS(to)) gn *= 0.5
} else {
if(ice(to)) {
gn = EIGain
if(IsLTS(to)) gn *= 0.5
} else {
gn = EEGain
if(sy==NM || sy==NM2) gn *= EENMGain // E->E NMDA gain
}
}
wmat[from][to][sy][c] *= gn
}
if(!TCNM) for i=0,colr-1 for case(&j,NM,NM2) wmat[TC][E4][j][i]=0 // stccol
}
// %con (con/pre) = %div (div/post)
//* prune using values in prumat
proc pruc () { local i,j
for i=0,CTYPi-1 for j=0,CTYPi-1{
if(div[i][j][0] && numc[i] && numc[j] && prumat[i][j]){
printf("Warning: pruning random %.2f%% of %s->%s syns\n",prumat[i][j]*100,CTYP.o(i).s,CTYP.o(j).s)
for ixt(i) XO.prune(prumat[i][j],j)
}
}
}
//* get sprouting value assuming 0% sprouting == 50% pruning
func getspr () { local pr
pr = $1
return ((0.5-pr)/.5)*100
}
//* turn off pruning
proc pruoff () { local i,j
for i=0,CTYPi-1 for j=0,CTYPi-1 prumat[i][j]=0
for i=0,allcells-1 INTF6[i].prune(0)
}
//* set all entries in pruning matrix to $1
proc setpru () { local from,to,val
val=$1
pruoff() // first turn off pruning
for from=0,CTYPi-1 for to=0,CTYPi-1 prumat[from][to]=val
}
//* print prumat
proc prumatpr () { local i,j
for i=0,CTYPi-1 { for j=0,CTYPi-1{
printf("%.2f ",prumat[i][j])
}
printf("\n")
}
}
//* clear sprmat entries to 0
proc clrsprmat () { local i,j
for i=0,CTYPi for j=0,CTYPi sprmat[i][j]=0
}
//* unkill/prune all cells
proc unkp () {
for i=0,allcells-1 {
ce.o(i).flag("dead",0)
ce.o(i).prune(0)
}
}
//* kill cells who's ids are in $o1
proc dokill () { local id
for vtr(&id,$o1) ce.o(id).flag("dead",1)
}
//* getkillids - gets ids of cells to kill in $o1 but excludes cells that are stim'ed
//$1=cell type to kill,$2=prct of cells to kill,$o3=vq stim nqs,$4=out vector of kill ids,$5=rnd seed
func getkillids () { local killcnt,i,j,ct,prct localobj vq,vkid,rd
ct=$1 prct=$2 vq=$o3 vkid=$o4 killcnt=int(prct*numc[ct]) vkid.resize(0) j=0 i=ix[ct]
rd=new Random() rd.ACG($5)
while(j<killcnt){
i=rd.discunif(ix[ct],ixe[ct])
if(!vq.v[0].contains(i)){
j+=1
vkid.append(i)
}
i+=1
}
return killcnt
}
//* read .net file
strdef netfile
{sp = new NQS() cp = new NQS()}
//* CREATE CELLS
// %con (con/pre) = %div (div/post)
n=ty=id=0
//* sprcells() sprout cells in specific pathways using sprmat, $1=seed for rand generator
//max div is still 0.75*poty
func sprcells () { local id,a,prty,poty,sz,ls,mx,i localobj vid,vnewid,vnewdel,rd,vd,vtmp
ls=$1 a=allocvecs(vid,vnewid,vnewdel,vd,vtmp) rd=new Random() rd.ACG(ls)
for prty=0,CTYPi-1 for poty=0,CTYPi-1 if(sprmat[prty][poty]) for id=ix[prty],ixe[prty] {
ce.o(id).getdvi(vid) ce.o(id).getdvi(0.2,vd)
sz=div[prty][poty][0]*sprmat[prty][poty] mx=0.75*numc[poty]
if(vd.x(poty)>=mx)continue//already @ max size
while(sz+vd.x(poty)>mx) sz-=1
vrsz(sz*4,vtmp,vnewid) rd.discunif(ix[poty],ixe[poty]) vtmp.setrand(rd)
vtmp.uniq(vnewid) vtmp.resize(0)
for i=0,vnewid.size-1 if(!vid.contains(vnewid.x(i))) vtmp.append(vnewid.x(i))
vtmp.resize(sz)
if(vtmp.size) {
vnewdel.resize(vtmp.size)
rd.uniform(delm[prty][poty]-deld[prty][poty],delm[prty][poty]+deld[prty][poty])
vnewdel.setrand(rd)
ce.o(id).setdvi(vtmp,vnewdel,2)
}
}
dealloc(a)
return 1
}
//** gethublims(col,hubtype,hubfactor,numhubs,mode)
// get a matrix of size CTYPi X CTYPi, specifying div with mat.x(hubtype,othertype)
// and conv with mat.x(othertype,hubtype)
// hubtype = type of hub. hubfactor = desired ratio of hub div/conv vs non-hub div/conv
// numhubs = # of hubs. col = COLUMN for which to set hubs.
// mode == 0 <-- hub div(conv) is set to hubfactor * original div(conv)
// mode == 1 <-- hub div(conv) is set so that final hub div = hubfactor * final non_hub div (same for conv)
// formula is based on: m / ((N-H*m) / (C-H)) = F , and then solving for m
// m = div for the hubs, F = desired ratio of final hub div to final non-hub div
// N = # of synapses (links), C = total # of postsynaptic cells (including hubs) , H = # of hubs
// similarly done for conv , but replace N with appropriate values
// (/u/samn/intfcol/notebook.dol_1:21933)
obfunc gethublims () { local ct,mode,from,to,lim,nc,nhubs,fctr localobj col,mat
{col=$o1 ct=$2 fctr=$3 nhubs=$4 mode=$5 mat=new Matrix(CTYPi,CTYPi)}
for to=0,CTYPi-1 if(col.numc[to] && col.div[ct][to]) {
{nc=col.numc[to] if(ct==to)nc-=1} // deduct for self-link
if(mode==0) {
lim = int( 0.5 + col.div[ct][to]*fctr )
} else {
lim = int( 0.5 + col.div[ct][to]*col.numc[ct]*fctr/(col.numc[ct]-nhubs+fctr*nhubs) )
}
mat.x(ct,to) = MINxy(lim, nc) // at most div to all postsynaptic cells
}
for from=0,CTYPi-1 if(col.numc[from] && col.div[from][ct]) {
{nc=col.numc[from] if(ct==from)nc-=1} // deduct for self-link
if(mode==0) {
lim = int( 0.5 + col.conv[from][ct]*fctr )
} else {
lim = int( 0.5 + col.div[from][ct]*col.numc[from]*fctr/(col.numc[ct]-nhubs+fctr*nhubs) )
}
mat.x(from,ct) = MAXxy(MINxy(lim, nc),1) // at most conv from all presynaptic cells, but at least 1
}
return mat
}
//** addhubs(column,cell-type,numhubs,scaling factor,skipI[,seed,allowz,hubmode,verbose])
// add hubs to the network by stealing wires from other neurons
// $o1 == column
// $2 == cell type of hub
// $3 == number of hubs to add
// $4 == scaling factor (should be > 1.0) for conv,div of hub
// $5 == skip div/conv of I cells
// $6 == seed - optional
// $7 == allowz - whether to allow pulling all links from/to another cell
// $8 == hubmode - which mode to use for gethublims (see above)
// $9 == verbose - optional
// function returns a Vector containing the ids of the cells selected as hubs (within column ids)
obfunc addhubs () { local a,ct,fctr,nhubs,idx,jdx,lseed,hubid,szorig,cursz,preid,poid,lim,skipI,to,from,vrb,changed,allowz,hmode\
localobj col,ce,vin,vout,nq,vd,vc,vdd,vdt,vddt,vpicked,vhubid,vw1,vw2,vsyn,vprob,vsynt,vtmp,vdsz,vcsz,mhlim
col=$o1 ct=$2 nhubs=$3 fctr=$4 skipI=$5
if(numarg()>5) lseed=$6 else lseed=1234
if(numarg()>6) allowz=$7 else allowz=1
if(numarg()>7) hmode=$8 else hmode=0
if(numarg()>8) vrb=$9 else vrb=0
{ce=col.ce hashseed_stats(lseed,lseed,lseed)}
a=allocvecs(vin,vout,vd,vc,vdd,vdt,vddt,vpicked,vw1,vw2,vsyn,vprob,vsynt,vtmp,vdsz,vcsz)
vrsz(col.allcells,vin,vout,vd,vc,vdd,vdt,vddt,vpicked,vw1,vw2,vsyn,vprob,vsynt,vdsz,vcsz,vtmp)
mhlim=gethublims(col,ct,fctr,nhubs,hmode)
//vin,vout = input/output markers. vd,vc = div/conv.
//vdd div/conv delays, vdt div/conv temp. vddt=div/conv delay temp
//vpicked=which cells already picked as hubs
vhubid=new Vector()
{vhubid.indgen(col.ix[ct],col.ixe[ct],1) vhubid.shuffle() vhubid.resize(nhubs)}
if(vrb) vlk(vhubid)
for idx=0,vhubid.size-1 vpicked.x(vhubid.x(idx))=1
for idx=0,vhubid.size-1 { hubid=vhubid.x(idx)
if(vrb) printf("hub%d id = %d\n",idx+1,hubid)
{ce.o(hubid).getdvi(vd,vdd,vw1,vw2,vprob,vsyn) ce.o(hubid).getconv(vc)}//IDs of post/presynaptic cells
{ce.o(hubid).getconv(1.2,vcsz) vdsz.resize(CTYPi) vdsz.fill(0)}//counts of post/pre types
for jdx=0,vd.size-1 vdsz.x(ce.o(vd.x(jdx)).type)+=1
{vout.fill(0) vin.fill(0)} //init as 0
for jdx=0,vd.size-1 vout.x(vd.x(jdx))=1 //mark current postsynaptic cells
for jdx=0,vc.size-1 vin.x(vc.x(jdx))=1 //mark current presynaptic cells
for to=0,CTYPi-1 if(col.numc[to] && col.div[ct][to] && (!skipI || !ice(to))) {
cursz=szorig=vdsz.x(to) // update divergence
if(vrb) print "\torig div -> " , CTYP.o(to).s, " = " , szorig
{lim=mhlim.x(ct,to) changed=1}
while(cursz<lim && changed==1) { changed=0
for(preid=col.ix[ct];preid<=col.ixe[ct] && cursz<lim;preid+=1) {// pick same presynaptic type
if(vpicked.x(preid)) continue //dont take from other hubs
ce.o(preid).getdvi(vdt,vddt,vw1,vw2,vprob,vsynt)
vtmp.fill(0)
for jdx=0,vdt.size-1 vtmp.x(ce.o(vdt.x(jdx)).type)+=1
if(!allowz && vtmp.x(to)<=1)continue//dont want to turn div of another cell to 0
for jdx=0,vdt.size-1 { poid=vdt.x(jdx) // go thru postsynaptic cells looking for target type
if(ce.o(poid).type==to && poid!=hubid && vout.x(poid)==0) { cursz+=1
{vd.append(poid) vdd.append(vddt.x(jdx)) vsyn.append(vsynt.x(jdx))}
{vdt.remove(jdx) vddt.remove(jdx) vsynt.remove(jdx)}
ce.o(preid).setdvi(vdt,vddt,vsynt) // update presynaptic cell
vout.x(poid)=changed=1 // this cell synapses on poid
break
}
}
}
}
if(vrb) print "\tnew div -> " , CTYP.o(to).s, " = " , cursz
}
ce.o(hubid).setdvi(vd,vdd,vsyn) // update hub dvi
for from=0,CTYPi-1 if(col.numc[from] && col.div[from][ct] && (!skipI || !ice(from))) {
cursz=szorig=vcsz.x(from) // update convergence
{lim=mhlim.x(from,ct) changed=1}
if(vrb) print "\torig conv <- ", CTYP.o(from).s, " = " , szorig
while(cursz<lim && changed==1) { changed=0
for(preid=col.ix[from];preid<=col.ixe[from]&&cursz<lim;preid+=1) {
if(preid==hubid || vin.x(preid)) continue // don't make self or double-connects
ce.o(preid).getdvi(vdt,vddt,vw1,vw2,vprob,vsynt)
for jdx=0,vdt.size-1{
poid = vdt.x(jdx)
if(vpicked.x(poid)) continue // don't take wires from other hubs
if(ce.o(poid).type==ct){ ce.o(poid).getconv(1.2,vtmp)
if(allowz || vtmp.x(from)>1) { // make sure not to remove all inputs of a type to a cell
vdt.x( jdx ) = hubid // reassign input to hub
ce.o(preid).setdvi(vdt,vddt,vsynt) // reset presynaptic cell's div
vin.x( preid ) = changed = 1 // mark input
cursz += 1
break
}
}
}
}
}
if(vrb) print "\tnew conv <- " , CTYP.o(from).s, " = " , cursz
}
}
{dealloc(a) return vhubid}
}
//* mkcolnqs - make an nqs with current pmat,wmat,delm,deld info for use by a COLUMN for wiring
// "dist" represents distance between columns: dist==0 for intra-COLUMN setup, dist>0 for INTER-COLUMN setup
proc mkcolnqs () { local from,to,sy,idx,d,c,fctr localobj froms,tos,sys
if(numarg()>0) fctr=$1 else fctr=1 // -- for adjusting pmat
idx=0 // For setting column -- ignore for now
// Adjust wmat if required
if (fctr!=1) for from=0,CTYPi-1 for to=0,CTYPi-1 for sy=0,STYPi-1 for c=0,colr wmat[from][to][sy][c] *= fctr
{nqsdel(colnq[idx]) colnq[idx]=new NQS("froms","tos","sys","from","to","sy","w","pij","delm","deld","loc","dist")}
colnq[idx].strdec("froms","tos","sys")
for from=0,CTYPi-1 { froms=CTYP.o(from)
for to=0,CTYPi-1 { tos=CTYP.o(to)
for d=0,colr if(pmat[from][to][d]>0) for sy=0,STYPi-1 if(wmat[from][to][sy][d]>0) { sys=STYP.o(sy)
colnq[idx].append(froms.s,tos.s,sys.s,from,to,sy,wmat[from][to][sy][d],pmat[from][to][d],delm[from][to],deld[from][to],synloc[from][to],d)
}
}
}
}
//* setcellpos(col,vector of z values by type[,z variance,pseed,columndiameter in microns])
proc setcellpos () { local i,z,x,y,zvar,c,ctyp,zmin,zmax localobj rdm,vz
col=$o1
vz=$o2
if(numarg()>1) zvar=$3
if(numarg()>2) pseed=$4
if(numarg()>3) cside=$5
{rdm=new Random() rdm.ACG(pseed)}
c=-1
if(col.cellsnq!=nil) c=col.cellsnq.fi("xloc")
for i=0,col.allcells-1 {
ctyp = col.ce.o(i).type
// If L2/3 cell...
if ((ctyp == E2) || (ctyp == I2) || (ctyp == I2L)) {
zmin = 143.0 // L2/3 upper bound (microns)
zmax = 451.8 // L2/3 lower bound (microns)
// Else, if L5 corticostriatal cell (goes in 5A or 5B)...
} else if (ctyp == E5R) {
zmin = 451.8 // L5A upper bound (microns)
zmax = 1059.0 // L5B lower bound (microns)
// Else, if L5 corticospinal cell (goes in 5B)...
} else if (ctyp == E5B) {
zmin = 663.6 // L5B upper bound (microns)
zmax = 1059.0 // L5B lower bound (microns)
// Else, if L5 interneuron (goes in 5A or 5B)...
} else if ((ctyp == I5) || (ctyp == I5L)) {
zmin = 451.8 // L5A upper bound (microns)
zmax = 1059.0 // L5B lower bound (microns)
// If L6 cell...
} else if ((ctyp == E6) || (ctyp == I6) || (ctyp == I6L)) {
zmin = 1059.0 // L6 upper bound (microns)
zmax = 1412.0 // L6 lower bound, white-matter (microns)
}
col.ce.o(i).xloc=x=rdm.uniform(0,cside)
col.ce.o(i).yloc=y=rdm.uniform(0,cside)
col.ce.o(i).zloc=z=rdm.uniform(zmin,zmax)
if(c!=-1) {
col.cellsnq.v[c+0].x(i)=x
col.cellsnq.v[c+1].x(i)=y
col.cellsnq.v[c+2].x(i)=z
}
}
}
//* reweightL2toL5() - reweight the E2->layer 5 connections
proc reweightL2toL5 () { local a,L2toL5_EE_wmat_scale,L2toL5_EI_wmat_scale, ii,jj,prid,poid,prty,poty,ic1,ic2,wtchanged \
localobj col,ce,vidx,vdel,vwt1,vwt2,ign1,ign2,opr,opo,st,rdm
// Set the scaling parameters to scale the data from Figure 3F of the (Apicella et al., 2012).
L2toL5_EE_wmat_scale = 0.1 / EEGain // 0.023
L2toL5_EI_wmat_scale = 0.1 / EIGain // 0.0048
// Allocate vectors.
a = allocvecs(vidx,vdel,vwt1,vwt2,ign1,ign2)
// Read the function arguments.
col = $o1
ce = col.ce
// If we have a connections NQS table, set its verbosity off.
if (col.mknetnqss) col.connsnq.verbose = 0
// Loop over all cells in the column...
for ii=0,ce.count-1 {
// Set this as the pre-synaptic cell.
opr = ce.o(ii)
// Get pre-synaptic values.
prid = opr.id
prty = opr.type
ic1 = ice(opr.type)
// Initialize the weight to not changed.
wtchanged = 0
// If the pre-synaptic type is E2...
if (prty == E2) {
// Get the divergence information for the pre-synaptic cell.
opr.getdvi(vidx,vdel,ign1,vwt1,vwt2,ign2)
// Loop over all of the post-synaptic cells...
for jj=0,vidx.size-1 {
// Get post-synaptic values.
opo = ce.o(vidx.x(jj))
poid = opo.id
poty = opo.type
ic2 = ice(opo.type)
// Deal with the E2->E5R (corticostriatal) case.
if (poty == E5R) {
// If the E5R cell is in layer 5A...
if (opo.zloc < 663.6) {
if ((opr.zloc >= 143.0) && (opr.zloc < 250.0)) {
vwt1.x(jj) = 25.6 * L2toL5_EE_wmat_scale * EEGain
} else if ((opr.zloc >= 250.0) && (opr.zloc < 350.0)) {
vwt1.x(jj) = 14.2 * L2toL5_EE_wmat_scale * EEGain
} else if ((opr.zloc >= 350.0) && (opr.zloc < 451.8)) {
vwt1.x(jj) = 6.8 * L2toL5_EE_wmat_scale * EEGain
}
// Else (if layer 5B), the weight is zero.
} else {
vwt1.x(jj) = 0.0
}
// Set the secondary weight (NMDA) to 0.1 * AMPA weight.
vwt2.x(jj) = vwt1.x(jj) * 0.1
// Set the weight to changed.
wtchanged = 1
}
// Deal with the E2->E5B (corticospinal) case.
if (poty == E5B) {
if ((opr.zloc >= 143.0) && (opr.zloc < 250.0)) {
vwt1.x(jj) = 32.6 * L2toL5_EE_wmat_scale * EEGain
} else if ((opr.zloc >= 250.0) && (opr.zloc < 350.0)) {
vwt1.x(jj) = 33.7 * L2toL5_EE_wmat_scale * EEGain
} else if ((opr.zloc >= 350.0) && (opr.zloc < 451.8)) {
vwt1.x(jj) = 21.2 * L2toL5_EE_wmat_scale * EEGain
}
// Set the secondary weight (NMDA) to 0.1 * AMPA weight.
vwt2.x(jj) = vwt1.x(jj) * 0.1
// Set the weight to changed.
wtchanged = 1
}
// Deal with the E2->I5L (layer 5A and 5B LTS) cases.
if (poty == I5L) {
if ((opr.zloc >= 143.0) && (opr.zloc < 250.0)) {
// If the I5L cell is in layer 5A...
if (opo.zloc < 663.6) {
vwt1.x(jj) = 40.2 * L2toL5_EI_wmat_scale * EIGain
// Else, if layer 5B...
} else {
vwt1.x(jj) = 40.2 * L2toL5_EI_wmat_scale * EIGain
}
} else if ((opr.zloc >= 250.0) && (opr.zloc < 350.0)) {
// If the I5L cell is in layer 5A...
if (opo.zloc < 663.6) {
vwt1.x(jj) = 36.3 * L2toL5_EI_wmat_scale * EIGain
// Else, if layer 5B...
} else {
vwt1.x(jj) = 14.2 * L2toL5_EI_wmat_scale * EIGain
}
} else if ((opr.zloc >= 350.0) && (opr.zloc < 451.8)) {
// If the I5L cell is in layer 5A...
if (opo.zloc < 663.6) {
vwt1.x(jj) = 18.7 * L2toL5_EI_wmat_scale * EIGain
// Else, if layer 5B...
} else {
vwt1.x(jj) = 5.3 * L2toL5_EI_wmat_scale * EIGain
}
}
// Set the secondary weight (NMDA) to 0.1 * AMPA weight.
vwt2.x(jj) = vwt1.x(jj) * 0.1
// Set the weight to changed.
wtchanged = 1
}
// If the weight has changed and we have a connectivity NQS...
if ((wtchanged) && (col.mknetnqss)) {
// Change the weights to the new values.
{col.connsnq.select("id1",prid,"id2",poid)}
{col.connsnq.fill("wt1",vwt1.x(jj))}
{col.connsnq.fill("wt2",vwt2.x(jj))}
{col.connsnq.delect()}
}
} // for loop over all post-synaptic cells
// Reset the divergent weights.
opr.setsywv(vwt1,vwt2)
} // if pre-synaptic type E2
} // for loop over all cells
// If we have a connections NQS table, set its verbosity back.
if (col.mknetnqss) col.connsnq.verbose = 1
// Deallocate vectors.
dealloc(a)
}
//** selectconns () - select conns from connsnq based on pre- and post-synaptic type filter
proc selectconns () { local a,ii,prty,poty,prtarg,potarg localobj ridx,v1,thecol
thecol = $o1
// Allocate temporary vectors.
a = allocvecs(ridx,v1)
prtarg = $2
potarg = $3
// Start with an empty index list.
ridx.resize(0)
// Loop over all rows of connsnq
for ii=0,thecol.connsnq.size-1 {
v1 = thecol.connsnq.getrow(ii)
prty = thecol.ce.o(v1.x(0)).type
poty = thecol.ce.o(v1.x(1)).type
// Include only the rows that match the valid cases.
if ((prty == prtarg) && (poty == potarg)) {
ridx.append(ii)
}
}
// Select the rows with for the ridx rows.
thecol.connsnq.select("IND_",EQW,ridx)
// Deallocate the temporary vectors.
dealloc(a)
}
// <-- stccol
//** swirecut (col,EEsq,EIsq,IEsq,IIsq[,seed,slambda]) - spatial wiring: wires the column using pmat and cell positions
// (wiring probability effected by distance btwn cells)
// seed is random # seed
// lambda specifies length-constant for spatially-dependent fall-off in wiring probability
// at one lambda away, probability of connect is value in pmat
// CK: WARNING, this function has been kludgily rewritten to have no distance of constant connectivity!
proc swirecut() { local x,y,z,ii,jj,a,del,prid,poid,prty,poty,dv,lseed,h,prob,slambda,dsq,dist,slambdasq,\
EEsq,EIsq,IEsq,IIsq,ic1,ic2,pdx\
localobj col,ce,vidx,vdel,vdist,vwt1,vwt2,vtmp,opr,opo,st,rdm,vprob
// Allocate vectors.
a=allocvecs(vidx,vdel,vtmp,vdist,vwt1,vwt2,vprob)
z=0
// Read the function arguments.
col=$o1 ce=col.ce EEsq=$2 EIsq=$3 IEsq=$4 IIsq=$5
if (argtype(6)==0) lseed=$6 else lseed=1234
if(numarg()>6) slambda=$7 else slambda=10
if(slambda<=0){
printf("swirecut WARN: invalid slambda=%g, setting slambda to %g\n",colside/3)
slambda=colside/3
}
slambdasq = slambda^2 // using squared distance
// Set up connectivity vectors.
vrsz(1e4,vidx,vdel,vdist,vtmp)
// Set up random number stuff.
hashseed_stats(lseed,lseed,lseed)
rdm=new Random() rdm.ACG(lseed) // initdivrnd(lseed)//initialize random # generator
rdm.uniform(0,1) vprob.resize(ce.count^2) vprob.setrand(rdm) pdx=0
// Create the connectivity NQS table.
if(col.mknetnqss) {nqsdel(col.connsnq) col.connsnq=new NQS("id1","id2","del","dist","wt1","wt2")}
// Loop over all cells in the column as pre-synaptic cells...
for y=0,ce.count-1 { opr=ce.o(y)
// Start with empty connection vectors.
vrsz(0,vidx,vdel,vdist,vwt1,vwt2)
// Get pre-synaptic values.
prid=opr.id prty=opr.type ic1=ice(opr.type)
// For all post-synaptic types where there are cells there that need connecting to
// (set h to pmat value for that pr->po)...
for poty=0,CTYPi-1 if (col.numc[poty]!=0 && (h=col.pmat[prty][poty])>0) {
// For all of the post-synaptic cells (assuming not same as pre-synaptic cell)...
for poid=col.ix[poty],col.ixe[poty] if(prid!=poid) {
// Get post-synaptic cell values.
opo = ce.o(poid) ic2=ice(ce.o(poid).type)
// Get the x,y, and z distances.
dx = opr.xloc - opo.xloc
dy = opr.yloc - opo.yloc
dz = opr.zloc - opo.zloc
// Get the square of the x,y distance.
dsq = dx^2 + dy^2 // Connectivity fall-off only depends in intra-layer distance
// Get the Euclidean x,y,z distance.
ds = sqrt(dx^2 + dy^2 + dz^2) // Axonal delay depends on all quantities
// Get probability of connection (pmat value at x,y distance 0, scaled * 0.368 when slambda = dist)
prob = h * exp(-sqrt(dsq)/slambda) // probability of connect
// If the random connection is to be made...
if( prob >= vprob.x(pdx) ) { // rdm.uniform(0,1)
// Set the axonal delay using the x,y,z distance and random values.
del = delmscalemod*(rdm.uniform(col.delm[prty][poty]-col.deld[prty][poty],col.delm[prty][poty]+col.deld[prty][poty])+ds/axonalvelocity) // Include both synaptic and axonal delays (ds/axonalvelocity is axonal)
// Add the post-synaptic cell ID and the delay.
{vidx.append(poid) vdel.append(del)}
// Add the distance value (dendrite vs. soma).
if(synloc[prty][poty]==DEND) vdist.append(1) else vdist.append(0)
// Add primary and secondary weights, if appropriate to do so.
if(col.mknetnqss || wsetting_INTF6==1) {
if(col.syty1[prty][poty]>=0) vwt1.append(col.wmat[prty][poty][col.syty1[prty][poty]]) else vwt1.append(0)
if(col.syty2[prty][poty]>=0) vwt2.append(col.wmat[prty][poty][col.syty2[prty][poty]]) else vwt2.append(0)
}
}
pdx += 1
} // for poid
} // for poty
// If there is more than one efferent connection...
if(vidx.size>0) {
// If we are working with individual synapses (vwt1,vwt2), use them. Otherwise, set up normally.
if(wsetting_INTF6==1) opr.setdvi(vidx,vdel,vdist,0,vwt1,vwt2) else opr.setdvi(vidx,vdel,vdist)
// If we are making NQSs, add the new connections to connsnq.
if(col.mknetnqss) {
for ii=0,vidx.size-1 col.connsnq.append(prid,vidx.x(ii),vdel.x(ii),vdist.x(ii),vwt1.x(ii),vwt2.x(ii))
}
}
} // for all (pre-synaptic) cells
// Deallocate vectors.
dealloc(a)
if(verbose) printf("\n")
}
//* checkerpos - arrange cells in checkerboard pattern
proc checkerpos () { local i,j,x,y,xn,yn,csz,ssz,incsz,mnx,mxx,mny,mxy,a localobj col,ce,rdm,xo,vox,vex,voy,vey
col=$o1 csz=col.cdiam ssz=$2 incsz=$3 rdm=new Random() rdm.ACG($4)
ce=col.ce
a=allocvecs(vox,vex,voy,vey)
mxx=csz
if(mxx%2==1) mxx-=1
vex.indgen(0,mxx,2)
vey.indgen(0,mxx,2)
mxx=csz
if(mxx%2==0) mxx-=1
vox.indgen(1,mxx,2)
voy.indgen(1,mxx,2)
for ltr(xo,ce,&i) {
if(rdm.discunif(0,1)==0) {
x = vex.x(rdm.discunif(0,vex.size-1))
y = vey.x(rdm.discunif(0,vey.size-1))
} else {
x = vox.x(rdm.discunif(0,vox.size-1))
y = voy.x(rdm.discunif(0,voy.size-1))
}
mnx=MAXxy(x-1,0)
mxx=MINxy(x,csz)
if(mnx==mxx && mnx==0) {
mnx+=1 mxx+=2
}
mny=MAXxy(y-1,0)
mxy=MINxy(y,csz)
if(mny==mxy && mny==0) {
mny+=1 mxy+=2
}
xn = rdm.uniform(mnx,mxx)
yn = rdm.uniform(mny,mxy)
xo.xloc=xn
xo.yloc=yn
if(col.cellsnq!=nil) {
col.cellsnq.v[5].x(i) = xo.xloc
col.cellsnq.v[6].x(i) = xo.yloc
}
}
}
// stccol -->
//* mkcols - make the COLUMNs
proc mkcols () { local id,x,y,seed localobj nq
id=0
for y=0,colH-1 for x=0,colW-1 {
if(dbgcols)seed=dvseed else seed=(id+1)*dvseed
lcol.append(gcol[y][x]=new COLUMN(id,vcpercol,colnq,seed,x,y,setdviPT,mknetnqss,1))
col[id]=gcol[y][x]
col[id].verbose=verbose_INTF6
// <-- stccol
if(strlen(wnqstr)>0) { // wire from NQS
nq=new NQS(wnqstr)
col[id].wirenq(nq)
nqsdel(nq)
} else if(swire>0) { // spatial-wiring
// col[id].setcellpos(vlayerz,layerzvar,pseed*(id+1),colside)
setcellpos(col[id],vlayerz,layerzvar,pseed*(id+1),colside) // use Gordon cell version
if(checkers) checkerpos(col[id],1,1,pseed*(id+1))
if(swire==1) {
col[id].swire(col[id].wseed,maxsfall)
} else if(swire==2) {
swirecut(col[id],EEsq,EIsq,IEsq,IIsq,col[id].wseed,slambda) // Don't read from disk, calculate
}
} else {
setcellpos(col[id],vlayerz,layerzvar,pseed*(id+1),colside) // use Gordon cell version
col[id].wire(col[id].wseed) // random wiring (no spatial dependence)
reweightL2toL5(col[id])
}
// stccol -->
id+=1
}
}
//* wirecols - setup inter-COLUMN connectivity with NetCon
proc wirecols () { local x1,y1,x2,y2,dx,dy,maxd,d localobj fromc,toc
if(numarg()>0) d=$1 else d=colr
if(torus) { // wraparound
//alternate coordinates: ( -colW+x , -colH+y )
//alternate system: -5 -4 -3 -2 -1
//original system: 0 1 2 3 4
//layed out as a line: -5 -4 -3 -2 -1 0 1 2 3 4
//only need to compare in normal system, and 1 alternate coordinate vs original (and vice versa)
for y1=0,colH-1 for x1=0,colW-1 for y2=0,colH-1 for x2=0,colW-1 {
if(y1==y2 && x1==x2) continue // skip self-self
dx=MINxy(abs(x1-x2), MINxy(abs((-colW+x1)-x2), abs(x1-(-colW+x2))) )
dy=MINxy(abs(y1-y2), MINxy(abs((-colH+y1)-y2), abs(y1-(-colH+y2))) )
if((maxd=MAXxy(dx,dy)) > d) continue // skip too far
gcol[y1][x1].wire2col(gcol[y2][x2],colnq,maxd,ncl) // unidirectional wiring
}
} else { // no wrap-around
for y1=0,colH-1 for x1=0,colW-1 for y2=0,colH-1 for x2=0,colW-1 {
if(y1==y2 && x1==x2) continue // skip self-self
if((maxd=MAXxy(abs(x1-x2),abs(y1-y2))) > d) continue // skip too far
gcol[y1][x1].wire2col(gcol[y2][x2],colnq,maxd,ncl) // unidirectional wiring
}
}
}
//* intercoloff - turn off all weights between COLUMNs
proc intercoloff () { local i localobj xo
for ltr(xo,ncl) if(isojt(xo.pre,col.ce.o(0)) && isojt(xo.syn,col.ce.o(0))) {
for i=0,6 xo.weight(i)=0
}
}
// <-- stccol
//* intercolmul(from,to,sy,w)
proc intercolsyw () { local from,to,sy,w localobj xo
from=$1 to=$2 sy=$3 w=$4
for ltr(xo,ncl) if(isojt(xo.pre,col.ce.o(0)) && isojt(xo.syn,col.ce.o(0))) {
if(xo.pre.type==from && xo.syn.type==to) xo.weight(sy)=w
}
}
// stccol -->
//* function calls to setup network
//** z location of each layer in microns
mkvlayerz()
//** # of cells per column
setcpercol() //new numbers (10aug30)
//** setup pmat
if(name_declared("nqpmat")==2) { // read pmat from NQS if available, else set to default
if(nqpmat!=nil) nq2pmat(nqpmat) else setpmat()
} else setpmat()
if(pmatscale!=1) scalepmat(pmatscale)
//** setup synapse locations,delays,wmat
setsynloc()
setdelmats()
setwmat() // new KMJ version
scrsz=50*1e3
double scr[scrsz]
//** make cells, columns, wire columns
print "Making cells..."
mkcolnqs(wmatscale) // Argument is the scale factor for the weight matrix
print "Making columns..."
mkcols()
if (loadsaveconnectivity==2) { // Save connectivity matrix
print "Saving connectivity NQS table..."
col.connsnq.sv("connectivity.nqs") // Save connectivity NQS to disk
print "...done...."
}
print "Wiring columns..."
wirecols(1)
print "...done..."
// <-- stccol
//** leftover code from tvis - may use at some point
//*** delm setup -- mean delays
proc ae () {
//**** horizontal intralaminar mean delays
delm[E2][E2] = 0.4
delm[E2][I2] = 0.4
delm[E5R][E5R] = 0.4
delm[E5R][I5] = 0.4
delm[E6][E6] = 0.4
delm[E6][I6] = 0.4
//**** vertical interlaminar mean delays
delm[E2][E5R] = 1.4
delm[E2][I5] = 1.4
delm[E2][E6] = 1.8
delm[E2][I6] = 1.8
delm[E5R][E2] = 1.4
delm[E5R][I2] = 1.4
delm[E5R][E6] = 1.25
delm[E5R][I6] = 1.25
delm[E6][E5R] = 1.85
delm[E6][I5] = 1.85
//**** intracortical inhibitory mean delays
delm[I2][E2] = 0.4
delm[I2][I2] = 0.4
delm[I5][E5R] = 0.4
delm[I5][I5] = 0.4
delm[I6][E6] = 0.4
delm[I6][I6] = 0.4
//
}
//*** dels setup -- stdev of delays
proc ae () {
//**** horizontal intralaminar stdev of delay
dels[E2][E2] = 0.1
dels[E2][I2] = 0.1
dels[E5R][E5R] = 0.1
dels[E5R][E5R] = 0.1
dels[E5R][I5] = 0.1
dels[E6][E6] = 0.1
dels[E6][I6] = 0.1
//**** vertical interlaminar stdev of delay
dels[E2][E5R] = 0.25
dels[E2][I5] = 0.25
dels[E2][E6] = 0.25
dels[E2][I6] = 0.25
dels[E5R][E2] = 0.25
dels[E5R][I2] = 0.25
dels[E5R][E6] = 0.25
dels[E5R][I6] = 0.25
dels[E6][E5R] = 0.25
dels[E6][I5] = 0.25
//**** intracortical inhibitory stdev of delay
dels[I2][E2] = 0.1
dels[I2][I2] = 0.1
dels[I5][E5R] = 0.1
dels[I5][I5] = 0.1
dels[I6][E6] = 0.1
dels[I6][I6] = 0.1
//
from=to=0 //set delay variance
for ctt(&from) for ctt(&to) if(dels[from][to]) delv[from][to]=dels[from][to]^2
}
// stccol -->
