Olfactory bulb microcircuits model with dual-layer inhibition (Gilra & Bhalla 2015)

 Download zip file 
Help downloading and running models
Accession:153574
A detailed network model of the dual-layer dendro-dendritic inhibitory microcircuits in the rat olfactory bulb comprising compartmental mitral, granule and PG cells developed by Aditya Gilra, Upinder S. Bhalla (2015). All cell morphologies and network connections are in NeuroML v1.8.0. PG and granule cell channels and synapses are also in NeuroML v1.8.0. Mitral cell channels and synapses are in native python.
Reference:
1 . Gilra A, Bhalla US (2015) Bulbar microcircuit model predicts connectivity and roles of interneurons in odor coding. PLoS One 10:e0098045 [PubMed]
Model Information (Click on a link to find other models with that property)
Model Type: Realistic Network;
Brain Region(s)/Organism: Olfactory bulb;
Cell Type(s): Olfactory bulb main mitral GLU cell; Olfactory bulb main interneuron periglomerular GABA cell; Olfactory bulb main interneuron granule MC GABA cell;
Channel(s): I A; I h; I K,Ca; I Sodium; I Calcium; I Potassium;
Gap Junctions:
Receptor(s): AMPA; NMDA; Gaba;
Gene(s):
Transmitter(s): Gaba; Glutamate;
Simulation Environment: Python; MOOSE/PyMOOSE;
Model Concept(s): Sensory processing; Sensory coding; Markov-type model; Olfaction;
Implementer(s): Bhalla, Upinder S [bhalla at ncbs.res.in]; Gilra, Aditya [aditya_gilra -at- yahoo -period- com];
Search NeuronDB for information about:  Olfactory bulb main mitral GLU cell; Olfactory bulb main interneuron periglomerular GABA cell; Olfactory bulb main interneuron granule MC GABA cell; AMPA; NMDA; Gaba; I A; I h; I K,Ca; I Sodium; I Calcium; I Potassium; Gaba; Glutamate;
import sys
sys.path.extend(["..","../networks","../generators","../simulations"])

from moose_utils import * # imports moose, and 'moose.utils import *'
from networkConstants import *
from stimuliConstants import *
from data_utils import *

def setupTables(network, nopgs, nosingles, nojoints, nomultis, args={}, spikes=False):
    pgtables = []
    singletables = []
    jointtables = []
    multitables = []
    ## if spikes == True i.e. only spiketimes are recorded,
    ## then record all cells, else less number of cells.
    allcells = spikes
    if not nopgs:
        ## Setup the tables to pull data from PGs
        ## get cells connected to mitrals specified in args. if not specified get arbitrary.
        PGlist = getCellsByMitralConnection(args, network, 'PG_mitral', 'PGs', allcells)
        for i,PG in enumerate(PGlist):
            # assumes at least one soma and takes the first!
            PG.soma = moose.Compartment(get_matching_children(PG, ['Soma','soma'])[0])
            PG._vmTableSoma = setupTable("vmTableSoma",PG.soma,'Vm')
            pgtables.append(PG._vmTableSoma)
            if spikes:
                PG._vmTableSoma.stepMode = TAB_SPIKE
                PG._vmTableSoma.stepSize = THRESHOLD
    if not nosingles:
        ## Setup the tables to pull data from single granules
        ## get cells connected to mitrals specified in args. if not specified get arbitrary.
        singlesList = getCellsByMitralConnection(args, network,
            'granule_mitral_inh_singles', 'granules_singles', allcells)
        for i,gran in enumerate(singlesList):
            ## assumes at least one soma and takes the first!
            gran.soma = moose.Compartment(get_matching_children(gran, ['Soma','soma'])[0])
            gran._vmTableSoma = setupTable("vmTableSoma",gran.soma,'Vm')
            singletables.append(gran._vmTableSoma)
            if spikes:
                gran._vmTableSoma.stepMode = TAB_SPIKE
                gran._vmTableSoma.stepSize = THRESHOLD
    if not nojoints:
        ## Setup the tables to pull data from joint granules
        ## get cells connected to mitrals specified in args.
        ## if not specified get arbitrary.
        jointsList = getCellsByMitralConnection(args, network,
            'granule_mitral_inh_joints', 'granules_joints', allcells)
        for i,gran in enumerate(jointsList):
            ## assumes at least one soma and takes the first!
            gran.soma = moose.Compartment(get_matching_children(gran, ['Soma','soma'])[0])
            gran._vmTableSoma = setupTable("vmTableSoma",gran.soma,'Vm')
            jointtables.append(gran._vmTableSoma)
            if spikes:
                gran._vmTableSoma.stepMode = TAB_SPIKE
                gran._vmTableSoma.stepSize = THRESHOLD
    if not nomultis:
        ## Setup the tables to pull data from multi granules
        ## get cells connected to mitrals specified in args.
        ## if not specified get arbitrary.
        multisList = getCellsByMitralConnection(args, network,
            'granule_mitral_inh_multis', 'granules_multis', allcells)
        for i,gran in enumerate(multisList):
            ## assumes at least one soma and takes the first!
            gran.soma = moose.Compartment(get_matching_children(gran, ['Soma','soma'])[0])
            gran._vmTableSoma = setupTable("vmTableSoma",gran.soma,'Vm')
            multitables.append(gran._vmTableSoma)
            if spikes:
                gran._vmTableSoma.stepMode = TAB_SPIKE
                gran._vmTableSoma.stepSize = THRESHOLD
    return (pgtables, singletables, jointtables, multitables)

def exportTable(network, neuronProj, neuronPop, colours, \
            args={}, spikes=True, allcells=True):
    """ Return tables of cells in neuronPop population name
        connected to mitrals specified in args,
        via neuronProj projection name
        if args is not specified get all.
        NOTE: All tables for all cells' somas should be present if allcells is True.
        If allcells is False, cells called here should have tables present.
    """
    exportDict = {'spikes':spikes,'data_tables':[]}
    if array(colours).shape == (3,): coloursList = False
    else: coloursList = True
    ## get cells connected to mitrals specified in args. if not specified get arbitrary.    
    celllist = getCellsByMitralConnection(args, network, neuronProj, neuronPop, allcells)
    for i,cell in enumerate(celllist):
        cellname = cell.path.split('/')[-1]
        ## assumes at least one soma and takes the first!
        cell.soma = moose.Compartment(get_matching_children(cell,['Soma','soma'])[0])
        cellTablePath = cell.soma.path+"/data/vmTableSoma"
        if moose.context.exists(cellTablePath):
            cell._vmTableSoma = moose.Table(cellTablePath)
        else:
            print "SimError: Did not find "+cellTablePath
            sys.exit(1)
        if coloursList: colour=colours[i]
        else: colour=colours
        exportDict['data_tables'].append((cellname,colour,array(cell._vmTableSoma)))
    return exportDict

def exportTables(network, nopgs, nosingles, nojoints, nomultis, mitspikes, args, colours):
    ## mitrals first each with a different colour
    allTables = [ exportTable(network, '', 'mitrals', colours, \
            args=args, spikes=mitspikes, allcells=True) ]
    ## Each colour-entry below is a tuple of (baseline/initial colour, spiking/peak colour, colourmap)
    ## Each colour is a tuple of (r,g,b,a)
    ## Set in Moogli's config file, whether to change color, and/or alpha, or use colourmap.
    if not nopgs:
        allTables.append( exportTable(network,\
            'PG_mitral', 'PGs', ((0.3,0,0,0.3),(1,0,0,1),'jet'), args) )
    if not nosingles:
        allTables.append( exportTable(network,\
            'granule_mitral_inh_singles', 'granules_singles', \
            ((0.3,0,0.3,0.3),(1,0,1,1),'jet'), args) )
    if not nojoints:
        allTables.append( exportTable(network,\
            'granule_mitral_inh_joints', 'granules_joints', \
            ((0.3,0.3,0,0.3),(1,1,0,1),'jet'), args) )
    if not nomultis:
        allTables.append( exportTable(network,\
            'granule_mitral_inh_multis', 'granules_multis', \
            ((0,0.3,0.3,0.3),(0,1,1,1),'jet'), args) )
    return allTables

def getCellsByMitralConnection(args, network, projection, population, allcells=False):
    """
    Returns a list of pre-synaptic MOOSE Cell-s in 'projection' that are connected to mitrals
    specificied in args={'mitrals':[mitid1, mitid2, ...]}.
    If args does not have 'mitrals' key, an arbitrary list of cells from 'population' are returned.
    if allcells is False, a few cells are returned, else all cells are returned.
    """
    cellList = []
    cellUniques = []
    if args.has_key('mitrals'):
        for mitid in args['mitrals']:
            mitpath = 'mitrals_'+str(mitid)
            cellnum = 0
            if projection in network.projectionDict:
                for conn in network.projectionDict[projection][2]:
                    if mitpath in conn[2]: # if mitrals_<mitid> is substring of post_seg_path of this connection
                        cellpath = string.split(conn[1],'/')[1] # take out cellname from '/cellname/segmentname
                        ## Take only those cells that have not been taken before.
                        if cellpath not in cellUniques:
                            cell = moose.Cell(cellpath)
                            cellList.append(cell)
                            cellUniques.append(cellpath)
                            cellnum += 1
                            if not allcells and cellnum == 30: break
    else:
        if allcells: cellList = network.populationDict[population][1].values()
        else: cellList = network.populationDict[population][1].values()[0:60]
    return cellList

def exportMitralTables(mitMOOSETables,mitcolours,mitspikes):
    mitTables = []
    for miti,(mitname,mitCellTables) in enumerate(mitMOOSETables):
        mitTables.append((mitname,mitspikes,mitcolours[miti],\
            [(compname,array(compTable)) for compname,compTable in mitCellTables]))
    return mitTables

def setupMitralTables(network,mitspikes):
    mitTables = []
    for mit in network.populationDict['mitrals'][1].values():
        mitTables.append((mit.name,setupCellTables(mit,mitspikes)))
    return mitTables

def setupCellTables(cellObj,spikes):
    cellTables = []
    for compartmentid in cellObj.getChildren(cellObj.id): # compartments
        comp = moose.Compartment(compartmentid)
        cellTable = setupTable('VmTableExtra',comp,'Vm')
        if spikes:
            cellTable.stepMode = TAB_SPIKE
            cellTable.stepSize = THRESHOLD
        cellTables.append((comp.name,cellTable))
    return cellTables

def numpy_convert_tables(tables):
    return [[array(table) for table in onetypetables] for onetypetables in tables]

def plot_extras(timevec, tables, nopgs, nosingles, nojoints, nomultis, plottitle=''):
    if not nopgs:
        figure()
        title("PGs "+plottitle)
        for pgtable in tables[0]:
            plot(timevec[:len(pgtable)], pgtable, ',')
    if not nosingles:
        figure()
        title("single granules "+plottitle)
        for singletable in tables[1]:
            plot(timevec[:len(singletable)], singletable, ',')
    if not nojoints:
        figure()
        title("joint granules "+plottitle)
        for jointtable in tables[2]:
            plot(timevec[:len(jointtable)], jointtable, ',')
    if not nomultis:
        figure()
        title("multi granules "+plottitle)
        for multitable in tables[3]:
            plot(timevec[:len(multitable)], multitable, ',')

def plot_extras_spikes(binvec, tables, nopgs, nosingles,\
    nojoints, nomultis, bins, runt, settlet, plottitle=''):
    if not nopgs:
        figure()
        title("PGs "+plottitle)
        for pgtable in tables[0]:
            plot(binvec, plotBins(pgtable, bins, runt, settlet), '.-')
        axes_labels(gca(),"time (s)","rate (Hz)",fontsize=label_fontsize+2)
    if not nosingles:
        figure()
        title("single granules "+plottitle)
        for singletable in tables[1]:
            plot(binvec, plotBins(singletable, bins, runt, settlet), '.-')
        axes_labels(gca(),"time (s)","rate (Hz)",fontsize=label_fontsize+2)
    if not nojoints:
        figure()
        title("joint granules "+plottitle)
        for jointtable in tables[2]:
            plot(binvec, plotBins(jointtable, bins, runt, settlet), '.-')
        axes_labels(gca(),"time (s)","rate (Hz)",fontsize=label_fontsize+2)
    if not nomultis:
        figure()
        title("multi granules "+plottitle)
        for multitable in tables[3]:
            plot(binvec, plotBins(multitable, bins, runt, settlet), '.-')
        axes_labels(gca(),"time (s)","rate (Hz)",fontsize=label_fontsize+2)

def print_extras_activity(tables, nopgs, nosingles, nojoints, nomultis, contextstr):
    spikestable = {}
    if not nopgs:
        numspikes = 0
        numcells = 0
        for totcells,pgtable in enumerate(tables[0]):
            pgtable = array(pgtable)
            ## MOOSE often inserts one or two spiketime = 0.0 entries
            ## when storing spikes, so discount those
            numspikescell = len(where(pgtable>0.0)[0])
            if numspikescell>0:
                numspikes += numspikescell
                numcells += 1
        print "The number of PG cells spiking for",contextstr,"is",\
            numcells,"of",totcells+1,"; number of spikes =",numspikes
        pgstable = (numcells,totcells,numspikes)
        spikestable['PGs'] = pgstable
    if not nosingles:
        numspikes = 0
        numcells = 0
        for totcells,singletable in enumerate(tables[1]):
            singletable = array(singletable)
            ## MOOSE often inserts one or two spiketime = 0.0 entries
            ## when storing spikes, so discount those
            numspikescell = len(where(singletable>0.0)[0])
            if numspikescell>0:
                numspikes += numspikescell
                numcells += 1
        print "The number of single granule cells spiking for",contextstr,"is",\
            numcells,"of",totcells+1,"; number of spikes =",numspikes
        singlestable = (numcells,totcells,numspikes)
        spikestable['singles'] = singlestable
    if not nojoints:
        numspikes = 0
        numcells = 0
        for totcells,jointtable in enumerate(tables[2]):
            jointtable = array(jointtable)
            ## MOOSE often inserts one or two spiketime = 0.0 entries
            ## when storing spikes, so discount those
            numspikescell = len(where(jointtable>0.0)[0])
            if numspikescell>0:
                numspikes += numspikescell
                numcells += 1
        print "The number of joint granule cells spiking for",contextstr,"is",\
            numcells,"of",totcells+1,"; number of spikes =",numspikes
        jointstable = (numcells,totcells,numspikes)
        spikestable['joints'] = jointstable
    if not nomultis:
        numspikes = 0
        numcells = 0
        for totcells,multitable in enumerate(tables[3]):
            multitable = array(multitable)
            ## MOOSE often inserts one or two spiketime = 0.0 entries
            ## when storing spikes, so discount those
            numspikescell = len(where(multitable>0.0)[0])
            if numspikescell>0:
                numspikes += numspikescell
                numcells += 1
        print "The number of multi granule cells spiking for",contextstr,"is",\
            numcells,"of",totcells+1,"; number of spikes =",numspikes
        multistable = (numcells,totcells,numspikes)
        spikestable['multis'] = multistable
    return spikestable

## non-overlapping bins
def rebin_pulses(mitral_responses_list, numbins, runtime, settletime):
    numtrials = len(mitral_responses_list)
    numtrains = len(mitral_responses_list[0])
    nummits = len(mitral_responses_list[0][0])
    return [[[ plotBins( mitral_responses_list[trialnum][trainnum][mitnum],\
        numbins, runtime, settletime) \
        for mitnum in range(nummits)] \
            for trainnum in range(numtrains)] \
               for trialnum in range(numtrials)]

### overlapping bins
#def rebin_pulses(mitral_responses_list, numbins, runtime, settletime, bin_width_time):
#    numtrials = len(mitral_responses_list)
#    numtrains = len(mitral_responses_list[0])
#    nummits = len(mitral_responses_list[0][0])
#    return [[[ plotOverlappingBins( mitral_responses_list[trialnum][trainnum][mitnum],\
#        numbins, runtime, settletime, bin_width_time ) \
#        for mitnum in range(nummits)] \
#            for trainnum in range(numtrains)] \
#                for trialnum in range(numtrials)]


def rebin(mitral_responses_list, numbins, bin_width_time, numresps=1):
    numtrials = len(mitral_responses_list)
    numodors = len(mitral_responses_list[0])
    nummits = len(mitral_responses_list[0][0])
    ## NUM_RESPS is number of RESPIRATION cycles simulated, we rebin and return the last 'numresps'
    return [[[  plotOverlappingBins( mitral_responses_list[trialnum][odornum][mitnum],\
        numbins, RESPIRATION*numresps, SETTLETIME+(NUM_RESPS-numresps)*RESPIRATION, bin_width_time )
        for mitnum in range(nummits)] \
            for odornum in range(numodors)] \
                for trialnum in range(numtrials)]

def build_tweaks(mitralsclub, nospineinh, nosingles,
    nojoints, nomultis, nopgs, onlytwomits, 
    includeProjections=[], twomitrals=(0,2), nolateral=False):
    """
    Excludes singles/joints/multis granules and their projections as appropriate
    Excludes extra-excitation from unmodeled sisters as appropriate
    Include only the two required mitrals (and its connected granules) if ONLY_TWO_MITS is True
    """
    excludePopulations = []
    excludeProjections = ['SA']
    ## In odor_pulses, odor_morphs, scaled_pulses, I have not specified to include 
    ## file-based inputs to 2nd order cells as below. If not specified, force include:
    if 'granule_baseline' not in includeProjections: includeProjections.append('granule_baseline')
    if 'ORN_PG' not in includeProjections: includeProjections.append('ORN_PG')
    if not mitralsclub:
        excludeProjections.append('mitral_granule_extra_exc')
    if nospineinh:
        excludeProjections.append('_spinesingles')
        excludeProjections.append('_spinejoints')
        excludeProjections.append('_spinemultis')
    if nosingles:
        excludePopulations.append('singles')
        excludeProjections.append('_singles') # _ to avoid deleting spinesingles
    if nojoints:
        excludePopulations.append('joints')
        excludeProjections.append('_joints') # _ to avoid deleting spinejoints
    if nomultis:
        excludePopulations.append('multis')
        excludeProjections.append('_multis') # _ to avoid deleting spinemultis
    if nopgs:
        excludePopulations.append('PGs')
        excludeProjections.append('PG')
    if onlytwomits:
        onlyInclude = {'includePopulation':('mitrals',[str(twomitrals[0]),str(twomitrals[1])]),
            'includeProjections':includeProjections}
        return {'excludePopulations':excludePopulations,
            'excludeProjections':excludeProjections,'onlyInclude':onlyInclude}
    else:
        if nolateral:
            ## remove other mitrals so that there is no lateral inhibition
            ## differs from nojoints, in keeping the joints self-inhibition
            print "EXCLUDING OTHER MITS, KEEPING ONLY mits 0 and 1"
            onlyInclude = {'includePopulation':('mitrals',['0','1']),
                'includeProjections':includeProjections}
            return {'excludePopulations':excludePopulations,
                'excludeProjections':excludeProjections,'onlyInclude':onlyInclude}
        else:
            return {'excludePopulations':excludePopulations,\
                'excludeProjections':excludeProjections}

Loading data, please wait...