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]
Citations  Citation Browser
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;
#!/usr/bin/env python
# -*- coding: utf-8 -*-

import sys
import pickle
sys.path.extend(["..","../networks","../generators","../simulations","../analysis"])

from OBNetwork import *
from stimuliConstants import * # has SETTLETIME, inputList and pulseList, GLOMS_ODOR, GLOMS_NIL
from simset_odor import * # has REALRUNTIME, NUMBINS
from sim_utils import * # has rebin() and imports data_utils.py for axes_off()
from data_utils import * # has mpi import and variables also

plot_images=False#True ## plot Adil style images?
## below requires scipy which requires lapack (present only on gj not on nodes)
## hence import only if not running in parallel
if plot_images and mpisize == 1:
    from fit_odor_morphs import * # has fit_morphs()

from pylab import * # part of matplotlib that depends on numpy but not scipy

PLOTRESP_NUM = 2 # whether to plot 2 respiration cycles or 1
BINS_PER_RESP = 17
NUMBINS = BINS_PER_RESP*PLOTRESP_NUM
BIN_WIDTH_TIME = RESPIRATION/float(BINS_PER_RESP)
bindt = RESPIRATION/float(NUMBINS)
fitted_mitral = 2*central_glom+0
## I take the last PLOTRESP_NUM of respiration cycles out of NUM_RESPS simulated
startplottime = SETTLETIME+(NUM_RESPS-PLOTRESP_NUM)*RESPIRATION
endplottime = SETTLETIME+NUM_RESPS*RESPIRATION
responsetlist = arange( startplottime+bindt/2.0, endplottime, RESPIRATION/NUMBINS)

def plot_morph_responses(mitral_responses_avg,mitral_responses_se):

    numodors = len(inputList)
    if ONLY_TWO_MITS: mitlist = range(MIT_SISTERS)
    else: mitlist = range(1)#range(6)#range(NUM_GLOMS*MIT_SISTERS)
    for mitnum in mitlist:
        if mitnum%MIT_SISTERS == 0:
            figure()
            title('Glomerulus '+str(mitnum/MIT_SISTERS))
        for odornum in range(numodors):
            odorA,odorB = inputList[odornum]
            sister_ratio = (mitnum%MIT_SISTERS)/float(MIT_SISTERS)
            errorbar(x=responsetlist,y=mitral_responses_avg[odornum,mitnum],\
                yerr=mitral_responses_se[odornum,mitnum],color=(odorA,odorB,sister_ratio))

    #### plot to compare air decorr
    fig = figure(facecolor='w')
    ax = fig.add_subplot(111)
    title('Air: mit0 (red) vs mit1 (blue)',fontsize=36)
    errorbar(x=responsetlist,y=mitral_responses_avg[6,0],\
        yerr=mitral_responses_se[6,0],color='r',linewidth=2)
    errorbar(x=responsetlist,y=mitral_responses_avg[6,1],\
        yerr=mitral_responses_se[6,1],color='b',linewidth=2)
    axes_labels(ax,'time (s)','firing rate (Hz)',adjustpos=True)

    #### plot to compare odor decorr
    glomnum = central_glom
    ## 1/3rd column 85mm/3/25.4 inches wide
    fig = figure(figsize=(columnwidth/3,linfig_height),dpi=300,facecolor='w')
    ax = fig.add_subplot(2,1,1,frameon=False)
    #title('OdorA: mit0 (red) vs mit1 (blue)',fontsize=36)
    simresponse = mitral_responses_avg[5,2*glomnum+0]
    simerr = mitral_responses_se[5,2*glomnum+0]
    fill_between(responsetlist, simresponse+simerr, simresponse-simerr,
        color='r',alpha=0.4)
    plot(responsetlist,simresponse,color='r',linewidth=linewidth)
    simresponse = mitral_responses_avg[5,2*glomnum+1]
    simerr = mitral_responses_se[5,2*glomnum+1]
    fill_between(responsetlist, simresponse+simerr, simresponse-simerr,
        color='b',alpha=0.4)
    plot(responsetlist,simresponse,color='b',linewidth=linewidth)
    ax.set_xlim(startplottime,endplottime)
    ax.get_xaxis().set_ticks_position('none')
    ax.get_yaxis().set_ticks_position('left')
    ## need to add the axes lines that I want, after deleting full frame.
    xmin, xmax = ax.get_xaxis().get_view_interval()
    ymin, ymax = ax.get_yaxis().get_view_interval()
    ax.set_xticks([])
    ax.set_yticks([ymin,ymax])
    ax.add_artist(Line2D((xmin, xmin), (ymin, ymax), color='black', linewidth=linewidth))
    ## separate respirations
    ax.add_artist(Line2D(((xmax+xmin)/2.0, (xmax+xmin)/2.0), (ymin, ymax), color='black', linewidth=linewidth))
    axes_labels(ax,'','') # sets default fontsize too

    ax = fig.add_subplot(2,1,2,frameon=False)
    #title('OdorB: mit0 (red) vs mit1 (blue)',fontsize=36)
    simresponse = mitral_responses_avg[0,2*glomnum+0]
    simerr = mitral_responses_se[0,2*glomnum+0]
    fill_between(responsetlist, simresponse+simerr, simresponse-simerr,
        color='r',alpha=0.4)
    plot(responsetlist,simresponse,color='r',linewidth=linewidth)
    simresponse = mitral_responses_avg[0,2*glomnum+1]
    simerr = mitral_responses_se[0,2*glomnum+1]
    fill_between(responsetlist, simresponse+simerr, simresponse-simerr,
        color='b',alpha=0.4)
    plot(responsetlist,simresponse,color='b',linewidth=linewidth)
    ax.set_xlim(startplottime,endplottime)
    ax.get_xaxis().set_ticks_position('bottom')
    ax.get_yaxis().set_ticks_position('left')
    ## need to add the axes lines that I want, after deleting full frame.
    xmin, xmax = ax.get_xaxis().get_view_interval()
    ymin, ymax = ax.get_yaxis().get_view_interval()
    ax.set_xticks([xmin,xmax])
    ax.set_yticks([ymin,ymax])
    ax.add_artist(Line2D((xmin, xmin), (ymin, ymax), color='black', linewidth=linewidth))
    ax.add_artist(Line2D((xmin, xmax), (ymin, ymin), color='black', linewidth=linewidth))
    ## separate respirations
    ax.add_artist(Line2D(((xmax+xmin)/2.0, (xmax+xmin)/2.0), (ymin, ymax), color='black', linewidth=linewidth))
    axes_labels(ax,'time (s)','firing rate (Hz)')

    ## Below text position is wrt to last axes drawn.
    ## transform = ax.transAxes sets the test position as axes units and not data units.
    #text(-0.42,1.4,'firing rate(Hz)', fontsize=label_fontsize, rotation='vertical', transform = ax.transAxes)
    #text(0.20,-0.27,'time (s)', fontsize=label_fontsize, transform = ax.transAxes)
    fig.tight_layout()
    fig.savefig('../figures/sistermorphs.png',dpi=fig.dpi)
    fig.savefig('../figures/sistermorphs.svg',dpi=fig.dpi)
    
def plot_morph_images(filename,mitral_responses_avg,mitral_responses_binned_list):
    params,chisq,inputsA,inputsB,fitted_responses,numavgs,firingbinsmeanList,firingbinserrList\
        = fit_morphs(filename, fitted_mitral)

    fig = figure(facecolor='none') # the string 'none' sets transparent facecolor
    mitnum = fitted_mitral
    numodors = len(inputList)
    mitral_responses_binned_list = array(mitral_responses_binned_list)
    ## air image will be padded on both sides of odor response below
    ## array[::-1] is for reversing elements in an array
    air_image = [ trialresponse[::-1] \
            for trialresponse in mitral_responses_binned_list[:,numodors-1,mitnum,:] ]
    numtrials = len(air_image)
    for odornum in range(numodors-1):
        # odors are mapped to odornum: B<->0, A<->5; so reverse the index = realodornum to show A to B
        realodornum = numodors-2-odornum
        ############## One image showing individual trials
        ax = fig.add_axes([0.15*(odornum)+0.025,0.471,0.15,0.2]) # [left,bottom,width,height]
        ## block of air responses are padded to odor on both sides
        ## array[::-1] is for reversing elements in an array
        response_image = []
        ## one has to do response_image = [] and then extend(),
        ## instead of response_image=air_image.
        ## the latter doesn't create a new array but is like a pointer to air_image
        ## so that later statements keep extending air_image
        response_image.extend(air_image)
        response_image.extend( [ trialresponse[::-1] \
            for trialresponse in mitral_responses_binned_list[:,realodornum,mitnum,:] ] )
        response_image.extend(air_image)
        im = ax.imshow(transpose(response_image), cmap=cm.jet, norm=None)
        title(['A','0.8A+\n0.6B','0.6A+\n0.4B','0.4A+\n0.6B','0.6A+\n0.8B','B','air'][odornum], fontsize=28)
        axes_off(ax)
        #cb = colorbar(im)
        #cb.set_clim(vmin=0,vmax=50)
        im.set_clim(vmin=0,vmax=50)
        ############### One image showing average trials
        ax = fig.add_axes([0.15*(odornum)+0.025,0.33,0.15,0.2]) # [left,bottom,width,height]
        ## block of air responses are padded to odor on both sides
        ## array[::-1] is for reversing elements in an array
        response_image = []
        response_image.extend( [fitted_responses[-1][::-1]]*numtrials )
        response_image.extend( [fitted_responses[realodornum][::-1]]*numtrials )
        response_image.extend( [fitted_responses[-1][::-1]]*numtrials )
        im = ax.imshow(transpose(response_image), cmap=cm.jet, norm=None)
        #title(['A','0.8A+\n0.6B','0.6A+\n0.4B','0.4A+\n0.6B','0.6A+\n0.8B','B','air'][odornum])
        axes_off(ax)
        #cb = colorbar(im)
        #cb.set_clim(vmin=0,vmax=50)
        im.set_clim(vmin=0,vmax=50)
    ax = fig.add_axes([0.94,0.38,0.01,0.24]) # [left,bottom,width,height]
    #ax.set_aspect(100.0)
    #ax.set_position([0.85,0.38,0.01,0.24]) # [left,bottom,width,height]
    cb = fig.colorbar(im, cax=ax)
    cb.set_clim(vmin=0,vmax=50)
    for t in cb.ax.get_yticklabels():
        t.set_fontsize(18)

def plot_morphs(filename):
    f = open(filename,'r')
    (mitral_responses_list,mitral_responses_binned_list) = pickle.load(f)
    f.close()
    mitral_responses_binned_list = \
        rebin(mitral_responses_list, numbins=PLOTRESP_NUM*NUMBINS,\
            bin_width_time=BIN_WIDTH_TIME, numresps=PLOTRESP_NUM)

    numavgs = len(mitral_responses_list)
    mitral_responses_avg = mean(mitral_responses_binned_list, axis=0)
    mitral_responses_std = std(mitral_responses_binned_list, axis=0)
    ## since I plot the mean response, I must plot standard error of the mean
    ## = standard deviation of a repeat / sqrt(num of repeats).
    ## NOTE: our plot depends on number of trials.
    mitral_responses_se = mitral_responses_std/sqrt(numavgs)

    if plot_images:
        #### Fit the morphs, then plot image plots of responses and fits a la Khan et al 2008.
        plot_morph_images(filename,mitral_responses_avg,mitral_responses_binned_list)
    else:
        #### Usual firing rate vs time plots of responses
        plot_morph_responses(mitral_responses_avg,mitral_responses_se)

    show()

if __name__ == "__main__":
    if len(sys.argv)<2:
        print "You need to specify the morph responses pickle filename."
        sys.exit(1)

    plot_morphs(sys.argv[1])