In [1]:
import sys
sys.path.append("../")
import glob
import matplotlib
import matplotlib.pyplot as plt
import numpy as np
from Linearity import Neuron
from IPython.display import set_matplotlib_formats
import scipy.stats as ss
from pickle import dump as dump
plt.style.use('neuron')
In [2]:
def simpleaxis(axes, every=False, outward=True):
if not isinstance(axes, (list, np.ndarray)):
axes = [axes]
for ax in axes:
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
if (outward):
ax.spines['bottom'].set_position(('outward', 10))
ax.spines['left'].set_position(('outward', 10))
if every:
ax.spines['bottom'].set_visible(False)
ax.spines['left'].set_visible(False)
ax.get_xaxis().tick_bottom()
ax.get_yaxis().tick_left()
ax.set_title('')
1 C (CA3 heatmap)¶
In [5]:
ca3_neuron = Neuron.load('/media/sahil/NCBS_Shares_BGStim/patch_data/160927/c4_CA3_CPP/plots/c4_CA3_CPP.pkl')
In [51]:
AP_dict = np.zeros((12,12))
SubThP_dict = np.zeros((12,12))
for expType, exp in ca3_neuron:
if expType == "Control":
for coord in exp[1].coordwise:
temp_coord = []
for trial in exp[1].coordwise[coord].trials:
temp_coord.append(trial.AP_flag)
AP_dict[list(coord)[0]] = np.sum(temp_coord)
SubThP_dict[list(coord)[0]] = exp[1].coordwise[coord].average_feature[0]
In [52]:
SubThP_dict
Out[52]:
In [98]:
SubThP_dict = np.ma.masked_where(SubThP_dict == 0., SubThP_dict)
SubThP_dict = np.ma.masked_where(np.isnan(SubThP_dict), SubThP_dict)
cmap = plt.cm.viridis
cmap.set_bad(color='white')
fig, ax = plt.subplots()
heatmap = ax.pcolormesh(SubThP_dict, cmap=cmap, vmin=np.nanmin(SubThP_dict), vmax=np.nanmax(SubThP_dict))
# # mark a specific square?
# zeros = np.zeros((12,12))
# AP_coords = np.where(AP_dict>0)
# zeros[AP_coords] = 1.
# c = np.ma.masked_array(zeros, zeros == 0.) # mask squares where value == 1
# ax.pcolormesh(np.arange(13), np.arange(13), c, alpha=1.0, zorder=2, facecolor='none', edgecolors='k',
# cmap='gray')
for y in range(AP_dict.shape[0]):
for x in range(AP_dict.shape[1]):
if AP_dict[y, x] > 0:
plt.text(x + 0.5, y + 0.5, "{}/4".format(int(AP_dict[y, x])),
horizontalalignment='center',
verticalalignment='center', size=10)
circle1 = plt.Circle((7.5, 7.5), 0.75, color='r', fill=False)
ax.invert_yaxis()
for axis in [ax.xaxis, ax.yaxis]:
axis.set_ticks(np.arange(1,13), minor=True)
axis.set(ticks=np.arange(0,13,2)+0.5, ticklabels=np.arange(0,13,2)) #Skipping square labels
ax.grid(True, which='minor', axis='both', linestyle='--', alpha=0.1, color='k')
ax.set_xlim((0,13))
ax.set_ylim((0,13))
#Colorbar stuff
cbar = plt.colorbar(heatmap, label="Average response (mV)")
cbar.ax.get_yaxis().labelpad = 6
tick_locator = matplotlib.ticker.MaxNLocator(nbins=5)
cbar.locator = tick_locator
cbar.update_ticks()
ax.add_artist(circle1)
ax.set_aspect(1)
fig.set_figheight(2.5)
fig.set_figwidth(3)
simpleaxis(ax,every=True,outward=False)
dump(fig,file('figures/fig1/1c.pkl','wb'))
plt.show()
1 D (CA3 jitter)¶
In [89]:
fig, ax = plt.subplots()
AP_dict = np.zeros((12,12))
tsp = []
for expType, exp in ca3_neuron:
if expType == "Control":
for trial in exp[1].coordwise[frozenset([(7,7)])].trials:
time = np.linspace(0,100,len(trial.interestWindow))
ax.plot(time, trial.interestWindow, color=(170/256., 0, 0))
tsp.append(time[np.argmax(trial.interestWindow)])
ax.set_xlabel("Time")
ax.set_ylabel("$V_m$")
ax.plot(tsp, [0]*len(tsp), 'k*', markersize=5)
print(np.std(tsp, ddof = 1))
fig.set_figheight(2)
fig.set_figwidth(2)
ax.get_xaxis().tick_bottom()
ax.get_yaxis().tick_left()
simpleaxis(ax, outward=True)
dump(fig,file('figures/fig1/1d.pkl','wb'))
plt.show()
1 E (CA3 jitter distribution)¶
In [90]:
fig, ax = plt.subplots()
AP_dict = np.zeros((12,12))
jitter = []
for expType, exp in ca3_neuron:
if expType == "Control":
for coord in exp[1].coordwise:
tsp = []
for trial in exp[1].coordwise[coord].trials:
time = np.linspace(0,100,len(trial.interestWindow))
tsp.append(time[np.argmax(trial.interestWindow)])
jitter.append(np.std(tsp, ddof = 1))
ax.set_xlabel("CA3 jitter (ms)")
ax.set_ylabel("# trials")
bins = np.linspace(0,5,20)
ax.hist(jitter, bins=bins)
ax.set_xlim((0,5))
fig.set_figheight(2)
fig.set_figwidth(2)
simpleaxis(ax)
dump(fig,file('figures/fig1/1e.pkl','wb'))
plt.show()
1 F (CA1 heatmap)¶
In [100]:
neuronList = []
ca1_neuron = Neuron.load('/media/sahil/NCBS_Shares_BGStim/patch_data/161013/c1/plots/c1.pkl')
In [101]:
coordwise = ca1_neuron.experiment["Control"][1].coordwise
epsp_dict = np.zeros((12,12))
for coord in coordwise:
epsp_dict[list(coord)[0]] = coordwise[coord].average_feature[0]
In [102]:
epsp_dict = np.ma.masked_where(epsp_dict == 0., epsp_dict)
cmap = plt.cm.viridis
cmap.set_bad(color='white')
fig, ax = plt.subplots()
heatmap = ax.pcolormesh(epsp_dict, cmap=cmap)
circle1 = plt.Circle((5.5, 7.5), 0.75, color='r', fill=False)
ax.set_title("CA1 cell response")
ax.invert_yaxis()
for axis in [ax.xaxis, ax.yaxis]:
axis.set_ticks(np.arange(1,13), minor=True)
axis.set(ticks=np.arange(0,13,2)+0.5, ticklabels=np.arange(0,13,2)) #Skipping square labels
ax.grid(True, which='minor', axis='both', linestyle='--', alpha=0.1, color='k')
ax.set_xlim((0,13))
ax.set_ylim((0,13))
# Colorbar stuff
cbar = plt.colorbar(heatmap, label="Average response (mV)")
cbar.ax.get_yaxis().labelpad = 10
tick_locator = matplotlib.ticker.MaxNLocator(nbins=5)
cbar.locator = tick_locator
cbar.update_ticks()
ax.add_artist(circle1)
ax.set_aspect(1)
simpleaxis(ax, every=True, outward=False)
dump(fig,file('figures/fig1/1f.pkl','wb'))
fig.set_figheight(2.5)
fig.set_figwidth(3)
plt.show()
1 G (CA1 jitter)¶
In [103]:
fig, ax = plt.subplots()
tsp = []
for expType, exp in ca1_neuron:
if expType == "Control":
for trial in exp[1].coordwise[frozenset([(7,5)])].trials:
time = np.linspace(0,100,len(trial.interestWindow))
ax.plot(time, trial.interestWindow, color=(0,.5,.5))
tsp.append(time[np.argmax(trial.interestWindow)])
ax.set_xlabel("Time")
ax.set_ylabel("$V_m$")
ax.plot(tsp, [0]*len(tsp), 'k*', markersize=5)
ax.get_xaxis().tick_bottom()
ax.get_yaxis().tick_left()
print(np.std(tsp, ddof=1))
fig.set_figheight(2)
fig.set_figwidth(2)
simpleaxis(ax)
dump(fig,file('figures/fig1/1g.pkl','wb'))
plt.show()
1 H (CA1 jitter distribution)¶
In [104]:
fig, ax = plt.subplots()
ca1_jitter = []
for expType, exp in ca1_neuron:
if expType == "Control":
for coord in exp[1].coordwise:
tsp = []
for trial in exp[1].coordwise[coord].trials:
time = np.linspace(0,100,len(trial.interestWindow))
tsp.append(time[np.argmax(trial.interestWindow)])
ca1_jitter.append(np.std(tsp, ddof = 1))
ax.set_xlabel("CA1 jitter (ms)")
ax.set_ylabel("# trials")
bins = np.linspace(0,50,20)
ax.hist(ca1_jitter, bins=bins)
ax.set_xlim((0,50))
fig.set_figheight(2)
fig.set_figwidth(2)
simpleaxis(ax)
dump(fig,file('figures/fig1/1h.pkl','wb'))
plt.show()
1 I, J (CA1 amplitude variability )¶
In [105]:
ca1_amp_jitter = []
average_psp = []
for expType, exp in ca1_neuron:
if expType == "Control":
for coord in exp[1].coordwise:
amp = []
for trial in exp[1].coordwise[coord].trials:
amp.append(trial.feature[0])
average_psp.append(exp[1].coordwise[coord].average_feature[0])
ca1_amp_jitter.append(np.std(amp, ddof = 1))
fig, ax = plt.subplots()
ax.set_xlabel("CA1 amplitude variability (mV)")
ax.set_ylabel("# trials")
bins = 10
ax.hist(ca1_amp_jitter, bins=bins)
ax.set_xlim((0,1))
fig.set_figheight(2)
fig.set_figwidth(2)
simpleaxis(ax)
dump(fig,file('figures/fig1/1i.pkl','wb'))
plt.show()
fig, ax = plt.subplots()
average_psp = np.array(average_psp)
indices = np.argsort(average_psp)
ca1_amp_jitter = np.array(ca1_amp_jitter)
ax.bar(np.arange(len(average_psp)), average_psp[indices], yerr = ca1_amp_jitter[indices], ecolor='0.5')
ax.set_xlabel("sorted stimulus index")
ax.set_ylabel("average amplitude")
fig.set_figheight(2)
fig.set_figwidth(2)
simpleaxis(ax)
dump(fig,file('figures/fig1/1j.pkl','wb'))
plt.show()
Population statistics
In [106]:
neuronList = []
with open("/media/sahil/NCBS_Shares_BGStim/patch_data/current_clamp_files.txt", 'r') as fp:
fileList = fp.read().splitlines()
for i,filename in enumerate(fileList):
fileList[i] += "plots/" + filename.split('/')[-2] + ".pkl"
try:
print ("Reading ", fileList[i])
neuronList.append(Neuron.load(fileList[i]))
except:
print ("Couldn't read", fileList[i])
print ("Read {} files.".format(len(fileList)))
1 K (distribution of EPSP one squares)¶
In [109]:
def figureK(feature = 0):
''' Get's the fraction of times an AP occurs verus PSPs '''
epsp_1sq = []
squares = 1
feature = 0
numBins = 200
for neuron in neuronList:
try:
trials = neuron.experiment["Control"][squares].trial
for trialNum in trials:
trial = neuron.experiment["Control"][squares].trial[trialNum]
epsp_1sq.append(trial.feature[feature])
except:
print("{}, {}: No control dataset!".format(neuron.index, neuron.date))
#print (np.mean(epsp_1sq), np.median(epsp_1sq), ss.mode(epsp_1sq))
fig, ax = plt.subplots()
n, bins, patches = ax.hist(epsp_1sq, bins=numBins)
maxBin = np.argmax(n)
modeEPSP = bins[maxBin]
ymin, ymax = ax.get_ylim()
ax.vlines(modeEPSP,ymin,ymax, linestyles='--', color='0.5')
ax.annotate("{:.2f} mV".format(modeEPSP), xy=(modeEPSP, ymax), xytext=(1.2*modeEPSP, ymax), textcoords='data')
ax.set_xlabel("PSP Amplitude (mV)")
ax.set_ylabel("# recorded PSPs")
fig.set_figheight(2)
fig.set_figwidth(2)
simpleaxis(ax, outward=False)
dump(fig,file('figures/fig1/1k.pkl','wb'))
plt.show()
In [ ]:
figureK()
1 L (Fraction of APs and sublinear responses)¶
In [111]:
def figureL(feature = 0):
''' Get's the fraction of times an AP occurs verus PSPs '''
numNeurons = 0
numAP = 0
numPSP = 0
for neuron in neuronList:
for experiment in neuron.experiment:
if experiment == "Control":
numNeurons+=1
for squares in neuron.experiment[experiment]:
nSq = neuron.experiment[experiment][squares].trial
for trialnum in nSq:
if nSq[trialnum].AP_flag:
numAP+=1
else:
numPSP+=1
fracAPs = float(numAP)/numPSP
fig, ax = plt.subplots()
index = np.arange(2)
width = 0.5
ax.bar(index, (numPSP, numAP), width)
ax.set_title("CA1 responses")
ax.set_ylabel("Frequency")
ax.set_xticks(index)
ax.set_xticklabels(('PSP', 'AP'))
#ax.set_yticks(list(ax.get_yticks()) + [numAP])
ax.annotate("{:.2f} %".format(fracAPs*1e2), xy=(0.8, numAP*2.0), xytext=(0.75, numAP*5), textcoords='data')
xaxisVal = ax.get_xlim()[0]
ax.hlines(numAP, xaxisVal, index[1],linestyle="--")
fig.set_figheight(2)
fig.set_figwidth(2)
ax.set_ylim(0,22500)
# ax.set_yscale('log')
simpleaxis(ax, outward=False)
ax.ticklabel_format(style='sci', axis='y', scilimits=(0,0))
ax.yaxis.major.formatter._useMathText = True
dump(fig,file('figures/fig1/1l.pkl','wb'))
plt.show()
print ("Legend text: Number of cells: {}, number of points: {}, fraction of APS: {}".format(numNeurons, numAP+numPSP, fracAPs))
In [112]:
figureL()
1 M (Desensitization of CA3)¶
In [113]:
ca3_cell = '/media/sahil/NCBS_Shares_BGStim/patch_data/161013/c2_CA3_CPP/CPP/desensitization_161013_c2.txt'
In [114]:
desenCell = np.loadtxt(ca3_cell, delimiter=',')
In [115]:
fig, ax = plt.subplots()
ax.plot(desenCell, 'o--', markersize=3, mfc='1', markeredgecolor='0.5')
# ax.bar(range(len(desenCell)), desenCell)
ax.set_ylim(0,10)
ax.set_xlabel("Trial number")
ax.set_ylabel("Number of APs")
simpleaxis(ax)
fig.set_figheight(2)
fig.set_figwidth(2)
dump(fig,file('figures/fig1/1m.pkl','wb'))
plt.show()
