-
Notifications
You must be signed in to change notification settings - Fork 1
/
lfp.jl
206 lines (166 loc) · 7.67 KB
/
lfp.jl
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
## LFP CALCULATION AND PLOTTING POWER SPECTRA
## - built from referencing 'spike_train_correlation' in utilities.jl
## ---------------------------------------------
include("utilities.jl");
default(show=false)
default(fontfamily="Computer Modern")
# HYPERPARAMS
n_runs = 3
patterns = 0:12
duration = 2000
labels = ["theta" , "ftheta", "alpha", "beta", "gamma"]
freqlabels = [L"\theta" L"\theta_{fast}" L"\alpha" L"\beta" L"\gamma"]
fig_ext = ".png"
# CREATE NECESSARY DIRECTORIES
create_directories_if_not_exist()
# IDENTIFY NEURON POPULATION RANGES
populations = Dict(
"DG" => [0, 500],
"BC" => [500,506],
"MC" => [506, 521],
"HIPP" => [521, 527]
)
# FUNCTIONS
function makechunks(convolution::Vector, n_neurons::Integer)
chunks = length(convolution) ÷ n_neurons
return [convolution[1+chunks*k:(k == n_neurons-1 ? end : chunks*k+chunks)] for k = 0:n_neurons-1]
end
function compute_LFP(
spike_train::DataFrame,
n_neurons::Int64;
delta::Int64=5,
n_bins::Int64=1,
duration::Float64=2000.0)
# Create kernel
tri_rng = collect(1:round(Int64, delta/n_bins))
triangular_kernel = [tri_rng; reverse(tri_rng)[2:end]]'
# Create bins for histograms
bins = collect(1:n_bins:(duration + n_bins))
convolved_series = zeros(round(Int64, n_neurons * duration / n_bins))
for i ∈ 1:n_neurons
s = spike_train[spike_train.Neuron .== (i-1), "Time"]
timeseries, _ = np.histogram(s, bins)
idx_low = round(Int64, (i-1)*duration/n_bins + 1)
idx_high = round(Int64, i*duration/n_bins)
convolved_series[idx_low:idx_high] = np.convolve(timeseries, triangular_kernel, "same") #TODO: consider scipy.signal.fftconvolve as alt.
end
LFP = makechunks(convolved_series, n_neurons)
return sum(LFP)
end
global ppsave = []
global gcsave = []
lfp_save_pp = OrderedDict("theta"=>[], "ftheta"=>[], "alpha"=>[], "beta"=>[], "gamma"=>[])
lfp_save_gc = OrderedDict("theta"=>[], "ftheta"=>[], "alpha"=>[], "beta"=>[], "gamma"=>[])
for run_ ∈ 1:n_runs
for i ∈ 1:length(labels)
spikes = load_spike_files(patterns, labels[i]*"-$run_", populations)
for p ∈ unique(spikes.Pattern)
# TODO: change so this works ∀ neurons
pp = spikes[(spikes.Population .== "PP") .& (spikes.Pattern .== p), :]
gc = spikes[(spikes.Population .== "DG") .& (spikes.Pattern .== p), :]
#compute LFP
pp_lfp = [compute_LFP(pp, 100)]
gc_lfp = [compute_LFP(gc, 500)]
#save LFP for each pattern:
global ppsave = append!(ppsave, pp_lfp)
global gcsave = append!(gcsave, gc_lfp)
end
# average LFP across all patterns
sum_lfp_pat_pp = [sum(ppsave)/length(ppsave[1])]
sum_lfp_pat_gc = [sum(gcsave)/length(gcsave[1])]
# save summated lfp/freq band
append!(lfp_save_pp[labels[i]], sum_lfp_pat_pp)
append!(lfp_save_gc[labels[i]], sum_lfp_pat_gc)
end
end
#### GC Raster plots only ####
gcs = Dict(
"DG" => [0, 500],
)
for run_ ∈ 1:n_runs
for i ∈ 1:length(labels)
spikes = load_spike_files(patterns, labels[i]*"-$run_", populations)
# CREATE RASTER PLOTS
for p ∈ unique(spikes.Pattern)
stimin = spikes[(spikes.Population .== "PP") .& (spikes.Pattern .== p), :]
plots = []
append!(plots, [raster_plot(stimin; ylab="PP")])
for pop ∈ keys(gcs)
lb, ub = populations[pop]
popspikes = spikes[(spikes.Population .== pop) .& (spikes.Pattern .== p),:]
append!(plots, [raster_plot(popspikes; xlab="", ylab=pop)])
end
fig = plot(reverse(plots)..., layout=grid(2, 1, heights=[0.8, 0.2]), size=(500, 400), dpi = 300)
savefig(fig, "figures/raster-plots/raster-gconly--"*string(p)*"-"*labels[i]*"-$run_"*".png")
end
end
end
#-------PLOT LFPS
l = @layout [grid(2,1, heights=[0.8, 0.2])]
theta_lfp_pp = plot(lfp_save_pp["theta"][1],
color = :black,
xlabel = "Time (ms)",
ylabel = "PP",
label = nothing,
)
theta_lfp_gc = plot(lfp_save_gc["theta"][1],
color = :black,
ylabel = "DG",
label = nothing,
)
sumfig = plot(theta_lfp_gc, theta_lfp_pp, layout=l, size=(500, 400), dpi=300)
savefig(sumfig, "figures/lfp/lfps_gcs_pp.png")
#-------POWER SPECTRA
restruct_gclfps = zeros((duration, length(labels)))
restruct_pplfps = zeros((duration, length(labels)))
for i ∈ 1:length(labels)
restruct_gclfps[:,i] = lfp_save_gc[labels[i]][n_runs]
restruct_pplfps[:,i] = lfp_save_pp[labels[i]][n_runs]
end
# Calculate power spectra
S = spectra(restruct_gclfps, 500, duration; tapering=hamming) #TODO: learn what this 'tapering' is.
# Calculate power spectra
S = spectra(restruct_gclfps, 500, duration; tapering=hamming) #TODO: learn what this 'tapering' is.
# SUM LOW GAMMA POWER FOR 30Hz-60Hz
norm_gamma_pwrL = OrderedDict("theta"=>0.0, "ftheta"=>0.0, "alpha"=>0.0, "beta"=>0.0, "gamma"=>0.0)
for freq ∈ 1:length(labels)
norm_gamma_pwrL[labels[freq]] =sum(S.y[30:39,freq])
end
# SUM MED GAMMA POWER FOR 30Hz-60Hz
norm_gamma_pwrM = OrderedDict("theta"=>0.0, "ftheta"=>0.0, "alpha"=>0.0, "beta"=>0.0, "gamma"=>0.0)
for freq ∈ 1:length(labels)
norm_gamma_pwrM[labels[freq]] =sum(S.y[40:59,freq])
end
# SUM HIGH GAMMA POWER FOR 60-200Hz
norm_gamma_pwrH = OrderedDict("theta"=>0.0, "ftheta"=>0.0, "alpha"=>0.0, "beta"=>0.0, "gamma"=>0.0)
for freq ∈ 1:length(labels)
norm_gamma_pwrH[labels[freq]] =sum(S.y[60:200,freq])
end
CSV.write("figures/lfp/gamma_pwr_low.csv", norm_gamma_pwrL)
CSV.write("figures/lfp/gamma_pwr_med.csv", norm_gamma_pwrM)
CSV.write("figures/lfp/gamma_pwr_high.csv", norm_gamma_pwrH)
# PLOT GAMMA POWER BAR GRAPH
gamma_pwr = bar(vec(freqlabels), [[norm_gamma_pwrM[i] for i ∈ labels] [norm_gamma_pwrL[i] for i ∈ labels] [norm_gamma_pwrH[i] for i ∈ labels]],
labels = ["Low "*L"\gamma" "Med. "*L"\gamma" "High "*L"\gamma"],
color = [:grey75 :grey50 :black], xlabel = "Frequency Band", ylabel = L"\gamma"*" Power",
xtickfont=font(12), size = (350,350), legend = :topleft, dpi=300)
savefig(gamma_pwr, "figures/lfp/gamma_pwr_bar.png")
# PLOT GAMMA POWER LINE GRAPH
gamma_pwr_line = plot(vec(freqlabels), [[norm_gamma_pwrL[i] for i ∈ labels] [norm_gamma_pwrM[i] for i ∈ labels] [norm_gamma_pwrH[i] for i ∈ labels]],
labels = ["Low "*L"\gamma" "Med. "*L"\gamma" "High "*L"\gamma"], xtickfont=font(12),
color = [:grey75 :grey50 :black], ylims = (0, 40), linewidth = 2, xlabel = "Input Frequency Band", ylabel = L"\gamma"*" Power",
size = (350,350), legend = :topleft, dpi=300)
savefig(gamma_pwr_line, "figures/lfp/gamma_pwr_line.png")
# PLOT RAW SPECTRA
powerfig = plot(S, fmax = 50, label = freqlabels, legend = :topright, ylabel = "Power "*L"(\mu V^2)", dpi=300)
savefig(powerfig, "figures/lfp/powerspectra_gcs.png")
# PLOT HIGH FREQ RAW SPECTRA SEPERATELY
powerfigstack = plot(S, layout=grid(5,1, heights=[0.2,0.2,0.2,0.2,0.2]), xticks = (collect(0:5:50)),
xlims = (0,50), ylims=(0,35),label = freqlabels, linewidth = 3,
legend = :topright, ylabel = "Power "*L"(\mu V^2)",
xlabel = "Frequency (Hz)", size=(500, 800), c=:black, grid=:none, dpi=300)
savefig(powerfigstack, "figures/lfp/powerspectra_gcs_stacked.png")
# PLOT FREQUENCY-TIME HEATMAP
A = TFamplitude(lfp_save_gc["theta"][n_runs], 400, duration; fmax=40)
freqtime = heatmap(A.y, xlabel = "Time (ms)", ylabel = "Frequency (Hz)", dpi=300)
savefig(freqtime, "figures/lfp/gcs_freqtime.png")