-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathrunJB12Model.m
143 lines (124 loc) · 6.72 KB
/
runJB12Model.m
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
%RUNJB12MODEL - Run the DynaSim implementation of the (Benita et al., 2012) model!
%{
This is a starter script to show both how to use the DynaSim mechanisms I've
made for simulating the (Benita et al., 2012) model in Dynasim, available
here (https://github.com/asoplata/dynasim-benita-2012-model) while also
giving you a good skeleton to start experimenting yourself.
- Install:
1. Install DynaSim (https://github.com/DynaSim/DynaSim/wiki/Installation),
including adding it to your MATLAB path.
2. `git clone` or download this code's repo
(https://github.com/asoplata/dynasim-benita-2012-model) into
'/your/path/to/dynasim/models', i.e. the 'models' subdirectory of your
copy of the DynaSim repo.
3. Run this file.
4. Believe it or not...that should be it! You should be able to start MATLAB
in your DynaSim code directory and run this script successfully! Let me
know if there are problems, at austin.soplata 'at-symbol-thingy' gmail
'dot' com
- Notes:
- The default model is rather large, so you may wish to decrease the
`numCellScaledownFactor` variable to decrease the size of your model.
- Due to the complexity of the synaptic mechanisms, this REQUIRES that you
only use EULER integration for solving the ODE system. In the future, other
integration methods may be supported.
- Dependencies:
- This has only been tested on MATLAB version 2017a.
- References:
- Benita, J. M., Guillamon, A., Deco, G., & Sanchez-Vives, M. V. (2012).
Synaptic depression and slow oscillatory activity in a biophysical
network model of the cerebral cortex. Frontiers in Computational
Neuroscience, 6. https://doi.org/10.3389/fncom.2012.00064
This is descended from DynaSim's `tutorial.m` script, at
https://github.com/DynaSim/DynaSim/blob/master/demos/tutorial.m
Author: Austin E. Soplata <[email protected]>
Copyright (C) 2018 Austin E. Soplata, Boston University, USA
%}
% -------------------------------------------------------------------
%% 1. Define simulation parameters
% -------------------------------------------------------------------
% Setting `study_dir` to `mfilename`, which is a reserved word, will
% automatically run the simulation and deposit its data and metadata into a
% new folder in the current directory that has the SAME NAME as this file:
study_dir = mfilename;
% % If you're on a cluster and instead want to save the simulation results to a
% % folder in a different location, use the commented-out line below instead:
% study_dir = strcat('/projectnb/crc-nak/asoplata/p25-anesthesia-grant-sim-data/', mfilename);
% Debug: If you want to completely clean the environment and remove all data,
% set `debug` to 1:
debug = 1;
if debug == 1
clf
close all
% Note: the below may be dangerous!
if 7==exist(study_dir, 'dir')
rmdir(study_dir, 's')
end
end
% This is where you set the length of your simulation, in ms
time_end = 2000; % in milliseconds
% While DynaSim uses a default `dt` of 0.01 ms, we must specify ours explicitly
% since `dt` is actually used to construct our model directly.
dt = 0.01; % in milliseconds
% For the full size model (1024 PYso's and PYdr's, 256 IN), use a
% `numCellsScaledownFactor` of 1. To lower the number of cells simulated, but
% keep the same proportions, decrease this number to something > 0. As an
% example, using a numCellsScaledownFactor of 0.1 (aka using a size of 10%)
% would decrease the population sizes to 102 PYso's and PYdr's, and 26 INs.
% numCellsScaledownFactor = 1;
numCellsScaledownFactor = 0.1;
% "Vary" parameters, aka parameters to be varied -- this tells DynaSim to run a
% simulation for all combinations of values. For a tutorial on how to use
% this, see
% https://github.com/DynaSim/DynaSim/wiki/DynaSim-Getting-Started-Tutorial#running-sets-of-simulations-by-varying-model-parameters
vary={
'PYso', 'appliedStim', 0.1;
'PYdr', 'appliedStim', 0.1;
'IN', 'appliedStim', 0.1;
};
simulator_options={
'cluster_flag', 0,... % Whether to submit simulation jobs to a cluster, 0 or 1
'cluster_matlab_version','2017b',...% TODO
'compile_flag', 0,... % Whether to compile simulation using MEX, 0 or 1
'disk_flag', 0,...% TODO
'downsample_factor', 10,... % How much to downsample data, proportionally {integer}
'dsPlot2_no_spikes', 1,...% TODO
'dt', 0.01,... % Fixed time step, in milliseconds
'memory_limit', '8G',... % Memory limit for use on cluster
'mex_flag', 0,...% TODO
'num_cores', 1,... % Number of CPU cores to use, including on cluster
'overwrite_flag', 1,... % Whether to overwrite simulation raw data, 0 or 1
'parfor_flag', 0,... % Whether to use parfor if running multiple local sims, 0 or 1
'random_seed', 'shuffle',...% What seed to use, or to randomize
'save_data_flag', 0,... % Whether to save raw output data, 0 or 1
'save_results_flag', 1,... % Whether to save output plots and analyses, 0 or 1
'solver', 'euler',... % Numerical integration method {'euler','rk1','rk2','rk4'}
'study_dir', study_dir,... % Where to save simulation results and code
'tspan', [0 time_end],... % Time vector of simulation, [beg end], in milliseconds
'verbose_flag', 1,... % Whether to display process info, 0 or 1
};
% 'plot_functions', {@dsPlot, @dsPlot, @dsPlot},...% Which plot functions to call automatically
% 'plot_options', {{'plot_type', 'waveform'},... % Arguments to pass to each of those plot functions
% {'plot_type', 'rastergram'},...
% {'plot_type', 'power'},...
% },...
% -------------------------------------------------------------------
%% 2. Assemble and customize the model
% -------------------------------------------------------------------
% This builds the complete model, including all populations and connections.
spec = assembleJB12Spec(numCellsScaledownFactor);
% % Only run this if you do NOT want any noise/randomness in your initial
% % conditions, which can be useful for reproducibility or debugging.
% spec = removeNoiseIC(spec);
% -------------------------------------------------------------------
%% 3. Run the simulation
% -------------------------------------------------------------------
% For an explanation of the arguments to `dsSimulate`, see the DynaSim code
% file `dsSimulate.m`.
data=dsSimulate(spec,'vary',vary,simulator_options{:});
% -------------------------------------------------------------------
%% 4. (Optional) Plot the results of the simulation post hoc
% -------------------------------------------------------------------
% For an explanation of the arguments to `dsSimulate`, see the DynaSim code
% If you want to run your own plotting interactively, load the data using:
dsPlot(data);