LNCD

Table of Contents

  • LNCD Home
  • Administration
  • Notebooks
  • Journal Club Presentations
  • Publications
  • Current Projects
  • Completed Projects
  • Current Grants
  • Datasets by Project
  • Brain ROIs and Measures
  • ️Tools And Methods
  • Big Data
  • RA Homepage
  • Undergrad Resources
  • Recent Changes
  • Maintenance
  • Site Map
  • Random Page
LNCD
Docs » INTs on Macaque LFPs

INTs on Macaque LFPs

  • The MATLAB script is named ACW_MacaqueLFPs.m but then I named the function appropriately playingWithLFPs lol
  • Takes in the task (delay or fix)
  • Adds appropriate paths, including the path to the raw data
  • Sets desired result path
 
function playingWithLFPs(task)

addpath((('/Volumes/Hera/Projects/7TBrainMech/scripts/eeg/Shane/Entropy/')));
addpath((('/Volumes/Hera/Projects/7TBrainMech/scripts/eeg/Shane/Entropy/code')));
addpath('/Volumes/Hera/Shane/LFPData-20250428T155619Z-001/LFPData/')


if strcmp(task, 'delay')

    resultPath = '/Volumes/Hera/Projects/7TBrainMech/scripts/eeg/Shane/Entropy/Results/acw/MonkeyData/individual_files_delay';

elseif strcmp(task, 'fix')

    resultPath = '/Volumes/Hera/Projects/7TBrainMech/scripts/eeg/Shane/Entropy/Results/acw/MonkeyData/individual_files_fix';

end
  • Loads an excel file called channel_info
  • Within this excel is a table with 10 columns as follows
    • session: monkeyID and session number (i.e. QUA001_1)
    • Neuron_age: time (in days) between the day of recording and the subjects mature age
    • Neuron_area: localization of electrode within the lateral prefrontal cortex
    • ID: monkeys ID
    • group: quantization of neuron age
    • mature_age: post-natal, chronological age (in days) of 'mid-adolescence', defined by the bone plate closure
    • mature_group: quantization of mature age
    • age_group: quantization of (mature age + neuron age)
    • delay_duration: time between cue offset and go signal for the task done in that session
    • channel: channel number
  • Important: the monkeys actual age at that time of the session is neuron_age + mature_age in days
channel_info = readtable('channel_info.xlsx');

get_filename = @(row) sprintf('%s_CH%d.mat', row.session{1}, row.channel); % Naming convention made up of behavior session name and electrode number
cue_duration = 0.5; % seconds
  • For each channel, do the following
  • Pull out the delay_duration from the channel info, aka the time between the cue offset and go signal
  • Load the LFP data structure
  • Define the filename, monkey, behavioral session, channel
for i_electrode=1:size(channel_info,1)
    
    delay_duration = channel_info.delay_duration(i_electrode); % Time between cue offset and go signal
    load(get_filename(channel_info(i_electrode, :)), 'LFPData');
    
    filename = get_filename(channel_info(i_electrode, :));
    monkey = filename(1:3);
    behavSess = filename(4:6);
    channel = filename(10:12);
  • Check if the file exists
  • The data is arranged into a session-class-ntr topology, where trials (ntr) in which the same stimulus was presented are grouped together (class), and the indices of the class array (1 through to 8) implicitly map onto the manipulated parameter of the stimulus, which is the polar angle of the stimulus location on the 2D plane of the monitor (0 through to 7π/4). “class_to_radian” does this conversion.
  • A specific period of LFP data can be extracted using the “get_lfp_slice” function. The 2nd argument t0 of the “get_lfp_slice” function specifies a time point (in seconds) around which to include LFP data points up to -t_slice(1) seconds prior to up until t_slice(2) seconds after said time point t0. The time point that we generally align each trial to is “Cue_onT” which is the start of the stimulus. To find the fixation period, we generally use t_slice = [-1, 0]; to find the delay period, we generally use t_slice = [cue_duration, cue_duration + delay_duration] (the two duration values are defined once each in the demo).
  • Grab sampling rate from the LFP data structure
  • Depending on if its delay or fix that we want, we define the target time as
    • Delay: target_time = [cue_duration, cue_duration+delay_duration]; % Time around cue onset from which LFP is to be aligned to
    • Fix: target_time = [-1, 0]; % Time around cue onset from which LFP is to be aligned to
  • Define a dummy slice of data that we will fill in
    
    if ~isfile([resultPath '/' monkey '_acw_' 'behavSess' behavSess '_' channel '.csv'])
        
        
        get_lfp_slice = @(fs, t0, t_slice) [round(t_slice(1)*fs):round(t_slice(2)*fs)] + round(t0*fs);
   
        fs = LFPData.class(1).ntr(1).Sampling_Rate;
         
        if strcmp(task, 'delay')

            target_time = [cue_duration, cue_duration+delay_duration]; % Time around cue onset from which LFP is to be aligned to

        elseif strcmp(task, 'fix')
            
            target_time = [-1, 0]; % Time around cue onset from which LFP is to be aligned to

        end

        dummy_lfp_slice = get_lfp_slice(fs, 0, target_time);
        t_slice = dummy_lfp_slice/fs;
        
        clear lfp_current_class
        
        lfp_by_class = cell(numel(LFPData.class), 1);  % one cell per class
       
  • For each class (each stimulus that was presented to the monkeys) and each trial, grab the slice of the LFP data corresponsing to the target time (defined by either the delay or fixation block)
  • Store the data in one variable
        for i_class = 1:numel(LFPData.class)
            n_trials = numel(LFPData.class(i_class).ntr);
            lfp_trials = zeros(n_trials, numel(get_lfp_slice(fs, 0, target_time)));  % preallocate per class
            
            for i_trial = 1:n_trials
                t_idx = get_lfp_slice(fs, LFPData.class(i_class).ntr(i_trial).Cue_onT, target_time);
                try
                lfp_trials(i_trial, :) = LFPData.class(i_class).ntr(i_trial).LFP(t_idx);
                catch
                end
            end
            
            lfp_by_class{i_class} = lfp_trials;  % store trials (2D array) for each class
        end
        
        lfp_all_trials = vertcat(lfp_by_class{:});  % size: [total_trials x timepoints]

        
        % plot(t_slice, mean(lfp_all_trials,1))
        % plot(t_slice, lfp_all_trials(2,:))
        
        
  • Take in each trial from the LFP, resample to 150 (or whatever you want)
  • Define the ACWout variable
  • Calculate the ACW values using the acw function
  • Write all necessary info to a table
  • Save table out
        
        clear ACWout
        
        for t = 1:size(lfp_all_trials,1)
            lfp_down = resample(lfp_all_trials(t,:), 150, 500);

            ACWout{t,1} = monkey;
            ACWout{t,2} = behavSess;
            ACWout{t,3} = channel;
            ACWout{t,4} = t;
            [ACWout{t,5}, ACWout{t,6}, acf, lags] = acw(lfp_down,150, 1);
        end
        
        % Add the new column to the existing table
        subjectTable = array2table(ACWout, 'VariableNames', {'Monkey', 'BehavSess', 'Channel', 'Trial','ACW_0', 'ACW_50'});
        
        subjectTable.NeuronArea = repmat(cell2mat(channel_info(i_electrode, :).Neuron_area), size(ACWout,1),1);
        subjectTable.MatureAge = repmat((channel_info(i_electrode, :).mature_age), size(ACWout,1),1);
        subjectTable.chronological_age = repmat((channel_info(i_electrode, :).Neuron_age + channel_info(i_electrode, :).mature_age), size(ACWout,1),1);
        subjectTable.group = repmat((channel_info(i_electrode, :).group), size(ACWout,1),1);
        subjectTable.mature_group = repmat((channel_info(i_electrode, :).mature_group), size(ACWout,1),1);
        subjectTable.age_group = repmat((channel_info(i_electrode, :).age_group), size(ACWout,1),1);

        subjectSavePath = [resultPath '/' monkey '_acw_' 'behavSess' behavSess '_' channel '.csv'];
        writetable(subjectTable, subjectSavePath)
        
    end
    
end
Previous Next