OSCR

Neural mechanisms of time-forward predictions for naturalistic auditory tone sequences.

Code ↔ Paper

26 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 26 matches · 2 of them tie a paragraph to a whole file, not to given lines: weak matches, whose lines are not tinted
  1. [1] § Methods › Directed connectivity analysis with Granger causality ↔ AnalysisCode/Analysis_and_Plotting/Connectivity/NASTD_ECoG_Connectivity_Main_lk.m, lines 996–1094 · score 0.88 · IFC ventral, dlPFC, auditory hierarchy, auditory cortex, BA, Brodmann
  2. [2] § Results › Changes in directed connectivity during sequence processing and sensory prediction ↔ AnalysisCode/Analysis_and_Plotting/Connectivity/NASTD_ECoG_Connectivity_Main_lk.m, lines 996–1094 · score 0.83 · dlPFC, Brodmann area, auditory hierarchy, auditory cortex, IPL, IFC
  3. [3] § Methods › iEEG data preprocessing ↔ AnalysisCode/Analysis_and_Plotting/Prediction/Helper Functions/NASTD_ECoG_Predict_CompPredp33_Subs_ManualArtifactCorrection.m, lines 43–149 · score 0.82 · Hilbert transform, amplitude envelop, baseline corrected, FieldTrip, smoothed, artifacts
  4. [4] § Methods › iEEG data preprocessing ↔ AnalysisCode/Preprocessing/IED_detector_rev3.7.2014/spike_detector_hilbert_v16_byISARG.m, lines 1–84 · score 0.79 · interictal epileptiform discharges, frequency bands, Hilbert, Butterworth, algorithm, IED
  5. [5] § Methods › Directed connectivity analysis with Granger causality ↔ AnalysisCode/Analysis_and_Plotting/Connectivity/HelperFunctions/NASTD_ECoG_Connectivity_CalculateGC_oldoption.m, lines 394–533 · score 0.76 · median model, Granger Causality, full model, selected model, pairwise, electrode pair
  6. [6] § Methods › iEEG data preprocessing ↔ AnalysisCode/Preprocessing/IED_detector_rev3.7.2014/spike_detector_hilbert_v16_byISARG.m, lines 1–84 · score 0.72 · interictal epileptiform discharges, detection, segments, algorithm, events, IEDs
  7. [7] § Methods › iEEG data preprocessing ↔ AnalysisCode/Preprocessing/NASTD_ECoG_Preproc_Main.m, lines 254–317 · score 0.65 · Heartbeat related artifacts, algorithm, ECG, pulse, phase, thresholding
  8. [8] § Methods › Neural correlates of sensory history integration (SHI) ↔ AnalysisCode/Analysis_and_Plotting/HistoryTracking/HelperFunctions/NASTD_ECoG_HisTrack_CombineExpKvals.m, lines 118–169 · score 0.63 · cross validation, Model selection, fold, squared, regression, winning
  9. [9] § Methods › Neural correlates of sensory history integration (SHI) ↔ AnalysisCode/Analysis_and_Plotting/HistoryTracking/HelperFunctions/NASTD_ECoG_HisTrack_CompExpKvals.m, lines 525–591 · score 0.63 · cross validation, Model selection, fold, squared, regression, winning
  10. [10] § Results › Changes in directed connectivity during sequence processing and sensory prediction ↔ AnalysisCode/Analysis_and_Plotting/Connectivity/HelperFunctions/NASTD_ECoG_Connectivity_CalculateGC_oldoption.m, lines 394–533 · score 0.61 · frequency domain GC, Granger causality, selected electrodes, pairwise, electrode pair, spectral
  11. [11] § Results › Paradigm and experimental design ↔ ParadigmCode/sfa_expt4_makeTrainingStim.m, lines 70–109 · score 0.61 · 220–880 Hz, penultimate tone, unique sequences, 220 Hz, semitones, tone pitch
  12. [12] § Results › Paradigm and experimental design ↔ ParadigmCode/sfa_expt4_makeStim.m, lines 11–115 · score 0.61 · 220–880 Hz, penultimate tone, unique sequences, 220 Hz, semitones, tone pitch
  13. [13] § Methods › Directed connectivity analysis with Granger causality ↔ AnalysisCode/Analysis_and_Plotting/Connectivity/HelperFunctions/NASTD_ECoG_Connectivity_PlotSignElecCon_AllSubTDLobes.m, lines 1640–1726 · score 0.60 · frontal parietal, parietal temporal, frontal temporal, lobes, prediction error, electrode pairs
  14. [14] § Methods › iEEG data preprocessing ↔ AnalysisCode/Preprocessing/NASTD_ECoG_Preproc_Main.m, lines 254–317 · score 0.60 · phase shift, pass filter, Butterworth, zero, block, preprocessing
  15. [15] § Methods › iEEG data preprocessing ↔ AnalysisCode/Preprocessing/NASTD_ECoG_Preproc_SubPreprocSettings.m, lines 412–483 · score 0.60 · band stop, pass filter, lying, noise, block, preprocessing
  16. [16] § Results › Neural correlates of tone sequence tracking ↔ AnalysisCode/Analysis_and_Plotting/StimulusCorrelation/NASTD_ECoG_StimCorr_Main_lk.m, lines 100–164 · score 0.58 · amplitude envelopes, range activity, high gamma, field, ERF, MEG
  17. [17] § Results › Neural correlates of tone sequence tracking ↔ AnalysisCode/Analysis_and_Plotting/Prediction/NASTD_ECoG_Predict_Main_lk.m, lines 119–216 · score 0.58 · amplitude envelopes, range activity, high gamma, field, ERF, MEG
  18. [18] § Methods › Paradigm and experimental setup ↔ ParadigmCode/sfa_expt4_getParamsTraining.m, the whole file · a weak match · score 0.57 · fixation point, instructed, keyboard, laptop, training, window
  19. [19] § Methods › Neural correlates of tone sequence tracking ↔ AnalysisCode/Analysis_and_Plotting/Prediction/NASTD_ECoG_Predict_Main_lk.m, lines 119–216 · score 0.55 · amplitude envelops, frequency band, high gamma, tone sequence, beta, alpha
  20. [20] § Methods › Neural correlates of prediction and prediction error ↔ AnalysisCode/Analysis_and_Plotting/Prediction/Helper Functions/NASTD_ECoG_Predict_CompPred_Sample_Subs_FTPL.m, lines 133–189 · score 0.55 · Temporal clusters, linear regressions, prediction error, permutation, shuffling, tone duration
  21. [21] § Methods › Directed connectivity analysis with Granger causality ↔ AnalysisCode/Analysis_and_Plotting/Connectivity/HelperFunctions/NASTD_ECoG_Connectivity_CalculateGC_pertone.m, lines 510–565 · score 0.53 · Granger Causality, full model, pairwise, connectivity, tone
  22. [22] § Methods › Paradigm and experimental setup ↔ ParadigmCode/sfa_expt4_getParams.m, the whole file · a weak match · score 0.52 · fixation point, keyboard, laptop, feedback, window, Paradigm
  23. [23] § Results › Changes in directed connectivity during sequence processing and sensory prediction ↔ AnalysisCode/Analysis_and_Plotting/Connectivity/HelperFunctions/NASTD_ECoG_Connectivity_PlotGC_statp1top31_acrossTDs+seqs.m, lines 1–71 · score 0.52 · linear fit, PE Pred, slopes, electrode pairing, broadband, lobe
  24. [24] § Results › Changes in directed connectivity during sequence processing and sensory prediction ↔ AnalysisCode/Analysis_and_Plotting/Connectivity/HelperFunctions/NASTD_ECoG_Connectivity_PlotGC_statp1top31_acrossTDs+seqs_alltones.m, lines 1–66 · score 0.52 · linear fit, PE Pred, slopes, electrode pairing, broadband, lobe
  25. [25] § Results › Changes in directed connectivity during sequence processing and sensory prediction ↔ AnalysisCode/Analysis_and_Plotting/Connectivity/NASTD_ECoG_Connectivity_Main_lk.m, lines 102–163 · score 0.52 · Granger causality, S2, ECoG, high gamma, lobe, parietal
  26. [26] § Methods › Experimental stimuli ↔ ParadigmCode/sfa_expt4_makeStim.m, lines 11–115 · score 0.51 · 220–880 Hz, 220 Hz, semitone, distance, log, pitches

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

MATLAB · 1,208 lines · 52 KB · no license · 3 matches

  1. %% Project: NASTD_ECoG
  2. %Compute Granger Causality (GC) directed connectivity measures to
  3. %determine:
  4. %1) the interplay between frontal and temporal prediction effect electrodes (time-wise GC in the HP2-LP30Hz band)
  5. %2) the interplay between frontal and temporal complex prediciton error (PE) effect eletrodes (frequency-wise GC in the high gamma band)
  6. %3) the interplay between frontal prediction and temporal PE effect electrodes
  7. %4) the interplay between global SHI and prediciton effect electrodes
  9. %% 0.1) Specify vars, paths, and setup fieldtrip
  10. addpath('/isilon/LFMI/VMdrive/Thomas/NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/')
  11. %Add project base path
  12. NASTD_ECoG_setVars
  13. paths_NASTD_ECoG = NASTD_ECoG_paths;
  14. % addpath(genpath(paths_NASTD_ECoG.BaseDir));
  15. addpath(genpath(paths_NASTD_ECoG.ScriptsDir));
  16. %Add project base and script dir
  18. %Determine subjects
  19. sub_list = vars.sub_list;
  21. %Load in file with individual preproc infos
  22. subs_PreProcSettings = NASTD_ECoG_Preproc_SubPreprocSettings;
  23. ToneDur_text = {'0.2' '0.4'};
  25. plot_poststepFigs = 0;
  26. save_poststepFigs = 1;
  28. %% 1) Select and plot electrodes for which GC will be analyzed
  30. param.pval_plotting = 0.01; %Pval thresh for plotting
  31. param.pval_FDR = 0.05;
  32. param.FDRcorrect = 0;
  33. param.plot_SubplotperTW = 1; %Common plot across TW
  34. param.ElecSelect = 'All'; %StimCorr, All
  35. param.SamplesTW = 25; %25 = 50ms TW
  36. param.Label_TW = num2str(round(param.SamplesTW/512,2));
  37. param.ToneIndex = 33;
  39. InputDataType = {'HP05toLP30Hz'};
  40. InputEffectType = {'PredEffect', 'ComplexPredErrEffect'};
  41. ToneDur_text = {'0.2' '0.4'};
  42. subs = sub_list(vars.validSubjs);
  44. %1.1 Highlight same-subject electrodes with selected p-value for either prediciton OR complex prediction error effect
  45. % for i_effect = 1:length(InputEffectType)
  46. % for i_inputData = 1:length(InputDataType)
  47. % NASTD_ECoG_Connectivity_PlotSignElec_AllSubTD...
  48. % (subs, ...
  49. % InputEffectType{i_effect}, InputDataType{i_inputData}, ToneDur_text, ...
  50. % param,...
  51. % save_poststepFigs, paths_NASTD_ECoG)
  52. % end
  53. % end
  55. % %1.2 Highlight same-subject electrodes with selected p-value
  56. % %with prediciton (for HP2toLP30Hz) AND complex prediction error (for HighGamma_LogAmp) effect
  57. % NASTD_ECoG_Connectivity_PlotSignElecPred2PE_AllSubTD...
  58. % (subs, ...
  59. % ToneDur_text, ...
  60. % param,...
  61. % save_poststepFigs, paths_NASTD_ECoG)
  63. % %1.3 Create output file specifiying selected electrodes
  64. subs = sub_list(vars.validSubjs);
  65. plot_poststepFigs = 0;
  67. param.pval_plotting = 0.05; %Pval thresh
  68. param.FDRcorrect = 0;
  69. param.pval_FDR = 0.05;
  70. param.ElecSelect = 'All'; %StimCorr, All
  71. InputDataType = {'HP05toLP30Hz'};%, 'HighGamma_LogAmp'};
  73. %Effect-based p-val thresholded electrode selection
  74. % SelElecs = NASTD_ECoG_Connectivity_ReadOutGCElecs_AllSubTD...
  75. % (subs, ...
  76. % InputDataType, ToneDur_text, ...
  77. % param, plot_poststepFigs, ...
  78. % paths_NASTD_ECoG);
  80. % %%Select all existing electrodes (independent of effect or threshold)
  81. % SelElecs = NASTD_ECoG_Connectivity_ReadOutAllElecs_AllSubTD...
  82. % (subs, ...
  83. % plot_poststepFigs, ...
  84. % paths_NASTD_ECoG);
  86. %1.5 Create Plot showing electrode connections that are the basis for GC calculation
  87. %Across subjects and TD, color-coded for region, sign-coded for prediction-effect-type.
  88. %Plot separately for each electrode-selection based on different prediction-effect-types.
  90. % NASTD_ECoG_Connectivity_PlotSignElecConnectionsforGC_AllSubTD...
  91. % (subs, ...
  92. % InputDataType, ToneDur_text, ...
  93. % param, plot_poststepFigs, ...
  94. % paths_NASTD_ECoG);
  95. %
  96. % NASTD_ECoG_Connectivity_PlotSignElecCon_AllSubTDLobes... %For Pred-Pred & PE-PE, all lobes together with optional connection lines
  97. % (subs, ...
  98. % InputDataType, ToneDur_text, ...
  99. % param, plot_poststepFigs, ...
  100. % paths_NASTD_ECoG);
  102. %% 2) Compute Granger Causality (GC) for specific electrode combinations
  104. %Hypotheses:
  105. %Spatial:
  106. %Frontal P -> Temporal P > Chance
  107. %Temporal PE -> Frontal PE > Chance (CAVE: few frontal PE)
  108. %Frontal P -> Temporal PE > Frontal P <- Temporal PE
  109. %Temporal:
  110. %P -> PE electrodes stronger in earlier TW compared to late TW
  111. %P -> PE electrodes stronger in earlier TW compared to P <- PE electrodes
  112. %Spectral:
  113. %P -> PE: low frequencies (alpha/beta)
  114. %P <- PE: high gamma
  115. %P -> P: low frequencies (alpha/beta)
  116. %PE -> PE: high gamma
  118. addpath('/isilon/LFMI/VMdrive/Thomas/toolboxes/mvgc_v1.3/');
  119. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.01uncorr.mat') %p < 0.01 thresh,uncorr elec selection
  120. load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.05uncorr.mat') %p < 0.05 thresh,uncorr elec selection
  121. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.025FDRcorr.mat') %p < 0.025 thresh, FDRcorr elec selection
  122. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_AllElecs.mat') % no thresh, all elec selection
  123. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.05uncorr_HighGamma.mat') %p < 0.05 thresh,uncorr elec selection
  124. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.01uncorr_HighGamma.mat') %p < 0.05 thresh,uncorr elec selection
  126. % Computing how many significant electrodes per lobe (frontal, parietal, and temporal) and prediction vs. PE effect for Table S2
  127. % Define keywords for each lobe
  128. % frontalKeywords = ["AntPFC", "PrecentralG", "IFG"];
  129. % parietalKeywords = ["SupParLob", "PostcentralG", "SupramarginalG"];
  130. % temporalKeywords = ["VentralT", "STG", "MTG"];
  131. %
  132. % % Extract AnatCat Label column
  133. % predLabels = SelElecs.PredEffect.("AnatCat Label");
  134. % predLabelsStr = string(predLabels);
  135. % predLabelsStr(contains(predLabelsStr, "OccipitalL")) = []; % Remove entries containing "OccipitalL"
  136. % [uniqueLabels1, ~, idx1] = unique(predLabelsStr);
  137. % counts1 = accumarray(idx1, 1);
  138. % PredlabelCountsTable = table(uniqueLabels1, counts1, 'VariableNames', {'Label', 'Count'});
  139. %
  140. % peLabels = SelElecs.PEEffect.("AnatCat Label");
  141. % peLabelsStr = string(peLabels);
  142. % peLabelsStr(contains(peLabelsStr, "OccipitalL")) = []; % Remove entries containing "OccipitalL"
  143. % [uniqueLabels2, ~, idx2] = unique(peLabelsStr);
  144. % counts2 = accumarray(idx2, 1);
  145. % PElabelCountsTable = table(uniqueLabels2, counts2, 'VariableNames', {'Label', 'Count'});
  146. %
  147. % countLobes = @(data, keywords) sum(any(contains(data, keywords), 2));
  148. %
  149. % % Count electrodes per lobe for each effect using the strict word-boundary matching
  150. % frontalPredCount = countLobes(predLabelsStr, frontalKeywords);
  151. % parietalPredCount = countLobes(predLabelsStr, parietalKeywords);
  152. % temporalPredCount = countLobes(predLabelsStr, temporalKeywords);
  153. %
  154. % frontalPECount = countLobes(peLabelsStr, frontalKeywords);
  155. % parietalPECount = countLobes(peLabelsStr, parietalKeywords);
  156. % temporalPECount = countLobes(peLabelsStr, temporalKeywords);
  157. %
  158. % % Create the output table
  159. % LobeSummaryTable = table(["Frontal"; "Parietal"; "Temporal"], ...
  160. % [frontalPredCount; parietalPredCount; temporalPredCount], ...
  161. % [frontalPECount; parietalPECount; temporalPECount], ...
  162. % 'VariableNames', {'Lobe', 'PredCount', 'PECount'});
  163. % writetable(LobeSummaryTable, ['/isilon/LFMI/VMdrive/Lua/NASTD/Figures/ElectrodeCountsPerLobe.csv']);
  165. %% Compute GC
  166. param.GC = [];
  167. param.GC.fs = 512;
  168. param.GC.downsample = 0; %Cave: Downsampling lead to problems with tone sample selection
  169. if param.GC.downsample == 0
  170. param.GC.newfs = param.GC.fs;
  171. else
  172. param.GC.newfs = param.GC.fs/param.GC.downsample;
  173. end
  174. param.GC.nvars = 2;
  175. param.GC.regmode = 'OLS'; % VAR model estimation regression mode ('OLS', 'LWR' or empty for default
  176. param.GC.maxmorder = 50; % maximum model order for model order estimation, rule of thumb = number of samples per input data snippet, but not > 100
  177. % param.GC.morder = 'AIC'; % model order to use ('actual', 'AIC', 'BIC' or supplied numerical value)
  179. param.GC.tstat = 'F'; % statistical test for MVGC: 'F' for Granger's F-test (default) or 'chi2' for Geweke's chi2 test
  180. param.GC.alpha = 0.05; % significance level for significance test
  181. param.GC.mhtc = 'FDRD'; % multiple hypothesis test correction (see routine 'significance')
  183. param.GC.ElecPairEffect = {'Pred_Pred','PE_PE','Pred_PE','PE_Pred'};
  184. %param.GC.ElecPairEffect = {'Pred_Pred', 'PE_PE'};
  185. %param.GC.ElecPairEffect = {'Pred_PE', 'PE_Pred'};
  186. param.GC.ElecPairRegion = 'AllRegions';
  187. param.GC.InputDataType = {'Broadband'}; %{'Broadband', 'HP05toLP30Hz', 'HighGamma_LogAmp'};
  189. % param.GC.tone_aggregation = {32, 33, 34};
  190. % param.GC.tone_aggregation = {31, 33, 34};
  191. %param.GC.tone_aggregation = {1, 6, 11, 16, 21, 26, 31};
  192. %param.GC.tone_aggregation = {2, 7, 12, 17, 22, 27, 32};
  193. % param.GC.tone_aggregation = {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};
  194. % param.GC.tone_aggregation = {30, 31, 32};
  195. %param.GC.tone_aggregation = {30, 31, 32, 33, 34};
  196. param.GC.tone_aggregation = {34};
  198. param.GC.tone_aggregation_label = cellfun(@(c)[c],param.GC.tone_aggregation);
  199. param.GC.tone_aggregation_label = num2str((param.GC.tone_aggregation_label),'_%d');
  200. param.GC.tone_aggregation_label = erase(param.GC.tone_aggregation_label, ' ');
  202. param.GC.epochdur_ms = 'full'; %'full', 200, 100
  203. if strcmp(param.GC.epochdur_ms, 'full')
  204. param.GC.epochdur_ms_label = 'fullTW';
  205. else
  206. param.GC.epochdur_ms_label = [num2str(param.GC.epochdur_ms) 'msTW'];
  207. end
  209. % Across-trial GC estimates
  210. % parfor i_sub = vars.validSubjs
  211. %
  212. % tic
  213. % GCdata{i_sub} = ... %Compute GC for aggregated tones
  214. % NASTD_ECoG_Connectivity_CalculateGC_aggrtone ...
  215. % (sub_list, i_sub, ...
  216. % ToneDur_text, SelElecs, ...
  217. % param, paths_NASTD_ECoG);
  218. % GCdata{i_sub} = ... %Compute GC for each single tone
  219. % NASTD_ECoG_Connectivity_CalculateGC_pertone ...
  220. % (sub_list, i_sub, ...
  221. % ToneDur_text, SelElecs, ...
  222. % param, paths_NASTD_ECoG);
  223. % disp(['-- GC computation for sub: ' sub_list{i_sub} ' finished after ' num2str(round(toc/60),2) ' min --'])
  224. %
  225. % end
  227. %save output
  228. % path_outputdata = ([paths_NASTD_ECoG.ECoGdata_Connectivity 'GC/']);
  229. % if (~exist(path_outputdata, 'dir')); mkdir(path_outputdata); end
  230. % label_outputdata = ['GCdata_n' num2str(length(vars.validSubjs)) '_' ...
  231. % param.GC.ElecPairRegion '_' ...
  232. % param.GC.InputDataType{1} param.GC.tone_aggregation_label '_' ...
  233. % param.GC.epochdur_ms_label '_ensemnorm'];
  234. % %save([path_outputdata label_outputdata], 'GCdata','-v7.3')
  235. % load([path_outputdata label_outputdata], 'GCdata')
  237. % Across-trial GC estimates for selected trial IDs
  238. FTPL_ratings = load([paths_NASTD_ECoG.Analysis_Behavior 'FTPLratings/Allsub_n8/Data/', 'Trialwise_FTPL_allsubs.mat']);
  239. FTPL_ratings = FTPL_ratings.Trialwise_FTPL;
  240. rows_all = strcmp(FTPL_ratings.ToneDur, 'all'); %Take ratings across both TD conditions
  241. FTPL_all = FTPL_ratings(rows_all, :);
  242. sub_list_FTPL = sub_list(2:9);
  244. % Averaging over trialwise data for low-FTPL trials and for high-FTPL
  245. % trials
  246. low_FTPL_all = {};
  247. high_FTPL_all = {};
  248. for i_sub = 1:length(sub_list_FTPL)
  249. current_subID = sub_list_FTPL(i_sub);
  251. % Find all rows in FTPL_all for this subject
  252. subj_rows = strcmp(FTPL_all.SubID, current_subID);
  254. % Get the FTPL ratings for this subject
  255. ratings_subj = FTPL_all.FTPLrating(subj_rows);
  256. median_subj = median(ratings_subj, 'omitnan');
  258. % Create logical indices for low and high FTPL relative to
  259. % subject-specific median of ratings
  260. low_idx = ratings_subj < median_subj;
  261. low_idx = FTPL_all.TrialNum(low_idx); % TrialIDs for low FTPL
  262. low_FTPL_all{i_sub} = low_idx;
  264. high_idx = ratings_subj > median_subj;
  265. high_idx = FTPL_all.TrialNum(high_idx);
  266. high_FTPL_all{i_sub} = high_idx;
  267. end
  269. path_outputdata = ([paths_NASTD_ECoG.ECoGdata_Connectivity 'GC/']);
  270. % parfor i_sub = vars.validSubjs(2:9)
  271. % tic
  272. % GCdata_lowFTPL{i_sub} = ... %Compute GC for aggregated tones
  273. % NASTD_ECoG_Connectivity_CalculateGC_aggrtone_SelTrials ...
  274. % (sub_list, i_sub, ...
  275. % ToneDur_text, SelElecs, ...
  276. % low_FTPL_all{i_sub-1}, ...
  277. % param, paths_NASTD_ECoG);
  278. % disp(['-- GC computation for sub: ' sub_list{i_sub} ' for LOW FTPL finished after ' num2str(round(toc/60),2) ' min --'])
  279. % GCdata_highFTPL{i_sub} = ... %Compute GC for aggregated tones
  280. % NASTD_ECoG_Connectivity_CalculateGC_aggrtone_SelTrials ...
  281. % (sub_list, i_sub, ...
  282. % ToneDur_text, SelElecs, ...
  283. % high_FTPL_all{i_sub-1}, ...
  284. % param, paths_NASTD_ECoG);
  285. % disp(['-- GC computation for sub: ' sub_list{i_sub} ' for LOW FTPL finished after ' num2str(round(toc/60),2) ' min --'])
  286. % end
  287. label_outputdata1 = ['GCdata_n' num2str(length(vars.validSubjs)) '_' ...
  288. param.GC.ElecPairRegion '_' ...
  289. param.GC.InputDataType{1} param.GC.tone_aggregation_label '_' ...
  290. param.GC.epochdur_ms_label '_lowFTPL'];
  291. %save([path_outputdata label_outputdata1], 'GCdata_lowFTPL','-v7.3')
  292. GCdata_lowFTPL=load([path_outputdata label_outputdata1]);
  294. label_outputdata2 = ['GCdata_n' num2str(length(vars.validSubjs)) '_' ...
  295. param.GC.ElecPairRegion '_' ...
  296. param.GC.InputDataType{1} param.GC.tone_aggregation_label '_' ...
  297. param.GC.epochdur_ms_label '_highFTPL'];
  298. %save([path_outputdata label_outputdata2], 'GCdata_highFTPL','-v7.3')
  299. GCdata_highFTPL = load([path_outputdata label_outputdata2]);
  301. % Trialwise GC estimates
  302. % parfor i_sub = vars.validSubjs
  303. %
  304. % tic
  305. % GCdata{i_sub} = ... %Compute GC for each single tone and for each trial; used for control analysis checking how GC values differ for different trial-wise FTPL ratings; only run for tone 34
  306. % NASTD_ECoG_Connectivity_CalculateGC_aggrtone_trialwise ...
  307. % (sub_list, i_sub, ...
  308. % ToneDur_text, SelElecs, ...
  309. % param, paths_NASTD_ECoG);
  310. % disp(['-- GC computation for sub: ' sub_list{i_sub} ' finished after ' num2str(round(toc/60),2) ' min --'])
  311. %
  312. % end
  313. %save output
  314. % path_outputdata = ([paths_NASTD_ECoG.ECoGdata_Connectivity 'GC/']);
  315. % if (~exist(path_outputdata, 'dir')); mkdir(path_outputdata); end
  316. % label_outputdata = ['GCdata_n' num2str(length(vars.validSubjs)) '_' ...
  317. % param.GC.ElecPairRegion '_' ...
  318. % param.GC.InputDataType{1} param.GC.tone_aggregation_label '_' ...
  319. % param.GC.epochdur_ms_label '_trialwise_ensemnorm'];
  320. % %save([path_outputdata label_outputdata], 'GCdata','-v7.3')
  321. % load([path_outputdata label_outputdata], 'GCdata')
  323. %% Compare GC values for low vs. high FTPL ratings for parietal -> frontal during p34 using trialwise data
  325. %First make a GCdata for low FTPL ratings, average across trials,NASTD_ECoG_Connectivity_stat save GC.
  326. %Do the same for high FTP ratings, average across trials, save GC.
  327. %Pass both averaged GC through anatomical averaging script (NASTD_ECoG_Connectivity_PlotGC_statp32p33p34.m in Thomas' folder)
  328. % This will output 10 (source anat regions) x 10 (target anat regions) x 2
  329. % (TDs) arrays for the low-FTPL-GC and the high-FTPL-GC. Then we can select
  330. % parietal to frontal by indexing source anat region = 7 and target anat
  331. % region = 1 for each of them, and do the stats.
  332. % GCdata_FTPL = GCdata(2:9);
  333. % numSubjects = length(GCdata_FTPL);
  334. % FTPL_ratings = load([paths_NASTD_ECoG.Analysis_Behavior 'FTPLratings/Allsub_n8/Data/', 'Trialwise_FTPL_allsubs.mat']);
  335. % FTPL_ratings = FTPL_ratings.Trialwise_FTPL;
  336. % rows_all = strcmp(FTPL_ratings.ToneDur, 'all'); %Take ratings across both TD conditions
  337. % FTPL_all = FTPL_ratings(rows_all, :);
  338. % all_trialIDs = cell(numSubjects, 1);
  339. % for i_sub = 1:numSubjects
  340. % % Combine trial IDs across tone durations (1x2 cell)
  341. % all_trialIDs{i_sub} = [GCdata{i_sub}.info.trialIDs{1}; GCdata{i_sub}.info.trialIDs{2}];
  342. % end
  343. % sub_list_FTPL = sub_list(2:9);
  344. %
  345. % % Averaging over trialwise data for low-FTPL trials and for high-FTPL
  346. % % trials
  347. % for i_sub = 1:numSubjects
  348. % current_subID = sub_list_FTPL(i_sub);
  349. %
  350. % % Find all rows in FTPL_all for this subject
  351. % subj_rows = strcmp(FTPL_all.SubID, current_subID);
  352. %
  353. % % Get the FTPL ratings for this subject
  354. % ratings_subj = FTPL_all.FTPLrating(subj_rows);
  355. % median_subj = median(ratings_subj, 'omitnan');
  356. % trialIDs_subj = all_trialIDs{i_sub};
  357. %
  358. % % Create logical indices for low and high FTPL relative to
  359. % % subject-specific median of ratings
  360. % low_idx = ratings_subj < median_subj;
  361. % low_idx = FTPL_all.TrialNum(low_idx); % TrialIDs for low FTPL
  362. %
  363. % high_idx = ratings_subj > median_subj;
  364. % high_idx = FTPL_all.TrialNum(high_idx);
  365. %
  366. % % Find the Trial IDs for the GC data
  367. % trialIDs_GC_TD1 = GCdata_FTPL{i_sub}.info.trialIDs{1}; %Trial IDs of GC data for TD1
  368. % trialIDs_GC_TD2 = GCdata_FTPL{i_sub}.info.trialIDs{2}; %Trial IDs for GC data for TD2
  369. %
  370. % % Identify corresponding trials in GC data for low and high FTPL
  371. % % ratings
  372. % [~,match_idx_low_TD1] = ismember(low_idx, trialIDs_GC_TD1); % Finds corresponding trials in the GC data for TD1
  373. % idx_low_in_GC_TD1 = match_idx_low_TD1(match_idx_low_TD1 > 0);
  374. % [~, match_idx_low_TD2] = ismember(low_idx, trialIDs_GC_TD2); % Finds corresponding trials in the GC data for TD2
  375. % idx_low_in_GC_TD2 = match_idx_low_TD2(match_idx_low_TD2 > 0);
  376. %
  377. % [~, match_idx_high_TD1] = ismember(high_idx, trialIDs_GC_TD1); % Finds corresponding trials in the GC data for TD1
  378. % idx_high_in_GC_TD1 = match_idx_high_TD1(match_idx_high_TD1 > 0);
  379. % [~, match_idx_high_TD2] = ismember(high_idx, trialIDs_GC_TD2); % Finds corresponding trials in the GC data for TD2
  380. % idx_high_in_GC_TD2 = match_idx_high_TD2(match_idx_high_TD2 > 0);
  381. %
  382. % for i_effect = 1:length(GCdata_FTPL{i_sub}.temporalGC)
  383. % for i_TD = 1:length(GCdata_FTPL{i_sub}.temporalGC{i_effect})
  384. % if i_TD == 1
  385. % idx_low_in_GC = idx_low_in_GC_TD1;
  386. % idx_high_in_GC = idx_high_in_GC_TD1;
  387. % else
  388. % idx_low_in_GC = idx_low_in_GC_TD2;
  389. % idx_high_in_GC = idx_high_in_GC_TD2;
  390. % end
  391. %
  392. % % === TEMPORAL GC ===
  393. % temp_struct = GCdata_FTPL{i_sub}.temporalGC{i_effect}{i_TD};
  394. % GCdata_average_lowFTPL{i_sub}.temporalGC{i_effect}{i_TD}.source2target = ...
  395. % mean(temp_struct.source2target(:,:,:,idx_low_in_GC), 4, 'omitnan');
  396. % GCdata_average_lowFTPL{i_sub}.temporalGC{i_effect}{i_TD}.target2source = ...
  397. % mean(temp_struct.target2source(:,:,:,idx_low_in_GC), 4, 'omitnan');
  398. %
  399. % GCdata_average_highFTPL{i_sub}.temporalGC{i_effect}{i_TD}.source2target = ...
  400. % mean(temp_struct.source2target(:,:,:,idx_high_in_GC), 4, 'omitnan');
  401. % GCdata_average_highFTPL{i_sub}.temporalGC{i_effect}{i_TD}.target2source = ...
  402. % mean(temp_struct.target2source(:,:,:,idx_high_in_GC), 4, 'omitnan');
  403. %
  404. % % === PVAL TEMPORAL GC ===
  405. % pval_struct = GCdata_FTPL{i_sub}.pval_temporalGC{i_effect}{i_TD};
  406. % GCdata_average_lowFTPL{i_sub}.pval_temporalGC{i_effect}{i_TD}.source2target = ...
  407. % mean(pval_struct.source2target(:,:,:,idx_low_in_GC), 4, 'omitnan');
  408. % GCdata_average_lowFTPL{i_sub}.pval_temporalGC{i_effect}{i_TD}.target2source = ...
  409. % mean(pval_struct.target2source(:,:,:,idx_low_in_GC), 4, 'omitnan');
  410. %
  411. % GCdata_average_highFTPL{i_sub}.pval_temporalGC{i_effect}{i_TD}.source2target = ...
  412. % mean(pval_struct.source2target(:,:,:,idx_high_in_GC), 4, 'omitnan');
  413. % GCdata_average_highFTPL{i_sub}.pval_temporalGC{i_effect}{i_TD}.target2source = ...
  414. % mean(pval_struct.target2source(:,:,:,idx_high_in_GC), 4, 'omitnan');
  415. %
  416. % % === SPECTRAL GC ===
  417. % spec_struct = GCdata_FTPL{i_sub}.spectralGC{i_effect}{i_TD};
  418. % GCdata_average_lowFTPL{i_sub}.spectralGC{i_effect}{i_TD}.source2target = ...
  419. % mean(spec_struct.source2target(:,:,:,:,idx_low_in_GC), 5, 'omitnan');
  420. % GCdata_average_lowFTPL{i_sub}.spectralGC{i_effect}{i_TD}.target2source = ...
  421. % mean(spec_struct.target2source(:,:,:,:,idx_low_in_GC), 5, 'omitnan');
  422. %
  423. % GCdata_average_highFTPL{i_sub}.spectralGC{i_effect}{i_TD}.source2target = ...
  424. % mean(spec_struct.source2target(:,:,:,:,idx_high_in_GC), 5, 'omitnan');
  425. % GCdata_average_highFTPL{i_sub}.spectralGC{i_effect}{i_TD}.target2source = ...
  426. % mean(spec_struct.target2source(:,:,:,:,idx_high_in_GC), 5, 'omitnan');
  427. % end
  428. % end
  429. % GCdata_average_lowFTPL{i_sub}.info = GCdata_FTPL{i_sub}.info;
  430. % GCdata_average_lowFTPL{i_sub}.label_allelecs = GCdata_FTPL{i_sub}.label_allelecs;
  431. % GCdata_average_lowFTPL{i_sub}.ind_pairedelecs_from = GCdata_FTPL{i_sub}.ind_pairedelecs_from;
  432. % GCdata_average_lowFTPL{i_sub}.ind_pairedelecs_to = GCdata_FTPL{i_sub}.ind_pairedelecs_to;
  433. %
  434. % GCdata_average_highFTPL{i_sub}.info = GCdata_FTPL{i_sub}.info;
  435. % GCdata_average_highFTPL{i_sub}.label_allelecs = GCdata_FTPL{i_sub}.label_allelecs;
  436. % GCdata_average_highFTPL{i_sub}.ind_pairedelecs_from = GCdata_FTPL{i_sub}.ind_pairedelecs_from;
  437. % GCdata_average_highFTPL{i_sub}.ind_pairedelecs_to = GCdata_FTPL{i_sub}.ind_pairedelecs_to;
  438. % end
  440. % AnatReg_CatLabels = ... %Select which anatomical regions should be plotted
  441. % {'AntPFC'; ...
  442. % 'VentralT'; ...
  443. % 'SupParLob'};
  444. %
  445. % Gavg_tempGC_peranatreg_lowFTPL = NASTD_ECoG_Connectivity_ObtainGCByAnatRegions ...
  446. % (sub_list_FTPL, ...
  447. % GCdata_average_lowFTPL, ...
  448. % ToneDur_text, ...
  449. % SelElecs, AnatReg_CatLabels, ...
  450. % save_poststepFigs, ...
  451. % param, paths_NASTD_ECoG); %This will have the format, for each field, of i_effect cells, with format nSourceElecs x nTargetElecs x nTD
  452. %
  453. % Gavg_tempGC_peranatreg_highFTPL = NASTD_ECoG_Connectivity_ObtainGCByAnatRegions ...
  454. % (sub_list_FTPL, ...
  455. % GCdata_average_highFTPL, ...
  456. % ToneDur_text, ...
  457. % SelElecs, AnatReg_CatLabels, ...
  458. % save_poststepFigs, ...
  459. % param, paths_NASTD_ECoG); %This will have the format, for each field, of i_effect cells, with format nSourceElecs x nTargetElecs x nTD
  462. % Compute GC low vs high FTPL for Parietal > Frontal
  463. %
  464. % index_of_interest_source = 7; %Parietal
  465. % index_of_interest_target = 1; %Frontal
  466. %
  467. % GC_Par_to_Front_lowFTPL = nan(numSubjects, 1); % one entry per subject
  468. % GC_Par_to_Front_highFTPL = nan(numSubjects, 1); % one entry per subject
  469. %
  470. % GC_Par_to_Front_lowFTPL_all = {}; % Pool across 4 different effect types, average across TDs
  471. % GC_Par_to_Front_highFTPL_all = {};
  472. %
  473. % for i_sub = 1:numSubjects
  474. % all_GC_rows_low = [];
  475. % all_GC_rows_high = [];
  476. %
  477. % for i_effect = 1:length(GCdata_FTPL{i_sub}.temporalGC)
  478. % for i_TD = 1:length(GCdata_FTPL{i_sub}.temporalGC{i_effect})
  479. % % Get data and subject indices for this effect and TD
  480. % GCvals_low = Gavg_tempGC_peranatreg_lowFTPL.source2target{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  481. % subjIDs_low = Gavg_tempGC_peranatreg_lowFTPL.subject_index{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  482. %
  483. % if ~isempty(GCvals_low) && ~isempty(subjIDs_low)
  484. % % Find rows belonging to this subject
  485. % rows_low = find(subjIDs_low == i_sub);
  486. % if ~isempty(rows_low)
  487. % all_GC_rows_low = [all_GC_rows_low; GCvals_low(rows_low, :)];
  488. % end
  489. % end
  490. %
  491. % % Get data and subject indices for this effect and TD
  492. % GCvals_high = Gavg_tempGC_peranatreg_highFTPL.source2target{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  493. % subjIDs_high = Gavg_tempGC_peranatreg_highFTPL.subject_index{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  494. %
  495. % if ~isempty(GCvals_high) && ~isempty(subjIDs_high)
  496. % % Find rows belonging to this subject
  497. % rows_high = find(subjIDs_high == i_sub);
  498. % if ~isempty(rows_high)
  499. % all_GC_rows_high = [all_GC_rows_high; GCvals_high(rows_high, :)];
  500. % end
  501. % end
  502. % end
  503. % end
  504. %
  505. % if ~isempty(all_GC_rows_low)
  506. % GC_Par_to_Front_lowFTPL(i_sub,1) = mean(all_GC_rows_low, 1, 'omitnan');
  507. % else
  508. % GC_Par_to_Front_lowFTPL(i_sub,1) = NaN; % or [] if you prefer
  509. % end
  510. %
  511. % if ~isempty(all_GC_rows_high)
  512. % GC_Par_to_Front_highFTPL(i_sub,1) = mean(all_GC_rows_high, 1, 'omitnan');
  513. % else
  514. % GC_Par_to_Front_highFTPL(i_sub,1) = NaN; % or [] if you prefer
  515. % end
  516. % end
  517. %
  518. % % Do stats
  519. % [~, pval, ~, stats] = ttest(GC_Par_to_Front_lowFTPL, GC_Par_to_Front_highFTPL);
  520. % [pval, h, stats] = signrank(GC_Par_to_Front_lowFTPL, GC_Par_to_Front_highFTPL);
  521. %
  522. % % Plot
  523. % % Combine data
  524. % % all_data = [GC_Par_to_Front_lowFTPL(:); GC_Par_to_Front_highFTPL(:)];
  525. % % group = [ones(length(GC_Par_to_Front_lowFTPL),1); 2*ones(length(GC_Par_to_Front_highFTPL),1)];
  526. % all_data = [GC_Par_to_Front_lowFTPL(:); mean(GC_Par_to_Front_lowFTPL); GC_Par_to_Front_highFTPL(:); mean(GC_Par_to_Front_highFTPL)];
  527. % group = [ones(length(GC_Par_to_Front_lowFTPL)+1,1); 2*ones(length(GC_Par_to_Front_highFTPL)+1,1)];
  528. %
  529. % % Create boxplot
  530. % figure; hold on;
  531. % h = boxplot(all_data, group, 'Labels', {'Low FTPL', 'High FTPL'}, ...
  532. % 'Colors', 'k', 'Widths', 0.5);
  533. %
  534. % % Change box colors
  535. % boxColors = [0.2 0.6 1; 1 0.4 0.4]; % Blue, Red
  536. % boxes = findobj(gca, 'Tag', 'Box');
  537. % for j = 1:length(boxes)
  538. % patch(get(boxes(j), 'XData'), get(boxes(j), 'YData'), ...
  539. % boxColors(3-j,:), 'FaceAlpha', 0.5, 'EdgeColor', 'none');
  540. % end
  541. %
  542. % % Overlay paired data
  543. % for i = 1:length(GC_Par_to_Front_lowFTPL)
  544. % plot([1, 2], [GC_Par_to_Front_lowFTPL(i), GC_Par_to_Front_highFTPL(i)], '-o', ...
  545. % 'Color', [0.6 0.6 0.6], ...
  546. % 'MarkerFaceColor', 'k', ...
  547. % 'MarkerEdgeColor', 'none', ...
  548. % 'LineWidth', 1.2);
  549. % end
  550. %
  551. % plot([1, 2], [mean(GC_Par_to_Front_lowFTPL), mean(GC_Par_to_Front_highFTPL)], '-o', ...
  552. % 'Color', [0.6 0.6 0.6], ...
  553. % 'MarkerFaceColor', 'k', ...
  554. % 'MarkerEdgeColor', 'none', ...
  555. % 'LineWidth', 1.2);
  556. %
  557. % % Significance annotation
  558. % [~, pval] = ttest(GC_Par_to_Front_lowFTPL, GC_Par_to_Front_highFTPL);
  559. % y_max = max([GC_Par_to_Front_lowFTPL(:); GC_Par_to_Front_highFTPL(:)]) + 0.01;
  560. % plot([1, 2], [y_max, y_max], 'k', 'LineWidth', 1.5)
  561. %
  562. % if pval < 0.001
  563. % sig_label = '***';
  564. % elseif pval < 0.01
  565. % sig_label = '**';
  566. % elseif pval < 0.05
  567. % sig_label = '*';
  568. % else
  569. % sig_label = 'n.s.';
  570. % end
  571. % text(1.5, y_max + 0.005, sig_label, ...
  572. % 'FontSize', 16, ...
  573. % 'FontWeight', 'bold', ...
  574. % 'HorizontalAlignment', 'center');
  575. %
  576. % % Formatting
  577. % title('GC during p34 Parietal > Frontal');
  578. % ylabel('Average GC');
  579. % ylim([min([GC_Par_to_Front_lowFTPL; GC_Par_to_Front_highFTPL]) - 0.01, y_max + 0.02]);
  580. % set(gca, 'FontSize', 12);
  581. % box on;
  582. %
  583. % filename = ['boxplot_compare_GC_lowhigh_FTPL_Par_to_Front.png'];
  584. % figfile = ['/isilon/LFMI/VMdrive/Lua/NASTD/Figures/' filename];
  585. % saveas(gcf, figfile, 'png');
  586. %
  587. %
  588. % %% Computing stats for FTPL low vs high GC comparison across pairs from different effect-type pairings using TRIALWISE data
  589. % % Compute GC low vs high FTPL for Parietal > Frontal
  590. %
  591. % index_of_interest_source = 7; %Parietal
  592. % index_of_interest_target = 1; %Frontal
  593. % effectTypeNames = {'Pred_Pred','PE_PE','Pred_PE','PE_Pred'};
  594. %
  595. % GC_Par_to_Front_HighMinusLow = struct(); % Final output
  596. %
  597. % for i_effect = 1:length(GCdata_FTPL{1}.temporalGC)
  598. % effect_field = effectTypeNames{i_effect};
  599. % for i_TD = 1:length(GCdata_FTPL{1}.temporalGC{1})
  600. % if ~isfield(GC_Par_to_Front_HighMinusLow, effect_field)
  601. % GC_Par_to_Front_HighMinusLow.(effect_field) = struct();
  602. % end
  603. % if ~isfield(GC_Par_to_Front_HighMinusLow.(effect_field), sprintf('TD_%d', i_TD))
  604. % GC_Par_to_Front_HighMinusLow.(effect_field).(sprintf('TD_%d', i_TD)) = [];
  605. % end
  606. %
  607. % pooled_diffs = []; % temporary holder for this TD across all subjects
  608. %
  609. % for i_sub = 1:numSubjects
  610. % % LOW
  611. % GCvals_low = Gavg_tempGC_peranatreg_lowFTPL.source2target{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  612. % subjIDs_low = Gavg_tempGC_peranatreg_lowFTPL.subject_index{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  613. %
  614. % if ~isempty(GCvals_low) && ~isempty(subjIDs_low)
  615. % rows_low = (subjIDs_low == i_sub);
  616. % GC_low = GCvals_low(rows_low, :);
  617. % else
  618. % GC_low = NaN;
  619. % end
  620. %
  621. % % HIGH
  622. % GCvals_high = Gavg_tempGC_peranatreg_highFTPL.source2target{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  623. % subjIDs_high = Gavg_tempGC_peranatreg_highFTPL.subject_index{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  624. %
  625. % if ~isempty(GCvals_high) && ~isempty(subjIDs_high)
  626. % rows_high = (subjIDs_high == i_sub);
  627. % GC_high = GCvals_high(rows_high, :);
  628. % else
  629. % GC_high = NaN;
  630. % end
  631. %
  632. % if ~isempty(GC_low) && ~isempty(GC_high)
  633. % n_rows = min(size(GC_low,1), size(GC_high,1));
  634. % GC_diff = GC_high(1:n_rows, :) - GC_low(1:n_rows, :); % pairwise subtraction
  635. % pooled_diffs = [pooled_diffs; GC_diff]; % append
  636. % end
  637. % end
  638. % GC_Par_to_Front_HighMinusLow.(effect_field).(sprintf('TD_%d', i_TD)) = pooled_diffs;
  639. % end
  640. % end
  641. %
  642. % % Average across TDs
  643. % GC_all = cell(1, length(effectTypeNames));
  644. % pvals = zeros(1, length(effectTypeNames));
  645. % numEffects = numel(effectTypeNames);
  646. %
  647. % % Prepare data
  648. % for i = 1:length(effectTypeNames)
  649. % data_TD1 = GC_Par_to_Front_HighMinusLow.(effectTypeNames{i}).TD_1;
  650. % data_TD2 = GC_Par_to_Front_HighMinusLow.(effectTypeNames{i}).TD_2;
  651. %
  652. % % Average across TDs
  653. % avgData = mean([data_TD1(:), data_TD2(:)], 2, 'omitnan');
  654. % GC_all{i} = avgData;
  655. %
  656. % % t-test against zero
  657. % [~, p] = ttest(avgData);
  658. % pvals(i) = p;
  659. % end
  660. %
  661. % % Plot
  662. % effectNames_plot = {'Pred --> Pred', 'PE --> PE', 'Pred --> PE', 'PE --> Pred'};
  663. % figure; hold on; box on;
  664. % colors = lines(numEffects); % distinct colors for each cloud
  665. % rng(0); % for reproducible jitter
  666. %
  667. % for i = 1:numEffects
  668. % x = i + (rand(size(GC_all{i})) - 0.5) * 0.4; % add jitter
  669. % scatter(x, GC_all{i}, 25, ...
  670. % 'MarkerFaceColor', colors(i,:), ...
  671. % 'MarkerEdgeColor', 'k', ...
  672. % 'MarkerFaceAlpha', 0.6, ...
  673. % 'MarkerEdgeAlpha', 0.6);
  674. % end
  675. %
  676. % % Formatting
  677. % xlim([0.5, numEffects + 0.5]);
  678. % xticks(1:numEffects);
  679. % xticklabels(effectNames_plot);
  680. % xtickangle(20);
  681. % ylabel('GC Difference (High FTPL − Low FTPL)');
  682. % yline(0, '--k', 'LineWidth', 1);
  683. %
  684. % % Add exact p-values above each scatter cloud
  685. % yl = ylim;
  686. % extra_space = 0.1 * range(yl); % 10% extra space
  687. % ylim([yl(1) + 2 * extra_space, yl(2) + extra_space]);
  688. %
  689. % % Update y-limit variable for use in placing the p-values
  690. % yl = ylim;
  691. % text_y = yl(2) - 0.05 * range(yl);
  692. % for i = 1:numEffects
  693. % text(i, text_y, sprintf('p = %.3g', pvals(i)), ...
  694. % 'HorizontalAlignment', 'center', ...
  695. % 'FontSize', 11, 'FontWeight', 'bold');
  696. % end
  697. %
  698. % % Save figure
  699. % set(gcf, 'Color', 'w');
  700. %
  701. %
  702. % filename = ['boxplot_compare_GC_lowhigh_FTPL_Par_to_Front_allpairs.png'];
  703. % figfile = ['/isilon/LFMI/VMdrive/Lua/NASTD/Figures/' filename];
  704. % saveas(gcf, figfile, 'png');
  705. %
  707. %% Computing stats for FTPL low vs high GC comparison across pairs from different effect-type pairings using AVERAGED data
  708. % Compute GC low vs high FTPL for Parietal > Frontal
  709. AnatReg_CatLabels = ... %Select which anatomical regions should be plotted
  710. {'AntPFC'; ...
  711. 'VentralT'; ...
  712. 'SupParLob'};
  713. sub_list_FTPL = sub_list(2:9);
  715. Gavg_tempGC_peranatreg_lowFTPL = NASTD_ECoG_Connectivity_ObtainGCByAnatRegions ...
  716. (sub_list_FTPL, ...
  717. GCdata_lowFTPL.GCdata_lowFTPL(2:9), ...
  718. ToneDur_text, ...
  719. SelElecs, AnatReg_CatLabels, ...
  720. save_poststepFigs, ...
  721. param, paths_NASTD_ECoG); %This will have the format, for each field, of i_effect cells, with format nSourceElecs x nTargetElecs x nTD
  723. Gavg_tempGC_peranatreg_highFTPL = NASTD_ECoG_Connectivity_ObtainGCByAnatRegions ...
  724. (sub_list_FTPL, ...
  725. GCdata_highFTPL.GCdata_highFTPL(2:9), ...
  726. ToneDur_text, ...
  727. SelElecs, AnatReg_CatLabels, ...
  728. save_poststepFigs, ...
  729. param, paths_NASTD_ECoG); %This will have the format, for each field, of i_effect cells, with format nSourceElecs x nTargetElecs x nTD
  731. index_of_interest_source = 7; %Parietal
  732. index_of_interest_target = 1; %Frontal
  733. effectTypeNames = {'Pred_Pred','PE_PE','Pred_PE','PE_Pred'};
  735. GC_Par_to_Front_HighMinusLow = struct(); % Final output
  737. for i_effect = 1:length(GCdata_lowFTPL.GCdata_lowFTPL{2}.temporalGC)
  738. effect_field = effectTypeNames{i_effect};
  739. for i_TD = 1:length(GCdata_lowFTPL.GCdata_lowFTPL{2}.temporalGC{1})
  740. if ~isfield(GC_Par_to_Front_HighMinusLow, effect_field)
  741. GC_Par_to_Front_HighMinusLow.(effect_field) = struct();
  742. end
  743. if ~isfield(GC_Par_to_Front_HighMinusLow.(effect_field), sprintf('TD_%d', i_TD))
  744. GC_Par_to_Front_HighMinusLow.(effect_field).(sprintf('TD_%d', i_TD)) = [];
  745. end
  747. pooled_diffs = []; % temporary holder for this TD across all subjects
  749. for i_sub = 1:8
  750. % LOW
  751. GCvals_low = Gavg_tempGC_peranatreg_lowFTPL.source2target{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  752. subjIDs_low = Gavg_tempGC_peranatreg_lowFTPL.subject_index{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  754. if ~isempty(GCvals_low) && ~isempty(subjIDs_low)
  755. rows_low = (subjIDs_low == i_sub);
  756. GC_low = GCvals_low(rows_low, :);
  757. else
  758. GC_low = NaN;
  759. end
  761. % HIGH
  762. GCvals_high = Gavg_tempGC_peranatreg_highFTPL.source2target{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  763. subjIDs_high = Gavg_tempGC_peranatreg_highFTPL.subject_index{i_effect}{index_of_interest_source, index_of_interest_target, i_TD};
  765. if ~isempty(GCvals_high) && ~isempty(subjIDs_high)
  766. rows_high = (subjIDs_high == i_sub);
  767. GC_high = GCvals_high(rows_high, :);
  768. else
  769. GC_high = NaN;
  770. end
  772. if ~isempty(GC_low) && ~isempty(GC_high)
  773. n_rows = min(size(GC_low,1), size(GC_high,1));
  774. GC_diff = GC_high(1:n_rows, :) - GC_low(1:n_rows, :); % pairwise subtraction
  775. pooled_diffs = [pooled_diffs; GC_diff]; % append
  776. end
  777. end
  778. GC_Par_to_Front_HighMinusLow.(effect_field).(sprintf('TD_%d', i_TD)) = pooled_diffs;
  779. end
  780. end
  782. % Average across TDs
  783. GC_all = cell(1, length(effectTypeNames));
  784. pvals = zeros(1, length(effectTypeNames));
  785. numEffects = numel(effectTypeNames);
  787. % Prepare data
  788. for i = 1:length(effectTypeNames)
  789. data_TD1 = GC_Par_to_Front_HighMinusLow.(effectTypeNames{i}).TD_1;
  790. data_TD2 = GC_Par_to_Front_HighMinusLow.(effectTypeNames{i}).TD_2;
  792. % Average across TDs
  793. minLen = min(numel(data_TD1), numel(data_TD2));
  794. avgData = mean([data_TD1(1:minLen), data_TD2(1:minLen)], 2, 'omitnan');
  795. GC_all{i} = avgData;
  797. % t-test against zero
  798. [~, p] = ttest(avgData);
  799. pvals(i) = p;
  800. end
  802. % Plot
  803. effectNames_plot = {'Pred --> Pred', 'PE --> PE', 'Pred --> PE', 'PE --> Pred'};
  804. figure; hold on; box on;
  805. colors = lines(numEffects);
  806. rng(0);
  808. % Prepare data for boxplot
  809. allData = [];
  810. groupData = [];
  811. for i = 1:numEffects
  812. allData = [allData; GC_all{i}(:)];
  813. groupData = [groupData; repmat(i, numel(GC_all{i}), 1)];
  814. end
  816. % Overlay scatter points (lower alpha)
  817. for i = 1:numEffects
  818. x = i + (rand(size(GC_all{i})) - 0.5) * 0.4;
  819. scatter(x, GC_all{i}, 25, ...
  820. 'MarkerFaceColor', colors(i,:), ...
  821. 'MarkerEdgeColor', 'k', ...
  822. 'MarkerFaceAlpha', 0.3, ...
  823. 'MarkerEdgeAlpha', 0.3);
  825. % Mean line
  826. m = mean(GC_all{i}, 'omitnan');
  827. plot([i-0.3, i+0.3], [m, m], 'Color', colors(i,:), 'LineWidth', 3);
  828. end
  830. % Formatting
  831. xlim([0.5, numEffects + 0.5]);
  832. xticks(1:numEffects);
  833. xticklabels(effectNames_plot);
  834. xtickangle(20);
  835. ylabel('GC Difference (High FTPL > Low FTPL)');
  836. yline(0, '--k', 'LineWidth', 1);
  838. % Add exact p-values or significance stars
  839. ylim([-0.01, 0.02]);
  840. yl = ylim;
  841. text_y = yl(2) - 0.05 * range(yl);
  842. for i = 1:numEffects
  843. if pvals(i) < 0.001
  844. sigLabel = '***';
  845. elseif pvals(i) < 0.01
  846. sigLabel = '**';
  847. elseif pvals(i) < 0.05
  848. sigLabel = '*';
  849. else
  850. sigLabel = sprintf('p=%.3g', pvals(i));
  851. end
  852. text(i, text_y, sigLabel, ...
  853. 'HorizontalAlignment', 'center', ...
  854. 'FontSize', 12, 'FontWeight', 'bold', 'Color', 'k');
  855. end
  857. set(gcf, 'Color', 'w');
  859. filename = ['boxplot_compare_GC_lowhigh_FTPL_Par_to_Front_allpairs_avgData.png'];
  860. figfile = ['/isilon/LFMI/VMdrive/Lua/NASTD/Figures/' filename];
  861. saveas(gcf, figfile, 'png');
  863. %% Plot GC results
  864. AnatReg_CatLabels = ... %Select which anatomical regions should be plotted
  865. {'AntPFC'; ...
  866. 'VentralT'; ...
  867. 'SupParLob'};
  868. % AnatReg_CatLabels = ... %Select which anatomical regions should be plotted
  869. % {'AntPFC_IFG'; 'AntPFC_PrecentralG'; ...
  870. % 'VentralT_STG'; 'VentralT_MTG'; ...
  871. % 'SupParLob_PostcentralG'; 'SupParLob_SupramarginalG'};
  872. % AnatReg_CatLabels = ... %Select which anatomical regions should be plotted
  873. % {'AntPFC'; 'AntPFC_IFG'; 'AntPFC_PrecentralG'; ...
  874. % 'VentralT'; 'VentralT_STG'; 'VentralT_MTG'; ...
  875. % 'SupParLob'; 'SupParLob_PostcentralG'; 'SupParLob_SupramarginalG'; ...
  876. % 'OccipitalL'};
  878. %Single subject per electrode pairing
  879. % NASTD_ECoG_Connectivity_PlotGC_persub_aggrtone ...
  880. % (sub_list(vars.validSubjs), ...
  881. % ToneDur_text, ...
  882. % save_poststepFigs, ...
  883. % param, paths_NASTD_ECoG);
  884. % NASTD_ECoG_Connectivity_PlotGC_persub_pertone ...
  885. % (sub_list(vars.validSubjs), ...
  886. % ToneDur_text, InputDataType, label_ElecPairSel, ...
  887. % save_poststepFigs, ...
  888. % paths_NASTD_ECoG);
  890. %Single subject aggregated GC results
  891. % NASTD_ECoG_Connectivity_PlotGC_persub_aggrtoneanat ...
  892. % (sub_list(vars.validSubjs), ...
  893. % ToneDur_text, SelElecs, AnatReg_CatLabels, ...
  894. % save_poststepFigs, ...
  895. % param, paths_NASTD_ECoG);
  897. %Plot group-level aggregated GC results
  898. % NASTD_ECoG_Connectivity_PlotGC_allsub_aggrtone ...
  899. % (sub_list(vars.validSubjs), ...
  900. % ToneDur_text, SelElecs, AnatReg_CatLabels, ...
  901. % save_poststepFigs, ...
  902. % param, paths_NASTD_ECoG);
  903. NASTD_ECoG_Connectivity_PlotGC_allsub_aggrtone_median ...
  904. (sub_list(vars.validSubjs), ...
  905. ToneDur_text, SelElecs, AnatReg_CatLabels, ...
  906. save_poststepFigs, ...
  907. param, paths_NASTD_ECoG);
  909. %Plot GC output with statistics
  910. if strcmp(param.GC.tone_aggregation_label,'_30_31_32_33_34')
  911. NASTD_ECoG_Connectivity_PlotGC_statp30top32_vs_p33p34_lk ...
  912. (sub_list(vars.validSubjs), ...
  913. ToneDur_text, SelElecs, AnatReg_CatLabels, ...
  914. save_poststepFigs, ...
  915. param, paths_NASTD_ECoG);
  916. else
  917. NASTD_ECoG_Connectivity_PlotGC_statp1top31 ...
  918. (sub_list(vars.validSubjs), ...
  919. ToneDur_text, SelElecs, AnatReg_CatLabels, ...
  920. save_poststepFigs, ...
  921. param, paths_NASTD_ECoG);
  922. end
  924. %% 3) Compute Granger Causality (GC) during p34 for different PE effects
  925. addpath('/isilon/LFMI/VMdrive/Thomas/toolboxes/mvgc_v1.3/');
  926. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.01uncorr.mat') %p < 0.01 thresh,uncorr elec selection
  927. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.05uncorr.mat') %p < 0.05 thresh,uncorr elec selection
  928. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.025FDRcorr.mat') %p < 0.025 thresh, FDRcorr elec selection
  929. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_AllElecs.mat') % no thresh, all elec selection
  930. %%%load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.05uncorr_HighGamma.mat') %p < 0.05 thresh,uncorr elec selection
  931. % load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.01uncorr_HighGamma.mat') %p < 0.05 thresh,uncorr elec selection
  933. param.GC.fs = 512;
  934. param.GC.downsample = 0; %Cave: Downsampling lead to problems with tone sample selection
  935. if param.GC.downsample == 0
  936. param.GC.newfs = param.GC.fs;
  937. else
  938. param.GC.newfs = param.GC.fs/param.GC.downsample;
  939. end
  940. param.GC.nvars = 2;
  941. param.GC.regmode = 'OLS'; % VAR model estimation regression mode ('OLS', 'LWR' or empty for default
  942. param.GC.maxmorder = 50; % maximum model order for model order estimation, rule of thumb = number of samples per input data snippet, but not > 100
  943. param.GC.morder = 'AIC'; % model order to use ('actual', 'AIC', 'BIC' or supplied numerical value)
  945. param.GC.tstat = 'F'; % statistical test for MVGC: 'F' for Granger's F-test (default) or 'chi2' for Geweke's chi2 test
  946. param.GC.alpha = 0.05; % significance level for significance test
  947. param.GC.mhtc = 'FDRD'; % multiple hypothesis test correction (see routine 'significance')
  949. param.GC.ElecPairEffect = {'Pred_Pred','PE_PE','Pred_PE','PE_Pred'};
  950. param.GC.ElecPairRegion = 'AllRegions';
  951. param.GC.InputDataType = {'Broadband'}; %{'Broadband', 'HP05toLP30Hz', 'HighGamma_LogAmp'};
  953. param.GC.epochdur_ms = 'full'; %'full', 200, 100
  954. if strcmp(param.GC.epochdur_ms, 'full')
  955. param.GC.epochdur_ms_label = 'fullTW';
  956. else
  957. param.GC.epochdur_ms_label = [num2str(param.GC.epochdur_ms) 'msTW'];
  958. end
  960. for i_sub = vars.validSubjs
  962. tic
  963. GCdata{i_sub} = ...
  964. NASTD_ECoG_Connectivity_CalculateGC_PEeffect ...
  965. (sub_list, i_sub, ...
  966. ToneDur_text, SelElecs, ...
  967. param, paths_NASTD_ECoG);
  969. disp(['-- GC PE computation for sub: ' sub_list{i_sub} ' finished after ' num2str(round(toc/60),2) ' min --'])
  971. end
  973. %save output
  974. path_outputdata = ([paths_NASTD_ECoG.ECoGdata_Connectivity 'GC/PEeffect/']);
  975. if (~exist(path_outputdata, 'dir')); mkdir(path_outputdata); end
  976. label_outputdata = ['GCdataPEeffect_n' num2str(length(vars.validSubjs)) '_' ...
  977. param.GC.ElecPairRegion '_' ...
  978. param.GC.InputDataType{1} '_' ...
  979. param.GC.epochdur_ms_label '_ensemnorm'];
  980. save([path_outputdata label_outputdata], 'GCdata')
  981. % load([path_outputdata label_outputdata], 'GCdata')
  983. %Plot GC results
  984. AnatReg_CatLabels = ... %Select which anatomical regions should be plotted
  985. {'AntPFC'; ...
  986. 'VentralT'; ...
  987. 'SupParLob'};
  989. NASTD_ECoG_Connectivity_PlotGC_statPEeffect ...
  990. (sub_list(vars.validSubjs), ...
  991. ToneDur_text, SelElecs, AnatReg_CatLabels, ...
  992. save_poststepFigs, ...
  993. param, paths_NASTD_ECoG);
  996. %% 4) Plot and do statistics for electrode combinations between regions of the auditory hierarchy
  998. addpath('/isilon/LFMI/VMdrive/Lua/Temp2A/Temp2A_fMRI_pilot/toolboxes/spm12');
  999. load('/isilon/LFMI/VMdrive/Thomas//NaturalisticAuditorySequences_ToneDuration(NAS_TD)/ECoG/Analysis/Connectivity/ElecSelect/Allsub_n9/SelElecs_p0.05uncorr.mat') %p < 0.05 thresh,uncorr elec selection
  1000. atlas_path = 'brodmann.nii';
  1002. % Determine if any of the electrodes are located in one of the auditory
  1003. % regions of interest
  1004. % Classify each electrode as part of a Brodmann's area
  1005. BA_areas_Pred = assign_BA_from_MNI(SelElecs.PredEffect.("Elec Coords"), atlas_path);
  1006. BA_areas_PE = assign_BA_from_MNI(SelElecs.PEEffect.("Elec Coords"), atlas_path);
  1008. % Define ROI map: Brodmann areas → auditory/cognitive labels
  1009. roi_map = containers.Map( ...
  1010. [41, 42, 44, 45, 8, 40, 6, 22], ...
  1011. {'Auditory Cortex', ...
  1012. 'Auditory Cortex', ...
  1013. 'IFC Dorsal', ...
  1014. 'IFC Ventral', ...
  1015. 'dlPFC', ...
  1016. 'IPL', ...
  1017. 'PMC', ...
  1018. 'STS'});
  1020. % For Prediction electrodes
  1021. roi_labels_pred = cell(size(BA_areas_Pred));
  1022. for i = 1:length(BA_areas_Pred)
  1023. roi_labels_pred{i} = get_roi_label(BA_areas_Pred(i), roi_map);
  1024. end
  1026. % For Prediction Error electrodes
  1027. roi_labels_PE = cell(size(BA_areas_PE));
  1028. for i = 1:length(BA_areas_PE)
  1029. roi_labels_PE{i} = get_roi_label(BA_areas_PE(i), roi_map);
  1030. end
  1032. % Combine everything into one table
  1033. SelElecs.PredEffect.BAlabel = roi_labels_pred;
  1034. SelElecs.PEEffect.BAlabel = roi_labels_PE;
  1036. % Also need to update SelElecs.Pairs for stats
  1037. % === Update Pred_Pred with PredEffect ===
  1038. predPairFields = fieldnames(SelElecs.Pairs.Pred_Pred.AllRegions.Pairs_perelec);
  1039. predEffect = SelElecs.PredEffect;
  1041. for i = 1:length(predPairFields)
  1042. pairName = predPairFields{i};
  1043. targetElecs = SelElecs.Pairs.Pred_Pred.AllRegions.Pairs_perelec.(pairName);
  1044. BAlabels = cell(height(targetElecs), 1);
  1045. for j = 1:height(targetElecs)
  1046. elecLabel = targetElecs.("Electrode Label"){j};
  1047. subjLabel = targetElecs.("Subject Label"){j};
  1049. % Find matching row in PredEffect
  1050. matchIdx = strcmp(predEffect{:,1}, elecLabel) & strcmp(predEffect{:,2}, subjLabel);
  1052. if any(matchIdx)
  1053. baLabel = predEffect{matchIdx,12}{1}; % Get BAlabel (12th column)
  1054. else
  1055. baLabel = 'NA';
  1056. end
  1058. BAlabels{j,1} = baLabel;
  1059. end
  1060. targetElecs.BAlabel = BAlabels; % Add as new column
  1061. SelElecs.Pairs.Pred_Pred.AllRegions.Pairs_perelec.(pairName) = targetElecs; % Update
  1062. end
  1064. % === Update PE_PE with PEEffect ===
  1065. pePairFields = fieldnames(SelElecs.Pairs.PE_PE.AllRegions.Pairs_perelec);
  1066. peEffect = SelElecs.PEEffect;
  1068. for i = 1:length(pePairFields)
  1069. pairName = pePairFields{i};
  1070. targetElecs = SelElecs.Pairs.PE_PE.AllRegions.Pairs_perelec.(pairName);
  1071. BAlabels = cell(height(targetElecs), 1);
  1072. for j = 1:height(targetElecs)
  1073. elecLabel = targetElecs.("Electrode Label"){j};
  1074. subjLabel = targetElecs.("Subject Label"){j};
  1076. % Find matching row in PEEffect
  1077. matchIdx = strcmp(peEffect{:,1}, elecLabel) & strcmp(peEffect{:,2}, subjLabel);
  1079. if any(matchIdx)
  1080. baLabel = peEffect{matchIdx,12}{1}; % Get BAlabel (12th column)
  1081. else
  1082. baLabel = 'NA';
  1083. end
  1085. BAlabels{j,1} = baLabel;
  1086. end
  1087. targetElecs.BAlabel = BAlabels; % Add as new column
  1088. SelElecs.Pairs.PE_PE.AllRegions.Pairs_perelec.(pairName) = targetElecs; % Update
  1089. end
  1091. % Output a surface project of Brodmanns areas with only the desired ROIs
  1092. % addpath '/isilon/LFMI/VMdrive/Lua/surfice_atlas-master'
  1093. % desired_ROIs = [41, 42, 44, 45, 8, 40, 6, 22];
  1094. % nii_nii2atlas_rois('brodmann.nii', 'mylut.lut', desired_ROIs);
  1096. %% Extract GC values for each pair
  1097. subs = sub_list(vars.validSubjs);
  1098. param.GC = [];
  1099. param.GC.fs = 512;
  1100. param.GC.downsample = 0; %Cave: Downsampling lead to problems with tone sample selection
  1101. if param.GC.downsample == 0
  1102. param.GC.newfs = param.GC.fs;
  1103. else
  1104. param.GC.newfs = param.GC.fs/param.GC.downsample;
  1105. end
  1106. param.GC.nvars = 2;
  1107. param.GC.regmode = 'OLS'; % VAR model estimation regression mode ('OLS', 'LWR' or empty for default
  1108. param.GC.maxmorder = 50; % maximum model order for model order estimation, rule of thumb = number of samples per input data snippet, but not > 100
  1109. % param.GC.morder = 'AIC'; % model order to use ('actual', 'AIC', 'BIC' or supplied numerical value)
  1111. param.GC.tstat = 'F'; % statistical test for MVGC: 'F' for Granger's F-test (default) or 'chi2' for Geweke's chi2 test
  1112. param.GC.alpha = 0.05; % significance level for significance test
  1113. param.GC.mhtc = 'FDRD'; % multiple hypothesis test correction (see routine 'significance')
  1115. %param.GC.ElecPairEffect = {'Pred_Pred','PE_PE','Pred_PE','PE_Pred'};
  1116. param.GC.ElecPairEffect = {'Pred_Pred','PE_PE'};
  1117. param.GC.ElecPairRegion = 'AllRegions';
  1118. param.GC.InputDataType = {'Broadband'}; %{'Broadband', 'HP05toLP30Hz', 'HighGamma_LogAmp'};
  1120. param.GC.tone_aggregation = {1, 6, 11, 16, 21, 26, 31};
  1121. param.GC.tone_aggregation = {2, 7, 12, 17, 22, 27, 32};
  1122. param.GC.tone_aggregation = {30, 31, 32, 33, 34};
  1124. param.GC.tone_aggregation_label = cellfun(@(c)[c],param.GC.tone_aggregation);
  1125. param.GC.tone_aggregation_label = num2str((param.GC.tone_aggregation_label),'_%d');
  1126. param.GC.tone_aggregation_label = erase(param.GC.tone_aggregation_label, ' ');
  1128. param.GC.epochdur_ms = 'full'; %'full', 200, 100
  1129. if strcmp(param.GC.epochdur_ms, 'full')
  1130. param.GC.epochdur_ms_label = 'fullTW';
  1131. else
  1132. param.GC.epochdur_ms_label = [num2str(param.GC.epochdur_ms) 'msTW'];
  1133. end
  1134. param.GC.ElecPairRegion = 'AllRegions';
  1135. param.GC.InputDataType = {'Broadband'}; %{'Broadband', 'HP05toLP30Hz', 'HighGamma_LogAmp'};
  1137. % Load in GC data
  1138. path_outputdata = ([paths_NASTD_ECoG.ECoGdata_Connectivity 'GC/']);
  1139. if (~exist(path_outputdata, 'dir')); mkdir(path_outputdata); end
  1140. label_outputdata = ['GCdata_n' num2str(length(vars.validSubjs)) '_' ...
  1141. param.GC.ElecPairRegion '_' ...
  1142. param.GC.InputDataType{1} param.GC.tone_aggregation_label '_' ...
  1143. param.GC.epochdur_ms_label '_ensemnorm'];
  1144. load([path_outputdata label_outputdata], 'GCdata')
  1146. %Add in BA label information to GCdata
  1147. % for i_sub = vars.validSubjs
  1148. % for i_effect = 1:2
  1149. % % Assign labels that correspond to ind_pairedelecs_from
  1150. % sub = subs(i_sub);
  1151. % if i_effect == 1
  1152. % effecttype = 'Pred';
  1153. % else
  1154. % effecttype = 'PE';
  1155. % end
  1156. % GCdata{i_sub}.BAlabels{i_effect} = MNI_coords_all.BAlabel(strcmp(MNI_coords_all.("Subject Label"), sub) & strcmp(MNI_coords_all.EffectType, effecttype),:);
  1157. % end
  1158. % end
  1160. % Now the GC results also have the BA labels for auditory cortex
  1162. %Plot GC results
  1163. AnatReg_CatLabels = {'Auditory Cortex', ... %Select which anatomical regions should be plotted
  1164. 'IFC Dorsal', ...
  1165. 'IFC Ventral', ...
  1166. 'dlPFC', ...
  1167. 'IPL', ...
  1168. 'PMC', ...
  1169. 'STS'};
  1171. %Extract GC per anat regions
  1172. NASTD_ECoG_Connectivity_BAregions_lk ...
  1173. (sub_list(vars.validSubjs), GCdata,...
  1174. ToneDur_text, SelElecs, AnatReg_CatLabels, ...
  1175. save_poststepFigs, ...
  1176. param, paths_NASTD_ECoG);
  1178. % GCdata = cell(1, 9); % Initialize output
  1179. %
  1180. % for i = 1:9
  1181. % % Start from GCdata_T1 and overwrite the field we want to average
  1182. % GCdata{i} = GCdata_T1{i};
  1183. %
  1184. % avg_tgc = cell(1, 4); % Correct shape: 1×4
  1185. %
  1186. % for j = 1:4
  1187. % avg_tgc{j} = cell(1, 2); % Each is a 1×2 cell
  1188. %
  1189. % for k = 1:2
  1190. % % Extract source2target matrices
  1191. % mat1 = GCdata_T1{i}.temporalGC{j}{k}.source2target;
  1192. % mat2 = GCdata_T2{i}.temporalGC{j}{k}.source2target;
  1193. %
  1194. % % Compute average
  1195. % avg_mat = (mat1 + mat2) / 2;
  1196. %
  1197. % % Copy one struct and overwrite source2target with the average
  1198. % avg_struct = GCdata_T1{i}.temporalGC{j}{k};
  1199. % avg_struct.source2target = avg_mat;
  1200. %
  1201. % % Save back
  1202. % avg_tgc{j}{k} = avg_struct;
  1203. % end
  1204. % end
  1205. %
  1206. % % Assign averaged temporalGC to output
  1207. % GCdata{i}.temporalGC = avg_tgc;
  1208. % end

NASTD_ECoG_Connectivity_Main_lk.m at commit 9cafa02, no license · at the source

Overview

Authors: Thomas J Baumgarten1,2, Lua Koenig1, Richard Hardstone1, Adeen Flinker3,4, Sasha Devore3, Daniel Friedman3, Patricia Dugan3, Orrin Devinsky3, Biyu J He1,3,4,5,6
ORCID iDs: Adeen Flinker, Orrin Devinsky, Biyu J He
  1. Neuroscience Institute, New York University Langone Health, New York, NY USA
  2. Institute of Clinical Neuroscience and Medical Psychology, Medical Faculty, Heinrich Heine University Düsseldorf, Düsseldorf, Germany
  3. Department of Neurology, New York University Langone Health, New York, NY USA
  4. Department of Biomedical Engineering, NYU Tandon School of Engineering, New York, NY USA
  5. Department of Neuroscience, New York University Langone Health, New York, NY USA
  6. Department of Radiology, New York University Langone Health, New York, NY USA
Institutions: NYU Langone Health (United States); Heinrich Heine University Düsseldorf (Germany); New York University (United States)
Journal: Nature communications, volume 17, issue 1, article 8459
Dates: received 10 March 2025; accepted 26 June 2026; published online 9 July 2026
Type: Research article · Language: English
License: CC BY-NC-ND
Identifiers: DOI 10.1038/s41467-026-75359-0 · PMID 42426016 · PMCID PMC13478490 · OpenAlex W7167837097
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: EEG (modality), human (organism), cognitive (subfield)
Methods: Spectral & time-frequency, Connectivity, Statistics, Machine learning, Preprocessing, fMRI & imaging, Smoothing, state filtering, decompositions, Physiology & signal measures
Keywords: Perception, Cortex
MeSH: Anticipation, Psychological*, Auditory Perception*, Cerebral Cortex*, Electroencephalography*, Adult, Auditory Pathways, Brain Mapping, Female, Humans, Male, Middle Aged, Young Adult (* major topic)
Topic: Neuroscience and Music Perception (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: EC | EU Framework Programme for Research and Innovation H2020 | H2020 Priority Excellent Science | H2020 Marie Skłodowska-Curie Actions (H2020 Excellent Science - Marie Skłodowska-Curie Actions) (795998)
Citations: not cited yet (Europe PMC); 53 references in the paper

Abstract

The abstract is not reproduced here: the paper's license (CC BY-NC-ND) does not allow it. Read it in the paper, at the publisher or on Europe PMC.

Repository

Its files are read in the Code ↔ Paper reader above, with 26 matches between paragraphs and lines of code.

BiyuHeLab/TimeForwardPrediction_iEEG_Baumgarten_Koenig

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: 9cafa02247de275fa5b20745e861a6cccaf1bab7, 29 May 2026
Languages: MATLAB (279), R (1)
Size: 389 files, 280 scripts
Software Heritage: not archived
Found in: “Code availability”
Holds: README
Not found: license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: Statistics and Machine Learning Toolbox (61 files), fdr_bh (Benjamini-Hochberg FDR) (31 files), FieldTrip (27 files), FreeSurfer (21 files), Signal Processing Toolbox (21 files), Psychtoolbox (18 files), shadedErrorBar (14 files), SPM (2 files)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
281 files
At the source: github.com/BiyuHeLab/TimeForwardPrediction_iEEG_Baumgarten_Koenig

Code availability statement

The paper has a code availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

Read it in the paper: doi.org/10.1038/s41467-026-75359-0.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 280 scripts, each with its path and the digest of its content;
  • 26 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data availability statement

The paper has a data availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

  • no repository, dataset or request procedure was recognized in it

Read it in the paper: doi.org/10.1038/s41467-026-75359-0.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 9 authors, 2 keywords, 12 MeSH terms, 1 funder, 48 references.

Cite

This paper

Baumgarten, T. J., Koenig, L., Hardstone, R., Flinker, A., Devore, S., Friedman, D., Dugan, P., Devinsky, O., & He, B. J. (2026). Neural mechanisms of time-forward predictions for naturalistic auditory tone sequences. Nature communications, 17(1), 8459. https://doi.org/10.1038/s41467-026-75359-0

BibTeX

@article{baumgarten2026neural,
author = {Baumgarten, Thomas J and Koenig, Lua and Hardstone, Richard and Flinker, Adeen and Devore, Sasha and Friedman, Daniel and Dugan, Patricia and Devinsky, Orrin and He, Biyu J},
title = {{Neural mechanisms of time-forward predictions for naturalistic auditory tone sequences}},
journal = {Nature communications},
year = {2026},
month = jul,
volume = {17},
number = {1},
pages = {8459},
publisher = {Nature Publishing Group},
issn = {2041-1723},
doi = {10.1038/s41467-026-75359-0},
url = {https://doi.org/10.1038/s41467-026-75359-0},
pmid = {42426016},
pmcid = {PMC13478490}
}

RIS

TY - JOUR
AU - Baumgarten, Thomas J
AU - Koenig, Lua
AU - Hardstone, Richard
AU - Flinker, Adeen
AU - Devore, Sasha
AU - Friedman, Daniel
AU - Dugan, Patricia
AU - Devinsky, Orrin
AU - He, Biyu J
TI - Neural mechanisms of time-forward predictions for naturalistic auditory tone sequences
T2 - Nature communications
J2 - Nat Commun
PY - 2026
DA - 2026/07/09
VL - 17
IS - 1
SP - 8459
SN - 2041-1723
PB - Nature Publishing Group
DO - 10.1038/s41467-026-75359-0
UR - https://doi.org/10.1038/s41467-026-75359-0
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s41467-026-75359-0",
"type": "article-journal",
"title": "Neural mechanisms of time-forward predictions for naturalistic auditory tone sequences",
"container-title": "Nature communications",
"author": [
{
"family": "Baumgarten",
"given": "Thomas J"
},
{
"family": "Koenig",
"given": "Lua"
},
{
"family": "Hardstone",
"given": "Richard"
},
{
"family": "Flinker",
"given": "Adeen"
},
{
"family": "Devore",
"given": "Sasha"
},
{
"family": "Friedman",
"given": "Daniel"
},
{
"family": "Dugan",
"given": "Patricia"
},
{
"family": "Devinsky",
"given": "Orrin"
},
{
"family": "He",
"given": "Biyu J"
}
],
"container-title-short": "Nat Commun",
"volume": "17",
"issue": "1",
"page": "8459",
"DOI": "10.1038/s41467-026-75359-0",
"PMID": "42426016",
"PMCID": "PMC13478490",
"ISSN": "2041-1723",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s41467-026-75359-0",
"language": "en",
"issued": {
"date-parts": [
[
2026,
7,
9
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1038/s41467-026-73540-z [code]
Predictive acoustical processing in human cortical layers.
Journal: Nature communications
In common: FieldTrip, FreeSurfer, SPM, 2 other tools, cognitive, 10 references
[2] doi:10.1002/advs.77857 [code]
Brain Network Dynamics of Local and Global Predictive Processing in Aging.
Journal: Advanced science (Weinheim, Baden-Wurttemberg, Germany)
In common: FieldTrip, SPM, Signal Processing Toolbox, 1 other tool, cognitive, 7 references
[3] doi:10.1038/s41467-026-73947-8 [code]
Frontal cortex organization supporting audiovisual processing during naturalistic viewing.
Journal: Nature communications
In common: cognitive, 3 references, 2 authors
[4] doi:10.7554/elife.108408 [code]
Frequency and laminar profile of feature-specific visual activity revealed by interleaved EEG-fMRI.
Journal: eLife
In common: FieldTrip, FreeSurfer, SPM, 2 other tools, EEG, 5 references
[5] doi:10.1371/journal.pbio.3003966 [code]
Behavioral engagement and stimulus regularities coordinate auditory-prefrontal network activity.
Journal: PLoS biology
In common: fdr_bh (Benjamini-Hochberg FDR), Statistics and Machine Learning Toolbox, 8 references
[6] doi:10.1126/sciadv.aec0518 [code]
Hybrid spatial organization and evidence for magnitude-independent neural coding of linguistic information during sentence production.
Journal: Science advances
In common: 3 references, 2 authors
[7] doi:10.1038/s41467-026-71151-2 [code]
Common and distinct neural correlates of social interaction processing and theory of mind in narratives.
Journal: Nature communications
In common: fdr_bh (Benjamini-Hochberg FDR), Psychtoolbox, FieldTrip, 4 other tools, cognitive, 1 reference
[8] doi:10.1038/s41593-026-02345-6 [code]
Human hippocampal ripples tune cortical responses based on predicted uncertainty.
Journal: Nature neuroscience
In common: FieldTrip, Signal Processing Toolbox, Statistics and Machine Learning Toolbox, cognitive, 7 references
[9] doi:10.1162/imag.a.1218 [code]
Reliability and signal comparison of OPM-MEG, fMRI &amp; iEEG in a repeated movie viewing paradigm.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: fdr_bh (Benjamini-Hochberg FDR), FieldTrip, Statistics and Machine Learning Toolbox, 5 references
[10] doi:10.1371/journal.pcbi.1014563 [code]
Single pulse electrical stimulation in white matter modulates iEEG visual responses in human early visual cortex.
Journal: PLoS computational biology
In common: fdr_bh (Benjamini-Hochberg FDR), FreeSurfer, SPM, 2 other tools, EEG, 2 references

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.