OSCR

Dentate gyrus interneurons modulate winner-take-all network dynamics in freely behaving mice.

Code ↔ Paper

40 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 40 matches · 17 of them tie a paragraph to a whole file, not to given lines: weak matches, whose lines are not tinted
  1. [1] § STAR METHODS › METHOD DETAILS › DG circuit model ↔ dendritic_somatic_transfer.py, lines 93–215 · score 0.96 · dendritic somatic transfer, NMDA voltage dependence, somatic voltage, v_soma, dendritic integration, GABA conductances
  2. [2] § STAR METHODS › METHOD DETAILS › DG circuit model ↔ optogenetic_experiment.py, lines 106–156 · score 0.96 · transduction failure rate, fiber tip, Light induced activation, log normal distribution, activation probability, light intensity
  3. [3] § STAR METHODS › METHOD DETAILS › DG circuit model ↔ DG_circuit_dendritic_somatic_transfer.py, lines 417–467 · score 0.93 · hilar region, Connection probabilities, dendritic inhibition, molecular layer, local connections, granule cell
  4. [4] § STAR METHODS › METHOD DETAILS › DG circuit model ↔ dendritic_somatic_network_clamp.py, lines 241–359 · score 0.92 · dendritic somatic transfer, NMDA voltage dependence, dendritic voltage, NMDA fraction, GABA conductances, membrane
  5. [5] § STAR METHODS › METHOD DETAILS › Putative monosynaptic connections ↔ buzcodeAddOns/MonoSyn/burstCCG.m, the whole file · a weak match · score 0.89 · inter spike intervals, Spike transmission probability, presynaptic spike, spike trains, 4 ms, 0.8 ms
  6. [6] § STAR METHODS › METHOD DETAILS › Putative monosynaptic connections ↔ buzcodeAddOns/MonoSyn/analyzeMonoSyn.m, the whole file · a weak match · score 0.85 · inter spike intervals, Spike transmission probability, putatively connected, 4 ms, 0.8 ms, subtracted
  7. [7] § STAR METHODS › METHOD DETAILS › Cell classification ↔ buzcodeAddOns/CellClassification/trainSVMpvssst.m, the whole file · a weak match · score 0.84 · Classification Learner App, cross validation, fold, AB, kernel, CV2
  8. [8] § STAR METHODS › METHOD DETAILS › Cell classification ↔ buzcodeAddOns/CellClassification/trainSVMpvsst.m, the whole file · a weak match · score 0.84 · Classification Learner App, cross validation, fold, AB, kernel, CV2
  9. [9] § STAR METHODS › METHOD DETAILS › Cell classification ↔ buzcodeAddOns/CellClassification/YutaClassifier.m, the whole file · a weak match · score 0.83 · principal components, maximum waveform, firing rate ratios, granule cell, labelling, derivative
  10. [10] § STAR METHODS › METHOD DETAILS › Spatial tuning ↔ buzcodeAddOns/PlaceFields/get_placeFieldsLinear.m, lines 155–216 · score 0.81 · Spatial coherence, Spatial tuning, place field, firing rate peak, SI, Bits
  11. [11] § STAR METHODS › METHOD DETAILS › Behavior ↔ buzcodeAddOns/makeRadialMazeMaps.m, the whole file · a weak match · score 0.80 · radial arm maze, rewarded arms, firing rate maps, minute, cm, Behavior
  12. [12] § STAR METHODS › METHOD DETAILS › Spatial tuning ↔ buzcodeAddOns/PlaceFields/get_placeFieldsOF.m, lines 167–232 · score 0.79 · Spatial coherence, Spatial tuning, place field, firing rate peak, Bits, median
  13. [13] § STAR METHODS › METHOD DETAILS › DG circuit model ↔ statistical_testing.py, lines 558–632 · score 0.77 · activation probability, optogenetic drive, light intensity, optogenetic stimulation, target population, Opsin expression
  14. [14] § STAR METHODS › METHOD DETAILS › LFP analysis ↔ buzcodeAddOns/LFP_analysis/DGlayerProfile.m, lines 1–74 · score 0.76 · 5–15 Hz, theta gamma, 125 ms, comodulogram, profiles, CSD
  15. [15] § STAR METHODS › METHOD DETAILS › Generalized linear model (peer prediction) ↔ buzcodeAddOns/PeerPrediction/peerPredictOpto1sec.m, lines 64–190 · score 0.76 · generalized linear model, peer, fitglm, predicted, R2, interactions
  16. [16] § STAR METHODS › METHOD DETAILS › Generalized linear model (peer prediction) ↔ buzcodeAddOns/PeerPrediction/CrossValidationAssemblyPrediction.m, the whole file · a weak match · score 0.75 · generalized linear model, Binned spike trains, peer, root, predicted, validation
  17. [17] § STAR METHODS › METHOD DETAILS › Behavior ↔ buzcodeAddOns/Behavior/OptitrackBehav.m, the whole file · a weak match · score 0.74 · rewarded arms, radial arm, radial maze, barrier, solenoid, cm
  18. [18] § STAR METHODS › METHOD DETAILS › DG circuit model ↔ DG_protocol.py, lines 2523–2649 · score 0.72 · standard deviation, paradoxical excitation, synaptic weight, optogenetic stimulation, opsin expressing, unchanged
  19. [19] § Results › Supervised classification of DG cells into four types based on physiological criteria ↔ DG_IN_Figures_final.m, lines 458–511 · score 0.69 · spike asymmetry, ACG rise, spike width, S1d, CV2, wake
  20. [20] § STAR METHODS › METHOD DETAILS › LFP analysis ↔ DG_IN_Ephys/DG_IN_LFP_Scripts.m, lines 165–249 · score 0.68 · theta gamma comodulogram, CSD, component, profiles, troughs, bands
  21. [21] § Results › Spatial tuning properties of DG cells ↔ buzcodeAddOns/PlaceFields/get_placeFieldsLinear.m, lines 155–216 · score 0.68 · spatial coherence, spatial tuning, place fields, SI, track, linear
  22. [22] § STAR METHODS › METHOD DETAILS › Cell classification ↔ DG_IN_Figures_final.m, lines 545–595 · score 0.66 · anatomical location, ground truth, granule cell, waking, SVM, waveforms
  23. [23] § Results › Identification of dentate interneurons with extracellular electrophysiology and optotagging ↔ DG_IN_Figures_final.m, lines 3203–3290 · score 0.64 · ChR2, eNpHR, PV Cre, SST Cre, mice, classes
  24. [24] § Results › Mechanisms of paradoxical activation upon DG interneuron stimulation ↔ buzcodeAddOns/PeerPrediction/peerPredictOpto1sec.m, lines 64–190 · score 0.64 · peer prediction, generalized linear models, GLMs, onset, trained, validation
  25. [25] § Results › Spatial tuning properties of DG cells ↔ buzcodeAddOns/PlaceFields/get_placeFieldsOF.m, lines 167–232 · score 0.63 · spatial coherence, spatial tuning, place fields, properties, behavioral, cells
  26. [26] § Results › Mechanisms of paradoxical activation upon DG interneuron stimulation ↔ DG_IN_Figures_final.m, lines 63–121 · score 0.61 · PV Cre mice, SST Cre mice, eNpHR, S6e, SVM, trained
  27. [27] § STAR METHODS › METHOD DETAILS › LFP analysis ↔ buzcodeAddOns/EventDetection/DetectDSpikes_v4.m, lines 1–99 · score 0.60 · 2–50 Hz, molecular layer, hilus, DS2, filtered, threshold
  28. [28] § STAR METHODS › METHOD DETAILS › LFP analysis ↔ buzcodeAddOns/EventDetection/DetectDSpikes_v5.m, lines 1–102 · score 0.60 · 2–50 Hz, molecular layer, hilus, DS2, filtered, threshold
  29. [29] § STAR METHODS › METHOD DETAILS › Opto-tagging and optogenetic manipulations ↔ buzcodeAddOns/Utilities/findAmbiguous.m, the whole file · a weak match · score 0.59 · ambiguous response, modulation ratio, stim, firing rate, Opto, classified
  30. [30] § STAR METHODS › METHOD DETAILS › Surgical Procedures ↔ DG_circuit_optogenetic_optimization.py, lines 2719–2853 · score 0.59 · theta oscillation, theta phase, Diagnostic, depth, parallel, amplitude
  31. [31] § STAR METHODS › METHOD DETAILS › Putative monosynaptic connections ↔ buzcodeAddOns/analysis/monosynapticPairs/bz_GetMonoSynapticallyConnected.m, the whole file · a weak match · score 0.59 · monosynaptic connections, Poisson, CCG, detection, ms, predictor
  32. [32] § Results › Synaptic connectivity in the DG network ↔ buzcodeAddOns/MonoSyn/analyzeMonoLTP.m, the whole file · a weak match · score 0.58 · synaptic strength, transmission probability, spike transmission, brain, putative, excitatory
  33. [33] § Results › Mechanisms of paradoxical activation upon DG interneuron stimulation ↔ buzcodeAddOns/PeerPrediction/CrossValidationAssemblyPrediction.m, the whole file · a weak match · score 0.57 · peer prediction, generalized linear models, validation, trained, spike, cells
  34. [34] § STAR METHODS › METHOD DETAILS › Opto-tagging and optogenetic manipulations ↔ buzcodeAddOns/Optogenetics/reclassifyOptoResponses2.m, the whole file · a weak match · score 0.56 · half rise, standard deviation, ambiguous, Opto, firing rate, classified
  35. [35] § STAR METHODS › METHOD DETAILS › Opto-tagging and optogenetic manipulations ↔ buzcodeAddOns/Utilities/findAmbiguous.m, the whole file · a weak match · score 0.54 · eNpHR, standard deviation, ambiguous, firing rate, Opto, opsin
  36. [36] § Results › Synaptic connectivity in the DG network ↔ buzcodeAddOns/analysis/monosynapticPairs/bz_GetMonoSynapticallyConnected.m, the whole file · a weak match · score 0.54 · monosynaptic connections, synaptic connectivity, likelihoods, CCG, synapses, network
  37. [37] § Results › Supervised classification of DG cells into four types based on physiological criteria ↔ buzcodeAddOns/CellClassification/trainSVMpvssst.m, the whole file · a weak match · score 0.53 · cross validation, accuracy, fold, CV2, trained, rise
  38. [38] § STAR METHODS › METHOD DETAILS › Putative monosynaptic connections ↔ buzcodeAddOns/analysis/monosynapticPairs/bz_fitPoissPlasticity.m, lines 1–30 · score 0.53 · monosynaptic connections, Poisson, timescale, postsynaptic, cross, binning
  39. [39] § STAR METHODS › METHOD DETAILS › Sleep scoring ↔ buzcodeAddOns/detectors/detectStates/SleepScoreMaster/ClusterStates_DetermineStates.m, the whole file · a weak match · score 0.53 · quiet wake, scoring, Sleep, NREM, theta
  40. [40] § STAR METHODS › METHOD DETAILS › Firing rate distribution equality ↔ DG_IN_Figures_final.m, lines 2790–2840 · score 0.53 · Lorenz curve, S8a, CV, Gini, 1–2, DG

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 · 3,691 lines · 144 KB · no license · 5 matches

  1. %% README
  2. % Download and install 'https://github.com/buzsakilab/buzcode'. Then
  3. % download this repository. Add all folders and subfolders to your matlab
  4. % path.
  5. %
  6. % Download data from 'https://buzsakilab.nyumc.org/datasets/' to a folder
  7. % on your computer and replace basedir (example 'Z:\Buzsakilabspace\LabShare\HainmuellerT\')
  8. % below with the the folder name. Run all scripts under 'preparations'
  9. % before trying to create any of the figures.
  10. %
  11. % [ ] CHANGE 'basedir' TO THE FOLDER WHERE YOU DOWNLOADED THE DATA
  12. basedir = 'Z:\buzsakilab\Buzsakilabspace\LabShare\HainmuellerT';
  14. %% Prepare handle and batch_metrics
  15. % Animal selection
  16. basedir = 'Z:\buzsakilab\Buzsakilabspace\LabShare\HainmuellerT\';
  17. ChRmice = {'SST2','SST4','YutaMouse20','YutaMouse21','YutaMouse37','YutaMouse38',...
  18. 'YutaMouse39','YutaMouse40','PV1','PV3','PV5','PV16','PV18','PV19','PV21','PV29'}; % PV16 1:1, PV18, PV23, SST6 1:3, SST7, others pure, PV28 thalamic injection
  19. HRmice = {'PV13','PV22','SST3','SST5'}; % 'PV20',
  20. linear = {'PV19','PV21','YutaMouse21','YutaMouse29'}; % Linear probe mice
  21. WTmice = {'TH10','TH11','TH12','TH13','TH14','TH15'}; % TH17 once downloaded
  22. animals = cat(2,ChRmice,HRmice,WTmice);
  24. % Querry for numbers of units meeting different criteria
  25. querryCells(1).tag = 'nDG';
  26. querryCells(1).include = struct('brainRegion','DG');
  27. querryFiles = {'DS','.DS2.events.mat'};
  29. handle = makeDShandle('basedir',basedir,'animals',animals,...
  30. 'querryCells',querryCells,'querryFiles',querryFiles);
  31. handle = handle([handle(:).hasDS] & [handle(:).hasMetrics]);
  33. % Load batch
  34. batch_metrics = LoadCellMetricBatch('basepaths',{handle(:).basepath});
  36. % Run Waveform and ACG quality metrics
  37. batch_metrics = get_waveformACGquality('metrics',batch_metrics,'save_results',false);
  39. % TODO: Rerun opto-tagging assignments
  40. % batch_metrics = reclassifyOptoResponses(batch_metrics,'maxHalfRise',.005);
  42. % Rerun Yuta Classifier on final cell selection
  43. batch_metrics = YutaClassWrapper(batch_metrics); % rng initialized, should be deterministic
  45. % Add monosyn curation status to batch_metrics
  46. batch_metrics.monoSynManual = ismember(batch_metrics.batchIDs,...
  47. find([handle(:).monoSynManual]))*1; % *1 to make double
  49. % Refine peaks for DS2 (% TODO: incorporate this somewhere else)
  50. batch_metrics = refinePeaks(batch_metrics);
  52. % % TODO: New version of excluding ambiguous & 0.1 Hz criterion for HR (Rev1)
  53. batch_metrics = reclassifyOptoResponses2(batch_metrics); % Added 12/16/2025
  54. % batch_metrics = addAmbiguousResponse(batch_metrics); % Added 11/14/2025
  55. batch_metrics = add_evtPeakActivation(batch_metrics,'opto_10ms','InEvent',[0.0005 .01]);
  56. batch_metrics = add_evtPeakActivation(batch_metrics,'opto_1sec','InEvent',[.05 .95]);
  58. % Sort fields in alphabetical order
  59. batch_metrics = orderfields(batch_metrics);
  61. clear basedir ChRmice HRmice WTmice linear animals querryCells querryFiles
  63. %% Make 'standard' SVM based on all PV-Cre and SST-Cre mice
  64. % Make table for classifier learning
  65. exclude = struct('putativeCellType',repmat({'Unknown'},1,4),...
  66. 'tags',repmat({{'InverseSpike','Bad'}},1,4));
  67. include = struct('putativeCellType',{'','','Granule Cell','Mossy Cell'},...
  68. 'groundTruthClassification',{'PV+','SST+','',''},...
  69. 'brainRegion','DG','geneticLine','Cre',...); % ...
  70. 'labels','eNpHR'); %-- Toggle this line for HR classifier control Fig S6e
  71. % Shorthand, will pick up PV-Cre and SST-Cre -- matlab automatically repeats singular entries
  73. % Get features
  74. fields = {'cv2','acg_tau_rise','troughToPeak','ab_ratio','firingRate','ds2amplitude','wakeNREMratio'};
  75. for ff = 1:length(fields)
  76. Tbl{ff} = get_metricfield(batch_metrics,fields{ff},...
  77. 'include',include,'exclude',exclude,'tukeyFence',Inf);
  78. end
  80. Tbl = cat(2,Tbl{:});
  81. Tbl = Tbl(:,[1:2:end-1 end]);
  82. Tbl = array2table(Tbl,'VariableNames',[fields,{'labels'}]);
  84. % Retrain SVM from classification learner
  85. % 250126 -- optimizable SVM, 10-fold cv
  86. % '>> classificationLearner' run 'optimizableSVM' and 'ExportModel'
  88. rng(0); % Initialize RNG for reproducibility
  89. [SVMecoc, validationAccuracy, partitionedModel] = trainSVMpvsst(Tbl); % - use this to add posterior fit
  90. SVM = SVMecoc.ClassificationSVM;
  91. fprintf('Validation accuracy %.2f%%\n',validationAccuracy*100)
  93. % Get predictions for high-certainty confusion matrix only
  94. % [labels,~,~,scoreProb] = resubPredict(SVM);
  95. % isHQ = max(scoreProb,[],2)>.5;
  96. % response = Tbl.labels;
  98. % Perform cross-validation
  99. % partitionedModel = crossval(SVM, 'KFold', 10);
  101. % Compute validation predictions
  102. [labels, ~, ~, scoreProb] = kfoldPredict(partitionedModel);
  103. isHQ = max(scoreProb,[],2)>.5;
  104. response = Tbl.labels;
  106. % All predictions
  107. figure;
  108. cm = confusionchart(response, labels);
  109. cm.RowSummary = 'row-normalized';
  110. cm.ColumnSummary = 'column-normalized';
  111. title('All classifications');
  113. % High-certainty predictions, only
  114. figure;
  115. cm = confusionchart(response(isHQ), labels(isHQ));
  116. cm.RowSummary = 'row-normalized';
  117. cm.ColumnSummary = 'column-normalized';
  118. title('High-certainty classifications');
  120. clear cm labels score pbscore scoreProb SVMecoc Tbl ff fields include exclude...
  121. validationAccuracy response labels scoreProb isHQ cm partitionedModel
  123. %% Classify batch_metrics with SVM classifier (keep all cells, set non-classifiable DG to 'Unknown')
  124. % Load batch metrics first
  125. clear AllTbl
  126. celltype = {'PV interneuron','SST interneuron','Granule Cell','Mossy Cell','Unknown'};
  128. % Remove cells with inverse spikes or bad waveforms
  129. [~, classIDs] = select_cells(batch_metrics,...
  130. 'include',struct('brainRegion',{'DG'}),...
  131. 'exclude',struct('putativeCellType',{'Unknown'},'tags',{{'InverseSpike','Bad'}}));
  132. class_metrics = batch_metrics;
  133. isDG = contains(class_metrics.brainRegion,'DG');
  134. isGood = false(1,length(batch_metrics.UID));
  135. isGood(classIDs) = true;
  136. %isGood = ~contains(class_metrics.tags,{{'InverseSpike','Bad'}}) & ~contains(class_metrics.putativeCellType,'Unknown');
  138. % Get features
  139. fields = {'cv2','acg_tau_rise','troughToPeak','ab_ratio','firingRate','ds2amplitude','wakeNREMratio'};
  140. for ff = 1:length(fields)
  141. AllTbl{ff} = class_metrics.(fields{ff});
  142. end
  144. % Assemble data into table
  145. AllTbl = array2table(cat(1,AllTbl{:})','VariableNames',fields);
  147. % Set rng for deterministic outcome
  148. rng(0);
  150. % Run Classifier
  151. [labels,~,~,scoreProb] = predict(SVM,AllTbl);
  153. % Set labels with posterior prob < 50% to 'Unknown'
  154. labels(max(scoreProb,[],2)<.5) = 5;
  156. % Set DG labels with bad waveforms to 'Unknown'
  157. labels(isDG & ~isGood) = 5;
  159. for ll = 1:5
  160. theseIDs = find(labels'==ll & isDG);
  161. for n = theseIDs
  162. class_metrics.putativeCellType{n} = celltype{ll};
  163. end
  164. end
  166. clear ll n theseIDs labels AllTbl scoreProb fields classIDs celltype ff isDG isGood
  168. %% Get synaptic parameters for each connected pair
  169. % TODO: Correct cell-type determination
  170. % TODO: Make analysis for each CELL (connectivity rates) and each PAIR
  171. % (mono_analysis)
  173. reanalyze = false;
  174. mono_analysis = getMonoConnAnalysis(class_metrics,'reanalyze',reanalyze);
  175. clear reanalyze;
  177. %% Connected and non-connected simultaneously recorded pairs and connection rates (class metrics)
  178. % Create in- and exclude structure
  179. %criteria = struct('batchIDs',{[5 15]}); % SUBSELECT DATASETS HERE!
  180. labels = {'PV','SST','GC','MC','MEC PYR','PVo','SSTo'};
  181. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  182. 'Granule Cell','Mossy Cell','Pyramidal Cell','',''},...
  183. 'brainRegion',{'DG','DG','DG','DG','ENTm','DG','DG'},...
  184. 'groundTruthClassification',{'','','','','','PV+','SST+'},...
  185. 'monoSynManual',1);
  186. exclude = struct('groundTruthClassification',{'','','+','+','+','',''});
  188. % Get the connectivity data (rates and pair global IDs)
  189. mono_res = getMonoConnRates(class_metrics,'include',include,'exclude',exclude);
  191. fprintf('Table rows = pre\n')
  192. array2table(sum(mono_res.nEcon,3),'RowNames',labels,'VariableNames',labels)
  194. fprintf('Table rows = pre(%%)\n')
  195. confrct = round(sum(mono_res.nEcon,3)./sum(mono_res.nEtotal,3)*100,2);
  196. array2table(confrct,'RowNames',labels,'VariableNames',labels)
  198. % Make heatmap
  199. figure('position',[100 100 220 175]); heatmap(log10(confrct(1:5,1:5)));
  200. colormap(bwyr); caxis([-1 .5]); % For color
  202. figure('position',[100 100 220 175]); heatmap(confrct(1:5,1:5)); % For values
  204. clear labels include exclude tbl
  206. %% Converging and non-converging pairs
  207. % Create in- and exclude structure
  208. %criteria = struct('batchIDs',{[5 15]}); % SUBSELECT DATASETS HERE!
  209. labels = {'PV','SST','GC','MC','MEC PYR','PVo','SSTo'};
  210. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  211. 'Granule Cell','Mossy Cell','Pyramidal Cell','',''},...
  212. 'brainRegion',{'DG','DG','DG','DG','ENTm','DG','DG'},...
  213. 'groundTruthClassification',{'','','','','','PV+','SST+'},...
  214. 'monoSynManual',1);
  215. exclude = struct('groundTruthClassification',{'','','+','+','+','',''});
  217. % Get the connectivity data (rates and pair global IDs)
  218. convg_res = getConvergingPairs(class_metrics,'include',include,'exclude',exclude);
  220. % fprintf('Table rows = pre\n')
  221. % array2table(sum(mono_res.nEcon,3),'RowNames',labels,'VariableNames',labels)
  222. %
  223. % fprintf('Table rows = pre(%%)\n')
  224. % array2table((sum(mono_res.nEcon,3)./sum(mono_res.nEtotal,3)*100),...
  225. % 'RowNames',labels,'VariableNames',labels)
  227. clear labels include exclude
  229. %% Load place fields (Figure 4)
  230. linear = load_LinearBatch(batch_metrics);
  231. openfield = load_OpenfieldBatch(batch_metrics);
  232. radial = load_RadialBatch(batch_metrics);
  234. %% Lii/iii separate connected and non-connected simultaneously recorded pairs and connection rates (class metrics)
  235. % Separate for MEC Lii and Liii
  236. %
  237. % Create in- and exclude structure
  238. %criteria = struct('batchIDs',{[5 15]}); % SUBSELECT DATASETS HERE!
  239. labels = {'PV','SST','GC','MC','PYR2','PYR3','PVo','SSTo'};
  240. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  241. 'Granule Cell','Mossy Cell','Pyramidal Cell','Pyramidal Cell','',''},...
  242. 'brainRegion',{'DG','DG','DG','DG','ENTm2','ENTm3','DG','DG'},...
  243. 'groundTruthClassification',{'','','','','','','PV+','SST+'},...
  244. 'monoSynManual',1);
  245. exclude = struct('groundTruthClassification',{'','','+','+','+','+','',''});
  247. % Get the connectivity data (rates and pair global IDs)
  248. mono_res_l23 = getMonoConnRates(class_metrics,'include',include,'exclude',exclude);
  250. fprintf('Table rows = pre\n')
  251. array2table(sum(mono_res_l23.nEcon,3),'RowNames',labels,'VariableNames',labels)
  253. fprintf('Table rows = pre(%%)\n')
  254. confrct = round(sum(mono_res_l23.nEcon,3)./sum(mono_res_l23.nEtotal,3)*100,2);
  255. array2table(confrct,'RowNames',labels,'VariableNames',labels)
  257. % Make heatmap
  258. figure('position',[100 100 250 205]); heatmap(log10(confrct(1:6,1:6)));
  259. colormap(bwyr); caxis([-1 .5]); % For color
  260. %xticklabels(labels(1:6)); xtickangle(45);
  262. figure('position',[100 100 250 205]); heatmap(confrct(1:6,1:6)); % For values
  264. clear labels include exclude tbl
  267. %% Figure 1
  269. %% 1c(1) Single cell example opto response & ACG (eNpHR - PV)
  270. % Load cell_metrics, spikes, opto_1sec
  271. session = 'PV22_240308';
  272. color = [.56 .39 .93];
  273. UID = 47; %PV22_240308-UID 47-20ms
  274. printfig = true;
  276. % Load stuff
  277. basepath = local2remote(session);
  278. basename = bz_BasenameFromBasepath(basepath);
  279. cell_metrics = load_cellmetrics_TH('basepath',basepath);
  280. spikes = load_spikes_TH('basepath',basepath);
  281. load([basepath,filesep,basename,'.opto_20ms.manipulation.mat'],'opto_20ms');
  283. % ACG
  284. acgT = cell_metrics.general.ccg_time;
  285. acgBinSz = mean(diff(acgT));
  286. acgEdg = [acgT-acgBinSz/2; acgT(end)+acgBinSz/2]';
  288. % Plot ACG
  289. f1 = figure('position',[200 200 400 200]);
  290. subplot(1,2,1); hold on;
  291. histogram('BinEdges',acgEdg,'BinCounts',...
  292. cell_metrics.general.ccg(:,UID,UID),...
  293. 'FaceColor',color,'FaceAlpha',1,'LineStyle','none');
  294. title(sprintf('%s (%i)',cell_metrics.putativeCellType{UID},UID));
  295. xlabel('Time (s)');
  296. yticks([]);
  298. % Plot Raster
  299. subplot(1,2,2);
  300. bz_eventRasterplot(spikes,opto_20ms.timestamps(1:10000,:),UID,'periEvtWdw',.02,...
  301. 'color',color,'subsample',5);
  303. sgtitle(cell_metrics.general.basename);
  305. % Save figure if desired
  306. if printfig
  307. pos = get(f1,'Position');
  308. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  309. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  310. print(f1,sprintf('%s-UID %i-20ms',cell_metrics.general.basename,UID),'-dpdf','-r0')
  311. end
  313. clear acgT acgBinSz acgEdg color UID printfig pos f1 session basepath basename...
  314. cell_metrics spikes opto_20ms
  316. %% 1c(2) Single cell example opto response & ACG (ChR - SST)
  317. % Load cell_metrics, spikes, opto_1sec
  318. session = 'YutaMouse20_140325';
  319. color = [0 .72 .53];
  320. UID = 6; % YutaMouse20-140325-UID 6-10ms
  321. printfig = true;
  323. % Load stuff
  324. basepath = strrep(local2remote(session),'_','-'); % bugfix, yuta sessions have '-'
  325. basename = bz_BasenameFromBasepath(basepath);
  326. cell_metrics = load_cellmetrics_TH('basepath',basepath);
  327. spikes = load_spikes_TH('basepath',basepath);
  328. load([basepath,filesep,basename,'.opto_10ms.manipulation.mat'],'opto_10ms');
  330. % ACG
  331. acgT = cell_metrics.general.ccg_time;
  332. acgBinSz = mean(diff(acgT));
  333. acgEdg = [acgT-acgBinSz/2; acgT(end)+acgBinSz/2]';
  335. % Plot ACG
  336. f1 = figure('position',[200 200 400 200]);
  337. subplot(1,2,1); hold on;
  338. histogram('BinEdges',acgEdg,'BinCounts',...
  339. cell_metrics.general.ccg(:,UID,UID),...
  340. 'FaceColor',color,'FaceAlpha',1,'LineStyle','none');
  341. title(sprintf('%s (%i)',cell_metrics.putativeCellType{UID},UID));
  342. xlabel('Time (s)');
  343. yticks([]);
  345. % Plot Raster
  346. subplot(1,2,2);
  347. bz_eventRasterplot(spikes,opto_10ms.timestamps,UID,'periEvtWdw',.02,...
  348. 'color',color,'subsample',5);
  350. sgtitle(cell_metrics.general.basename);
  352. % Save figure if desired
  353. if printfig
  354. pos = get(f1,'Position');
  355. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  356. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  357. print(f1,sprintf('%s-UID %i-10ms',cell_metrics.general.basename,UID),'-dpdf','-r0')
  358. end
  360. clear acgT acgBinSz acgEdg color UID printfig pos f1 session cell_metrics basepath...
  361. spikes opto_10ms
  363. %% 1d PSTH for PV/SST ChR/eNpHR responses 10/20 ms pulses
  364. baseline = [-Inf -.002]; % [-Inf 0] standard, [4 40] for freq
  365. minRate = 0; minPrctile = 1; % Minimum percentile of bins during baseline above minRate
  366. nGr = 4;
  367. printfig = false;
  369. % EVENT SELECTION
  370. type = 'manipulations'; % Manipulations, events
  371. field = {'opto_20ms','opto_20ms','opto_10ms','opto_10ms'};
  372. sortCriterion = '_modulationIndex'; % '_modulationSD' (default) or '_peakFreq' for chirp
  374. % CELL SELECTION
  375. labels = {'PV-eNpHR','SST-eNpHR','PV-ChR','SST-ChR'};
  376. include = struct('groundTruthClassification',{'PV+','SST+','PV+','SST+'},...
  377. 'labels',{'eNpHR','eNpHR','ChR2','ChR2'},...
  378. 'brainRegion',{'DG','DG','DG','DG'});
  379. exclude = struct('putativeCellType',repmat({'Unknown'},1,nGr),...
  380. 'tags',repmat({{'InverseSpike','Bad'}},1,nGr));
  382. % PLOT
  383. f1 = figure('position',[200 20 130 165*nGr]); % Same abs height
  384. for gr = 1:4
  385. subplot(4,1,gr);
  386. plot_psth(batch_metrics,field{gr},'type',type,'labels',labels(gr),...
  387. 'include',include(gr),'exclude',exclude(gr),...
  388. 'colors',[0 0 0],'minRate',minRate,'minPrctile',minPrctile,...
  389. 'sortCriterion',sortCriterion,'baseline',baseline,'SDrange',[-5 5],...
  390. 'x_units','ms');
  391. end
  393. % Save figure if desired
  394. if printfig
  395. pos = get(f1,'Position');
  396. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  397. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  398. print(f1,'Opto_pulse_psth.pdf','-dpdf','-r0')
  399. end
  401. clear include exclude nGr gr field type labels minRate minPrctile baseline...
  402. sortCriterion printfig f1 pos colors
  404. %% 1e Waveforms superimposed
  405. include = struct('putativeCellType',{'','','Granule Cell','Mossy Cell'},...
  406. 'groundTruthClassification',{'PV+','SST+','',''},...
  407. 'brainRegion',{'DG','DG','DG','DG'});
  408. exclude = struct('putativeCellType',repmat({'Unknown'},1,4),...
  409. 'tags',repmat({{'InverseSpike','Bad'}},1,4),...
  410. 'groundTruthClassification',{'~','~','+','+'});
  411. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7];
  412. printfig = false;
  414. % Get z-scored waveforms for all cells:
  415. stdTS = batch_metrics.waveforms.time{1};
  416. for n = length(class_metrics.UID):-1:1
  417. % Get z-scored waveforms with standard timestamps
  418. waveforms(n,:) = zscore(interp1(batch_metrics.waveforms.time{n},...
  419. batch_metrics.waveforms.filt{n},stdTS));
  420. % % Filtered waveforms in microvolt
  421. % waveforms(n,:) = interp1(class_metrics.waveforms.time{n},...
  422. % class_metrics.waveforms.filt{n},stdTS);
  423. end
  425. % Plot all mean waveforms and SEM
  426. f1 = figure('position',[100 100 170 200]); hold on
  427. %figure('position',[100 100 800 200]); -- Toggle this for separate plots
  428. for ct = 1:length(include)
  429. % subplot(1,length(sel),ct);-- Toggle this for separate plots
  430. [~,cellIDs] = select_cells(batch_metrics,'include',include(ct),...
  431. 'exclude',exclude(ct));
  433. shadedErrorBar(stdTS, nanmean(waveforms(cellIDs,:)),...
  434. std(waveforms(cellIDs,:))./sqrt(length(cellIDs)),...
  435. 'lineProps',{'color',colors(ct,:)});
  437. % Label with N;
  438. %unitstr = sprintf('N = %i',length(cellIDs));
  439. %text(prctile(stdTS,10),-3.5,unitstr,'Color','k');
  441. %title(labels{sel(ct)});
  442. xlabel('Time (ms)');
  443. xlim([min(stdTS) max(stdTS)]);
  444. ylim([-4 2]);
  445. if ct == 1; ylabel('Amplitude (normalized)'); else; yticks([]); end
  446. end
  448. % Save figure if desired
  449. if printfig
  450. pos = get(f1,'Position');
  451. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  452. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  453. print(f1,'Spike-waveforms.pdf','-dpdf','-r0')
  454. end
  456. clear waveforms stdTS sel ct n cellIDs include exclude printfig f1 colors pos
  458. %% 1f,h, S1d-h ACG rise time, CV2, spike width (troughtopeak), spike Asymmetry (AB)
  459. printfig = false;
  461. % Parameter selection
  462. fields = {'troughToPeak','acg_tau_rise','cv2','ab_ratio',...
  463. 'firingRate','ds2amplitude','wakeNREMratio'}; % 1e-h
  464. transformations = {'','log10','','','log10','','log2'}; % 1e-h
  466. % Cell selection
  467. labels = {'PV+','SST+','GC','MC'};
  468. include = struct('putativeCellType',{'','','Granule Cell','Mossy Cell'},...
  469. 'groundTruthClassification',{'PV+','SST+','',''},...
  470. 'brainRegion',{'DG','DG','DG','DG'});
  471. exclude = struct('putativeCellType',repmat({'Unknown'},1,4),...
  472. 'tags',repmat({{'InverseSpike','Bad'}},1,4),...
  473. 'groundTruthClassification',{'','','+','+'});
  474. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7];
  476. % Plot
  477. f1 = figure('position',[50 450 155*length(fields) 115]); % Same aspect ratio
  478. %f1 = figure('position',[50 450 240*length(fields) 190]); % Same aspect ratio
  480. for ff = 1:length(fields)
  481. figure(f1);
  482. subplot(1,length(fields),ff);
  483. [~,~,values] = plot_metricfield(batch_metrics, fields{ff},'labels',labels,...
  484. 'include',include,'exclude',exclude,'tukeyFence',3,...
  485. 'colors',colors,'transformation',transformations{ff});
  486. set(gca,'fontsize',6);
  488. fprintf('\nDG statistics for %s:\n',fields{ff});
  489. %kruskalwallisNdunns(catuneven(values,NaN),labels);
  490. onewayGroupTest(catuneven(values,NaN),labels);
  492. % Nested glmm statistics
  493. fprintf('\nDG mixed-effects model statistics for %s:\n',fields{ff});
  495. valTbl = get_metricfieldTbl(class_metrics, fields{ff}, 'include',include,...
  496. 'exclude',exclude, 'labels',labels, 'transformation',transformations{ff});
  497. glme_stats(valTbl)
  499. fprintf('\n ====================================== \n\n');
  500. end
  502. % Save figure if desired
  503. if printfig
  504. pos = get(f1,'Position');
  505. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  506. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  507. print(f1,'Fig1fhS1d-h_CellParameters.pdf','-dpdf','-r0')
  508. end
  510. clear include exclude thiscriteria values field transformation sel f1 printfig...
  511. pos ff fields labels transformations
  513. %% 1g ACG superimposed
  514. include = struct('putativeCellType',{'','','Granule Cell','Mossy Cell'},...
  515. 'groundTruthClassification',{'PV+','SST+','',''},...
  516. 'brainRegion',{'DG','DG','DG','DG'});
  517. exclude = struct('putativeCellType',repmat({'Unknown'},1,4),...
  518. 'tags',repmat({{'InverseSpike','Bad'}},1,4),...
  519. 'groundTruthClassification',{'~','~','+','+'});
  520. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7];
  522. % Get z-scored waveforms for all cells:
  523. ccgts = batch_metrics.general.batch{1}.acgs.log10;
  524. acgs = batch_metrics.acg.log10_rate'; % .log10_rate = peak normalized, .log10 = firing rate (Hz)
  526. % Plot all mean waveforms and SEM
  527. figure('position',[100 100 200 200]); hold on
  528. %figure('position',[100 100 800 200]); -- Toggle this for separate plots
  529. for ct = 1:length(include)
  530. % subplot(1,length(sel),ct);-- Toggle this for separate plots
  531. [~,cellIDs] = select_cells(batch_metrics,'include',include(ct),...
  532. 'exclude',exclude(ct));
  534. shadedErrorBar(ccgts, nanmean(acgs(cellIDs,:)),...
  535. std(acgs(cellIDs,:))./sqrt(length(cellIDs)),...
  536. 'lineProps',{'color',colors(ct,:)});
  538. set(gca,'XScale','log');
  539. xlabel('Time (s)');
  540. xlim([0 1]);
  541. end
  543. clear waveforms stdTS sel ct n cellIDs include exclude colors ccgts acgs
  545. %% 1i TSNE for SVM classified metrics (Include Unknown)
  546. % To change TSNE parameters in cell explorer, console '>> edit CellExplorer_Preferences'
  547. fields = {'cv2','acg_tau_rise','troughToPeak','ab_ratio','firingRate','ds2amplitude','wakeNREMratio'};
  548. % MANU: firing irregularity (CV2), autocorrelogram (ACG) rise time, spike width, anatomical location, waveform asymmetry and firing rate
  549. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .7 .7 .7; .56 .39 .93; 0 .72 .53];
  551. % Cell selection
  552. include = struct(...
  553. 'putativeCellType',{'PV interneuron','SST interneuron','Granule Cell','Mossy Cell','Unknown','',''},...
  554. 'groundTruthClassification',{'','','','','','PV+','SST+'});
  555. [include.brainRegion] = deal('DG');
  556. [include.geneticLine] = deal('Cre'); % 'Cre' for Fig1j, 'Wild type' For supplementary Fig1
  558. % Remove Inverse spike and Bad waveform metrics
  559. plot_metrics = select_cells(class_metrics,...
  560. 'include',struct('brainRegion','DG'),...
  561. 'exclude',struct('tags',{{'Bad','InverseSpike'}}));
  563. clear values
  564. for ff = length(fields):-1:1
  565. values(:,ff) = plot_metrics.(fields{ff});
  566. end
  568. Y = tsne(values,'standardize',true);
  570. % Bugfix: Remove cells from batch that don't have all features
  571. plot_metrics.tags((any(isnan(values),2))) = {'incomplete'};
  572. plot_metrics = select_cells(plot_metrics,'exclude',struct('tags','incomplete'));
  574. figure('position',[100 100 280 260]); hold on
  576. for ct = 1:length(include)
  577. [~,cellIDs] = select_cells(plot_metrics,'include',include(ct));
  579. if any(ct==[6 7])
  580. % Use different markers for ChR and HR
  581. chrIDs = intersect(find(contains(plot_metrics.labels,'ChR')),cellIDs);
  582. hrIDs = intersect(find(contains(plot_metrics.labels,'eNpHR')),cellIDs);
  584. scatter(Y(chrIDs,1),Y(chrIDs,2),36,colors(ct,:),'+',...
  585. 'LineWidth',.5,'MarkerFaceAlpha',1);
  586. scatter(Y(hrIDs,1),Y(hrIDs,2),36,colors(ct,:),'x',...
  587. 'LineWidth',.5,'MarkerFaceAlpha',1);
  588. else
  589. scatter(Y(cellIDs,1),Y(cellIDs,2),12,colors(ct,:),'o',...
  590. 'filled','MarkerFaceAlpha',.25);
  591. end
  592. end
  594. clear fields values sel include exclude ct markers cellIDs chrIDs colors hrIDs...
  595. plot_metrics Y
  597. %% Supplementary Fig1b: Burst-index vs. ACG-tau (PC-IN classification)
  598. % Peter CE IN criteria: acg_tau_decay > 30ms; acg_tau_rise>3; troughToPeak <= 0.425ms
  599. fields = {'troughToPeak','acg_tau_rise','acg_tau_decay'}; % 'acg_tau_rise','acg_tau_decay','burstIndex_Royer2012';
  600. colors = [.46 .29 .83; 0 .72 .53]; % Opto only
  602. % Basic cell selection criteria
  603. include = struct('brainRegion',{'DG'});
  604. exclude = struct('tags',{{'Bad','InverseSpike'}},'putativeCellType','Unknown');
  606. f1 = figure('position',[200 200 330 350]);
  608. % PC AND INTERNEURONS (DG ONLY)
  609. include.putativeCellType = 'Cell';
  610. [~,PCids] = select_cells(batch_metrics,'include',include,'exclude',exclude);
  611. % scatter(batch_metrics.(fields{1})(selected), batch_metrics.(fields{2})(selected), 10,...
  612. % 'ro','filled','MarkerFaceAlpha',.2);
  613. scatter3(batch_metrics.(fields{1})(PCids), batch_metrics.(fields{3})(PCids),...
  614. batch_metrics.(fields{2})(PCids), 10, 'ro','filled','MarkerFaceAlpha',.2);
  615. hold on
  617. include.putativeCellType = 'Interneuron';
  618. [~,INids] = select_cells(batch_metrics,'include',include,'exclude',exclude);
  619. % scatter(batch_metrics.(fields{1})(selected), batch_metrics.(fields{2})(selected), 10,...
  620. % 'bo','filled','MarkerFaceAlpha',.2);
  621. scatter3(batch_metrics.(fields{1})(INids), batch_metrics.(fields{3})(INids),...
  622. batch_metrics.(fields{2})(INids), 10, 'bo','filled','MarkerFaceAlpha',.2);
  624. % OPTO TYPES
  625. include.putativeCellType = '';
  626. include.groundTruthClassification = 'PV+';
  627. [~,PVids] = select_cells(batch_metrics,'include',include,'exclude',exclude);
  628. %scatter(batch_metrics.(fields{1})(selected), batch_metrics.(fields{2})(selected),100,'k+');
  629. scatter3(batch_metrics.(fields{1})(PVids), batch_metrics.(fields{3})(PVids),...
  630. batch_metrics.(fields{2})(PVids), 50, colors(1,:), '+');
  632. include.groundTruthClassification = 'SST+';
  633. [~,SSTids] = select_cells(batch_metrics,'include',include,'exclude',exclude);
  634. %scatter(batch_metrics.(fields{1})(selected), batch_metrics.(fields{2})(selected),100,'kx');
  635. scatter3(batch_metrics.(fields{1})(SSTids), batch_metrics.(fields{3})(SSTids),...
  636. batch_metrics.(fields{2})(SSTids), 50, colors(2,:), '+');
  638. set(gca,'Yscale','log');
  639. set(gca,'Zscale','log');
  640. set(f1,'renderer','painters');
  642. xlabel('Spike-half-width (ms)');
  643. ylabel('ACG decay-constant (ms)');
  644. zlabel('ACG rise-constant (ms)');
  645. clear pcind inind
  647. % PIE CHARTS
  648. figure('position',[100 50 600 300]);
  649. subplot(1,2,1);
  650. pie([length(intersect(PVids,INids)) length(intersect(PVids,PCids))]);
  651. ax = gca;
  652. ax.Colormap = [0 0 1; 1 0 0];
  653. title('PV-Cre Opto+')
  655. subplot(1,2,2);
  656. pie([length(intersect(SSTids,INids)) length(intersect(SSTids,PCids))]);
  657. ax = gca;
  658. ax.Colormap = [0 0 1; 1 0 0];
  659. title('SST-Cre Opto+')
  661. legend({'Interneuron','Principal Cell'});
  662. fprintf('N PV cells: %i, N SST cells: %i\n',length(PVids),length(SSTids))
  664. clear include exclude criteria PVids SSTids PCids INids colors ax f1 fields
  666. %% Supplementary Fig 1c: GC/MC classification
  667. % Run YutClassifier (see preparations above)
  669. %% Supplementary Fig 1i: TSNE WT only
  670. % See Fig 1i above
  673. %% Figure 2
  675. %% 2a Overall Firing rates for class_metrics (violin)
  676. % + GLME stats (20251206)
  677. % + Split IN plots (opto vs. non-opto)
  679. % Plot parameters
  680. fields = {'firingRate'};
  681. transformations = {'log10'}; % set to {''} for mean +/- SEM values in console
  682. printfig = false;
  684. % Cell selection
  685. labels = {'PV','SST','GC','MC','PV+','SST+',};
  686. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  687. 'Granule Cell','Mossy Cell','',''},'groundTruthClassification',{'','','','','PV+','SST+'});
  688. [include.brainRegion] = deal('DG');
  689. exclude = struct('tags',{{'Bad','InverseSpike'}});
  690. criteria = struct('UID',repmat({[0 Inf]},1,6));
  691. %[criteria.radialRatePeak] = deal([1 Inf]); % Specific exclusion criteria here
  692. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  695. for ff = 1:length(fields)
  696. % PLOTTING
  697. fi{ff} = figure('position',[200*ff 450 115 115]); % Same aspect ratio
  699. [~, valMtrx, ~] = plot_metricfield(class_metrics,fields{ff},...
  700. 'include',include,'exclude',exclude,'criteria',criteria,...
  701. 'transformation',transformations{ff},'tukeyFence',3,'labels',labels,...
  702. 'colors',colors,'splitgroups',[5 1; 6 2]);
  703. set(gca,'fontsize',6);
  705. % Save figure
  706. if printfig
  707. pos = get(fi{ff},'Position');
  708. set(fi{ff},'PaperPositionMode','Auto','PaperUnits','points',...
  709. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  710. print(fi{ff},sprintf('%s_svm_opto.pdf',fields{ff}),'-dpdf','-r0')
  711. end
  713. % ANOVA for SVM classified neurons
  714. fprintf('\nDG statistics for %s, SVM types:\n',fields{ff});
  715. %kruskalwallisNdunns(valMtrx(:,1:4),labels(1:4));
  716. onewayGroupTest(valMtrx(:,1:4),labels(1:4));
  718. % ANOVA for optotagged cells only
  719. if ~all(isnan(valMtrx(:,5:6)),'all')
  720. fprintf('\nDG statistics for %s, opto vs GC/MC:\n',fields{ff});
  721. onewayGroupTest(valMtrx(:,[5 6 3 4]),labels([5 6 3 4]));
  722. %kruskalwallisNdunns(valMtrx(:,[5 6 3 4]),labels([5 6 3 4]));
  723. end
  725. % LMM nested data stats
  726. valTbl = get_metricfieldTbl(class_metrics, fields{ff}, 'include',include,...
  727. 'exclude',exclude, 'criteria',criteria, 'labels',labels);
  729. glme_stats(valTbl)
  730. end
  732. % Print mean +/- SEM values in console
  733. for ct = 1:6
  734. % fprintf('%s firing rate: %.2f +/- %.2f (mean +/- STD)\n',labels{ct},...
  735. % nanmean(valMtrx(:,ct)), nanstd(valMtrx(:,ct)))
  736. fprintf('%s firing rate: Median %.2f (%.2f - %.2f IQR)\n',labels{ct},...
  737. 10.^nanmedian(valMtrx(:,ct)), 10.^prctile(valMtrx(:,ct),25), 10.^prctile(valMtrx(:,ct),75))
  738. end
  741. clear include exclude criteria valGrp valMtrx valCell fields transformation...
  742. sel fi printfig ff colors labels pos transformations
  744. %% 2b Incl Run/Rest Brain state dependent firing rates class_metrics (PSTH + boxplot)
  745. printfig = false;
  747. % Create in- and exclude structure
  748. labels = {'PV','SST','GC','MC'};
  749. limrate = [50 25 1.5 4]; % Hard coding here, no easy fix...
  750. include = struct('putativeCellType',{'PV interneuron','SST interneuron','Granule Cell','Mossy Cell'},...
  751. 'brainRegion',{'DG','DG','DG','DG'});
  752. exclude = struct('putativeCellType',repmat({'Unknown'},1,4),...
  753. 'tags',repmat({{'InverseSpike','Bad'}},1,4));
  754. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7];
  756. f1 = figure('position',[100 100 1250 200]);
  758. for ct = 1:4
  759. [~,cellIDs] = select_cells(class_metrics,'include',include(ct),'exclude',exclude(ct));
  761. rates = cat(1,class_metrics.firingRateRun(cellIDs),class_metrics.firingRateWake(cellIDs),...
  762. class_metrics.firingRateNREM(cellIDs),class_metrics.firingRateREM(cellIDs));
  763. rates = rates(:,~any(isnan(rates),1));
  764. figure(f1)
  765. ax(ct) = subplot(1,4,ct); hold(ax(ct),'on');
  767. % Make rates heatmap in background
  768. [~,order] = sort(rates(1,:));
  769. imagesc(1:4, linspace(0, limrate(ct), size(rates,2)), rates(:,order)');
  770. colormap(bwyr);
  771. colorbar
  772. clim([0 limrate(ct)]);
  774. % Plot boxes
  775. boxplot(rates','symbol', '');
  776. h = findobj('LineStyle','--'); set(h, 'LineStyle','-');
  777. h = findobj(gca,'Tag','Box');
  778. for j=1:length(h)
  779. patch(get(h(j),'XData'),get(h(j),'YData'),colors(ct,:),'FaceAlpha',.7);
  780. end
  782. % Markup
  783. title(labels{ct});
  784. ylabel('Firing Rate (Hz)');
  785. xticklabels({'Run','Quiet','NREM','REM'});
  786. xtickangle(45);
  787. ylim([0 limrate(ct)]);
  789. % Stats (repeated measures ANOVA on ranks)
  790. fprintf([labels{ct} '\n'])
  791. onewayGroupTest(rates',{'Run','Quiet','NREM','REM'});
  792. end
  794. if printfig
  795. pos = get(f1,'Position');
  796. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  797. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  798. print(f1,sprintf('StateRates_psth.pdf'),'-dpdf','-r0')
  799. end
  801. clear ct h field j rates f1 wi order pos printfig ax cellIDs colors exclude include labels
  803. %% Suppl 2a,b: NREM-Wake, REM-Wake ratios
  804. % Plot parameters
  805. fields = {'NREMwakeRatio','REMwakeRatio'};
  806. transformations = {'log2','log2'};
  807. printfig = false;
  809. % Cell selection
  810. labels = {'PV','SST','GC','MC','PV+','SST+',};
  811. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  812. 'Granule Cell','Mossy Cell','',''},'groundTruthClassification',{'','','','','PV+','SST+'});
  813. [include.brainRegion] = deal('DG');
  814. exclude = struct('tags',{{'Bad','InverseSpike'}});
  815. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  818. for ff = 1:length(fields)
  819. % PLOTTING
  820. fi{ff} = figure('position',[200*ff 450 135 135]); % Same aspect ratio
  822. [~, valMtrx, ~] = plot_metricfield(class_metrics,fields{ff},...
  823. 'include',include,'exclude',exclude,...
  824. 'transformation',transformations{ff},'tukeyFence',3,'labels',labels,...
  825. 'colors',colors,'markercolors',[repmat([.2 .2 .2],6,1)],... % 'markercolors',[repmat([.2 .2 .2],4,1); colors([1 2],:)]
  826. 'splitgroups',[5 1; 6 2]);
  827. set(gca,'fontsize',6);
  828. plot([0 length(include)+1],[0 0],':','color',[.5 .5 .5]);
  830. % Save figure
  831. if printfig
  832. pos = get(fi{ff},'Position');
  833. set(fi{ff},'PaperPositionMode','Auto','PaperUnits','points',...
  834. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  835. print(fi{ff},sprintf('%s_svm_opto.pdf',fields{ff}),'-dpdf','-r0')
  836. end
  838. % ANOVA for SVM classified neurons
  839. fprintf('\nDG statistics for %s, SVM types:\n',fields{ff});
  840. onewayGroupTest(valMtrx(:,1:4),labels(1:4));
  842. % ANOVA for optotagged cells only
  843. if ~all(isnan(valMtrx(:,5:6)),'all')
  844. fprintf('\nDG statistics for %s, opto vs GC/MC:\n',fields{ff});
  845. onewayGroupTest(valMtrx(:,[5 6 3 4]),labels([5 6 3 4]));
  846. end
  848. % LMM nested data stats
  849. valTbl = get_metricfieldTbl(class_metrics, fields{ff}, 'include',include,...
  850. 'exclude',exclude, 'labels',labels);
  852. glme_stats(valTbl)
  853. end
  855. clear include exclude criteria valGrp valMtrx valCell fields transformation...
  856. sel fi printfig ff colors labels pos transformations valTbl
  859. %% Figure 3
  861. %% 3a Plot all inputs to given target w/ place fields
  862. post = 8570; % Global ID (get e.g. from connectivity plot below)
  863. pairs = class_metrics.putativeConnections.excitatory(...
  864. class_metrics.putativeConnections.excitatory(:,2) == post,:);
  865. printfig = false;
  867. for pp = 1:size(pairs,1)
  868. f1 = figure('position',[10*pp+100 200 500 700]);
  869. plot_pairWplacefields(pairs(pp,1),pairs(pp,2),class_metrics);
  870. drawnow;
  872. % Save figure
  873. if printfig
  874. pos = get(f1,'Position');
  875. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  876. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  877. print(f1,sprintf('%i-%i_connExample.pdf',pairs(pp,1),pairs(pp,2)),'-dpdf','-r0')
  878. end
  879. end
  881. clear post pairs fs pos
  883. %% 3b Plot probe layout w/ classified units, connections, etc
  884. makeConnectivityPlot(class_metrics,'batchID',97,'showIDs',''); % 'global','local'
  886. %% 3c CCG PSTH
  887. % TODO: Milliseconds and % for transmission probability
  888. % TODO: class_metrics.general.batch{1, 1}.ccg , .ccg_times could be used
  889. % instead getting around the whole mono_analysis thing!
  890. pairlabels = {'GC-PV','GC-SST','MC-PV','MC-SST','MEC-PV','MEC-SST'};
  891. prelabels = {'GC','MC','MEC'};
  892. postlabels = {'PV','SST'};
  893. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; 1 0 0; 0 0 1];
  894. printfig = true;
  896. preinclude = struct('putativeCellType',{'Granule Cell','Mossy Cell','Pyramidal Cell'},...
  897. 'brainRegion',{'DG','DG','ENTm'});
  898. postinclude = struct('putativeCellType',{'PV','SST'},'brainRegion',{'DG','DG'}); % Consider adding opto types extra
  900. % Sanity check, local ID's must be identical for pairs
  901. globIDs = class_metrics.putativeConnections.excitatory;
  902. locIDs = class_metrics.UID(class_metrics.putativeConnections.excitatory);
  903. if ~all(globIDs == mono_analysis.UIDs,'all')
  904. warning('Mismatch between batch_metrics and mono_analysis, do not continue!');
  905. return
  906. end
  908. f1 = figure('Position',[100 10 150 980]); hold on; sp = 0;
  909. % Plot PSTH
  910. for pre = 1:length(preinclude)
  911. [~,presel] = select_cells(class_metrics,'include',preinclude(pre));
  912. for post = 1:length(postinclude)
  913. sp = sp+1;
  915. % Get PSTH
  916. [~,postsel] = select_cells(class_metrics,'include',postinclude(post));
  917. isRightPair = ismember(globIDs(:,1),presel) & ismember(globIDs(:,2),postsel);
  918. values = mono_analysis.latency_allspk(isRightPair);
  919. [~,order] = sort(values,'ascend');
  920. PairIDs = find(isRightPair);
  921. thisccg = 100*(mono_analysis.CCG_allspk(PairIDs(order),:) -...
  922. mono_analysis.pred_allspk(PairIDs(order),:)); % Convert to percent
  923. ccgts = mono_analysis.ccgTimeStamps * 1000; % Convert to ms
  925. subplot(6,1,sp);
  926. plim = [-1 2];
  928. % Plot PSTH
  929. imagesc(ccgts, linspace(plim(1), plim(2), size(thisccg,1)), thisccg);
  930. set(gca,'YDir','normal')
  931. xlim([-2 6]);
  932. xlabel('Time (ms)');
  933. colormap(bwyr);
  934. clim([plim(1) plim(2)]);
  936. % Overlay mean +/- SEM
  937. shadedErrorBar(ccgts, nanmean(thisccg,1), nanstd(thisccg,[],1)./sqrt(size(thisccg,1)),...
  938. 'lineProps',{'-','color',[0 0 0]});
  940. % Label with N;
  941. text(prctile(.002,55), plim(1)+ (plim(2)-plim(1))/10,...
  942. sprintf('N = %i',size(thisccg,1)),'Color','k')
  943. ylabel('Transm prob (%)')
  945. title(pairlabels{sp});
  946. drawnow
  947. end
  948. end
  950. % Save figure if desired
  951. if printfig
  952. pos = get(f1,'Position');
  953. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  954. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  955. print(f1,'MonoSynCCG_psth.pdf','-dpdf','-r0')
  956. end
  958. clear globIDs locIDs pre post presel postsel iRightPair f1 sp values ccgts colors...
  959. isRightPair order PairIDs pairlabels plim postinclude postlabels preinclude...
  960. prelabels thisccg f1 pos printfig
  962. %% 3d PV and SST input fractions and stats (svm and opto separately!)
  963. % Run section 'Get connected and non-connected...' above (preparations) first!
  964. % TODO: Consider: Opto+ cells should be separate plot (supplement) as the fractions
  965. % of pairs per session depends on the number of post in session, ergo not
  966. % same distribution
  967. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; 1 0 0; 0 0 1];
  968. printfig = false;
  970. % Include only sessions w/ manually corrected monosynaptic connections
  971. % Technically unnecesary, this is already tested when mono_res are created
  972. nEcon = mono_res.nEcon(:,:,mono_res.monoSynManual);
  973. nEtotal = mono_res.nEtotal(:,:,mono_res.monoSynManual);
  975. % Plot MEC,GC,MC connections to svm PV/SST IN
  976. EinRates = nEcon([3 4 5],[1 2],:)./nEtotal([3 4 5],[1 2],:) * 100;
  977. EinRatesPl = reshape(permute(EinRates,[2 1 3]),6,[]);
  979. % Rates for GC,MC connections to opto PV/SST IN
  980. oEinRates = nEcon([3 4],[6 7],:)./nEtotal([3 4],[6 7],:) * 100;
  981. oEinRatesPl = reshape(permute(oEinRates,[2 1 3]),4,[]);
  983. % Get numbers for console
  984. EconTot = sum(nEcon,3);
  985. Etot = sum(nEtotal,3);
  986. ErateTot = EconTot./Etot *100;
  989. % MAIN FIGURE (3d)
  990. f1 = figure('position',[100 50 190 220]);
  991. % Violin plots
  992. distributionPlot(mat2cell(EinRatesPl,ones(6,1)),...
  993. 'color',mat2cell(colors([3 3 4 4 5 5],:),ones(6,1)),'showMM',6,'distWidth',.9,...
  994. 'addSpread',0,'histOpt',1,'divFactor',2,'FaceAlpha',.5);%,'histOri','left','widthDiv',[2 1]);
  996. plotSpread(mat2cell(EinRatesPl,ones(6,1)),...
  997. 'spreadFcn',{'lin',10},'spreadWidth',.9,...
  998. 'distributionColors',{[0 0 0]},...%mat2cell(colors([3 3 4 4],:),ones(4,1)),...
  999. 'distributionMarkerSize',5,'distributionMarkerAlpha',.5,'binWidth',.05);
  1001. plotSpread(mat2cell(oEinRatesPl,ones(4,1)),...
  1002. 'spreadFcn',{'lin',10},'spreadWidth',.9,...
  1003. 'distributionColors',{[0 0 0]},...%mat2cell(colors([3 3 4 4],:),ones(4,1)),...
  1004. 'distributionMarkerSize',2,'distributionMarkerAlpha',.5,...
  1005. 'binWidth',.05,'distributionMarkers','+');
  1007. xticks(1:6);
  1008. xticklabels({'GC-DGPV','GC-DGSST','MC-DGPV','MC-DGSST','MEC-DGPV','MEC-DGSST'});
  1009. xtickangle(45);
  1010. ylim([0 55]);
  1011. ylabel('% connected');
  1013. % SEPARATE FIGURE OPTO ONLY FOR SUPPLEMENT (S3d)
  1014. f2 = figure('position',[100 50 140 220]);
  1015. % Violin plots
  1016. distributionPlot(mat2cell(oEinRatesPl,ones(4,1)),...
  1017. 'color',mat2cell(colors([3 3 4 4],:),ones(4,1)),'showMM',6,'distWidth',.9,...
  1018. 'addSpread',0,'histOpt',1,'divFactor',2,'FaceAlpha',.5);
  1020. plotSpread(mat2cell(oEinRatesPl,ones(4,1)),...
  1021. 'spreadFcn',{'lin',10},'spreadWidth',.9,...
  1022. 'distributionColors',{[0 0 0]},...%mat2cell(colors([3 3 4 4],:),ones(4,1)),...
  1023. 'distributionMarkerSize',2,'distributionMarkerAlpha',.5,...
  1024. 'binWidth',.05,'distributionMarkers','+');
  1026. xticks(1:4);
  1027. xticklabels({'GC-DGPV','GC-DGSST','MC-DGPV','MC-DGSST'});
  1028. xtickangle(45);
  1029. ylim([0 55]);
  1030. ylabel('% connected');
  1033. % STATS FOR SVM CELL TYPES
  1034. % Print total con/total numbers and stats (ranksum) in console
  1035. fprintf('GC to DG-PV: %i/%i (%.2f%%), to DG-SST %i/%i (%.2f%%)\n',...
  1036. EconTot(3,1),Etot(3,1),ErateTot(3,1),EconTot(3,2),Etot(3,2),ErateTot(3,2));
  1037. pvals(1) = unpairedSampleTest(EinRatesPl(1,:),EinRatesPl(2,:),'labels',{'PV','SST'},'verbose',true);
  1039. fprintf('MC to DG-PV: %i/%i (%.2f%%), to DG-SST %i/%i (%.2f%%)\n',...
  1040. EconTot(4,1),Etot(4,1),ErateTot(4,1),EconTot(4,2),Etot(4,2),ErateTot(4,2));
  1041. pvals(2) = unpairedSampleTest(EinRatesPl(3,:),EinRatesPl(4,:),'labels',{'PV','SST'},'verbose',true);
  1043. fprintf('MEC to DG-PV: %i/%i (%.2f%%), to DG-SST %i/%i (%.2f%%)\n',...
  1044. EconTot(5,1),Etot(5,1),ErateTot(5,1),EconTot(5,2),Etot(5,2),ErateTot(5,2));
  1045. pvals(3) = unpairedSampleTest(EinRatesPl(5,:),EinRatesPl(6,:),'labels',{'PV','SST'},'verbose',true);
  1047. % Multiple-comparison correction
  1048. pvals = multCompCorr(pvals,'method','Dunn-Sidak');
  1049. fprintf('Corrected p-values (Dunn-Sidak)\nGC PV vs. SST: %.4g\nMC PV vs. SST: %.4g\nMEC PV vs. SST: %.4g\n',pvals)
  1050. clear pvals
  1053. % STATS FOR OPTO-TAGGED CELL TYPES
  1054. fprintf('Opto: GC to DG-PV: %i/%i (%.2f%%), to DG-SST %i/%i (%.2f%%)\n',...
  1055. EconTot(3,6),Etot(3,6),ErateTot(3,6),EconTot(3,7),Etot(3,7),ErateTot(3,7));
  1056. pvals(1) = unpairedSampleTest(oEinRatesPl(1,:),oEinRatesPl(2,:),'labels',{'PV','SST'},'verbose',true); % [sic!], EinRates is only 4 categories
  1058. fprintf('Opto: MC to DG-PV: %i/%i (%.2f%%), to DG-SST %i/%i (%.2f%%)\n',...
  1059. EconTot(4,6),Etot(4,6),ErateTot(4,6),EconTot(4,7),Etot(4,7),ErateTot(4,7));
  1060. pvals(2) = unpairedSampleTest(oEinRatesPl(3,:),oEinRatesPl(4,:),'labels',{'PV','SST'},'verbose',true);
  1062. % Multiple-comparison correction (opto)
  1063. pvals = multCompCorr(pvals,'method','Dunn-Sidak');
  1064. fprintf('Corrected p-values (Dunn-Sidak) opto\nGC PV vs. SST: %.4g\nMC PV vs. SST: %.4g\n',pvals)
  1067. % SAVE FIGURE
  1068. if printfig
  1069. pos = get(f1,'Position');
  1070. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1071. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1072. print(f1,'ConnectivityRates.pdf','-dpdf','-r0')
  1074. pos = get(f2,'Position');
  1075. set(f2,'PaperPositionMode','Auto','PaperUnits','points',...
  1076. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1077. print(f2,'ConnectivityRatesOptoOnly.pdf','-dpdf','-r0')
  1078. end
  1080. clear EconTot Etot ErateTot EinRates EinRatesPl nEcon nEtotal f1 pos printfig...
  1081. pvals f2
  1083. %% 3e, S3e: Synaptic latencies
  1084. % TODO: Make function to load synaptic results aligned w/ batch_metrics,
  1085. % for now just do sanity check.
  1086. printfig = false;
  1087. field = 'latency_allspk'; % 'peak_allspk','latency_allspk'
  1088. ylims = [0 5]; %[0 .005],[0 .4]
  1089. ylab = 'Latency (ms)'; %'Latency (s)','Transm prob'
  1090. labels = {'GC-PV','GC-SST','MC-PV','MC-SST','MEC-PV','MEC-SST'}; % 'o' = opto! --
  1091. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; 1 0 0; 0 0 1];
  1094. preinclude = struct(...
  1095. 'putativeCellType',{'Granule Cell','Mossy Cell','Pyramidal Cell'},...
  1096. 'brainRegion',{'DG','DG','ENTm'});
  1098. postinclude = struct(...
  1099. 'putativeCellType',{'PV','SST','',''},...
  1100. 'brainRegion',{'DG','DG','DG','DG'},...
  1101. 'groundTruthClassification',{'','','PV+','SST+'});
  1103. % Sanity check, local ID's must be identical for pairs
  1104. globIDs = class_metrics.putativeConnections.excitatory;
  1105. locIDs = class_metrics.UID(class_metrics.putativeConnections.excitatory);
  1106. if ~all(globIDs == mono_analysis.UIDs,'all')
  1107. warning('Mismatch between batch_metrics and mono_analysis, do not continue!');
  1108. return
  1109. end
  1111. % Get latencies for desired pair types
  1112. for pre = 1:length(preinclude)
  1113. [~,presel] = select_cells(class_metrics,'include',preinclude(pre));
  1114. for post = 1:length(postinclude)
  1115. [~,postsel] = select_cells(class_metrics,'include',postinclude(post));
  1116. isRightPair = ismember(globIDs(:,1),presel) & ismember(globIDs(:,2),postsel);
  1117. values{pre,post} = mono_analysis.(field)(isRightPair)*1000;
  1118. end
  1119. end
  1121. % FIGURE (SVM)
  1122. f1 = figure('position',[100 50 250 300]);
  1123. plotSpread(mat2cell(catuneven(reshape(values(:,[1 2])',6,1),NaN)',ones(6,1)),...
  1124. 'spreadFcn',{'xp',[]},'spreadWidth',.9,'distributionMarkerAlpha',.3,... %'spreadFcn',{'lin',75}
  1125. 'distributionColors',{[0 0 0]},...
  1126. 'distributionMarkerSize',5,'binWidth',.05);
  1128. plotSpread(mat2cell(catuneven(reshape(values(:,[3 4])',6,1),NaN)',ones(6,1)),...
  1129. 'spreadFcn',{'xp',[]},'spreadWidth',.9,'distributionMarkerAlpha',.3,... % 'spreadFcn',{'lin',10}
  1130. 'distributionColors',{[0 0 0]},...
  1131. 'distributionMarkerSize',4,'binWidth',.05,'distributionMarkers','+');
  1133. customErrorBar(catuneven(reshape(values(:,[1 2])',6,1),NaN), 'errtype','STD',...
  1134. 'colors',colors([3 3 4 4 5 5],:), 'MarkerSize',100, 'alpha',.8, 'LineWidth',1.5);
  1136. xticklabels(labels);
  1137. xtickangle(45);
  1138. ylim(ylims);
  1139. ylabel(ylab);
  1141. % FIGURE (OPTO) - S3e
  1142. f2 = figure('position',[100 50 180 300]);
  1143. plotSpread(mat2cell(catuneven(reshape(values([1 2],[3 4])',4,1),NaN)',ones(4,1)),...
  1144. 'spreadFcn',{'xp',[]},'spreadWidth',.9,'distributionMarkerAlpha',.3,... % 'spreadFcn',{'lin',10}
  1145. 'distributionColors',{[0 0 0]},...
  1146. 'distributionMarkerSize',4,'binWidth',.05,'distributionMarkers','+');
  1148. customErrorBar(catuneven(reshape(values([1 2],[3 4])',4,1),NaN), 'errtype','STD',...
  1149. 'colors',colors([3 3 4 4],:), 'MarkerSize',100, 'alpha',.8, 'LineWidth',1.5);
  1151. xticklabels(labels);
  1152. xtickangle(45);
  1153. ylim(ylims);
  1154. ylabel(ylab);
  1156. % STATS
  1157. % Print mean +/- STD in console per cell type
  1158. fprintf('%s values:\n',field);
  1159. for pre = 1:length(preinclude)
  1160. for post = [1 2]
  1161. fprintf('%s: %.2f +/- %.2f\n',labels{post+(pre-1)*2},...
  1162. nanmean(values{pre,post}), nanstd(values{pre,post}))
  1163. end
  1164. end
  1166. % Separate ANOVA for opto and SVM identified interneurons!
  1167. fprintf('\nDG statistics for %s, SVM types:\n',field);
  1168. kruskalwallisNdunns(catuneven(reshape(values(:,[1 2])',6,1),NaN),labels);
  1170. fprintf('\nDG statistics for %s, opto vs GC/MC:\n',field);
  1171. kruskalwallisNdunns(catuneven(reshape(values([1 2],[3 4])',4,1),NaN),...
  1172. labels(1:4)); % Exclude [3] pre, there are no opto in MEC mice
  1174. % Save figure if desired
  1175. if printfig
  1176. pos = get(f1,'Position');
  1177. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1178. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1179. print(f1,'Fig3e_SynapticLatencies.pdf','-dpdf','-r0')
  1181. pos = get(f2,'Position');
  1182. set(f2,'PaperPositionMode','Auto','PaperUnits','points',...
  1183. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1184. print(f2,'FigS3e_SynapticLatencies_opto.pdf','-dpdf','-r0')
  1185. end
  1187. clear locIDs globIDs presel postsel preinclude postinclude labels isRightPair...
  1188. values f1 post pre printfig pos f1 isRightPair colors labels postinclude...
  1189. preinclude ylab ylims field
  1191. %% 3f, S3f: Transmission probability
  1192. % TODO: Make function to load synaptic results aligned w/ batch_metrics,
  1193. % for now just do sanity check.
  1195. field = 'peak_allspk'; % 'peak_allspk','latency_allspk'
  1196. ylims = [-3 -.5]; %[0 .005],[0 .4]
  1197. ylab = 'Transm prob (%)'; %'Latency (s)','Transm prob'
  1198. labels = {'GC-PV','GC-SST','MC-PV','MC-SST','MEC-PV','MEC-SST'}; % 'o' = opto! --
  1199. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; 1 0 0; 0 0 1];
  1200. printfig = true;
  1202. preinclude = struct(...
  1203. 'putativeCellType',{'Granule Cell','Mossy Cell','Pyramidal Cell'},...
  1204. 'brainRegion',{'DG','DG','ENTm'});
  1206. postinclude = struct(...
  1207. 'putativeCellType',{'PV','SST','',''},...
  1208. 'brainRegion',{'DG','DG','DG','DG'},...
  1209. 'groundTruthClassification',{'','','PV+','SST+'});
  1211. % Sanity check, local ID's must be identical for pairs
  1212. globIDs = class_metrics.putativeConnections.excitatory;
  1213. locIDs = class_metrics.UID(class_metrics.putativeConnections.excitatory);
  1214. if ~all(globIDs == mono_analysis.UIDs,'all')
  1215. warning('Mismatch between batch_metrics and mono_analysis, do not continue!');
  1216. return
  1217. end
  1219. % Get Transmission probabilities for desired pair types
  1220. for pre = 1:length(preinclude)
  1221. [~,presel] = select_cells(class_metrics,'include',preinclude(pre));
  1222. for post = 1:length(postinclude)
  1223. [~,postsel] = select_cells(class_metrics,'include',postinclude(post));
  1224. isRightPair = ismember(globIDs(:,1),presel) & ismember(globIDs(:,2),postsel);
  1226. % Get log10 values
  1227. theseVal = log10(mono_analysis.(field)(isRightPair));
  1228. theseVal = real(theseVal(~isnan(theseVal) & ~isinf(theseVal) & ~imag(theseVal)>0));
  1229. values{pre,post} = theseVal;
  1230. clear theseVal
  1231. end
  1232. end
  1234. % FIGURE (SVM)
  1235. f1 = figure('position',[100 50 190 220]);
  1236. plotSpread(mat2cell(catuneven(reshape(values(:,[1 2])',6,1),NaN)',ones(6,1)),...
  1237. 'spreadFcn',{'xp',[]},'spreadWidth',.9,'distributionMarkerAlpha',.3,...
  1238. 'distributionColors',{[0 0 0]},...
  1239. 'distributionMarkerSize',5,'binWidth',.05);
  1241. plotSpread(mat2cell(catuneven(reshape(values(:,[3 4])',6,1),NaN)',ones(6,1)),...
  1242. 'spreadFcn',{'xp',[]},'spreadWidth',.9,'distributionMarkerAlpha',.3,... % 'spreadFcn',{'lin',10}
  1243. 'distributionColors',{[0 0 0]},...
  1244. 'distributionMarkerSize',4,'binWidth',.05,'distributionMarkers','+');
  1246. customErrorBar(catuneven(reshape(values(:,[1 2])',6,1),NaN), 'errtype','STD',...
  1247. 'colors',colors([3 3 4 4 5 5],:), 'MarkerSize',100, 'alpha',.8, 'LineWidth',1.5);
  1249. yticks([-3 -2 -1]);
  1250. yticklabels([.1 1 10]);
  1251. xticklabels(labels);
  1252. xtickangle(45);
  1253. ylim(ylims);
  1254. ylabel(ylab);
  1256. % FIGURE (OPTO)
  1257. f2 = figure('position',[100 50 125 220]);
  1258. plotSpread(mat2cell(catuneven(reshape(values([1 2],[3 4])',4,1),NaN)',ones(4,1)),...
  1259. 'spreadFcn',{'xp',[]},'spreadWidth',.9,'distributionMarkerAlpha',.3,... % 'spreadFcn',{'lin',10}
  1260. 'distributionColors',{[0 0 0]},...
  1261. 'distributionMarkerSize',4,'binWidth',.05,'distributionMarkers','+');
  1263. customErrorBar(catuneven(reshape(values([1 2],[3 4])',4,1),NaN), 'errtype','STD',...
  1264. 'colors',colors([3 3 4 4 5 5],:), 'MarkerSize',100, 'alpha',.8, 'LineWidth',1.5);
  1266. yticks([-3 -2 -1]);
  1267. yticklabels([.1 1 10]);
  1268. xticklabels(labels);
  1269. xtickangle(45);
  1270. ylim(ylims);
  1271. ylabel(ylab);
  1273. % Print mean +/- STD in console per cell type
  1274. fprintf('%s values:\n',field);
  1275. for pre = 1:length(preinclude)
  1276. for post = [1 2]
  1277. fprintf('%s: %.2f +/- %.2f\n',labels{post+(pre-1)*2},...
  1278. 100*nanmean(10.^values{pre,post}), 100*nanstd(10.^values{pre,post}))
  1279. end
  1280. end
  1282. % Separate ANOVA for opto and SVM identified interneurons!
  1283. fprintf('\nDG statistics for %s, SVM types:\n',field);
  1284. kruskalwallisNdunns(catuneven(reshape(values(:,[1 2])',6,1),NaN),labels);
  1286. fprintf('\nDG statistics for %s, opto vs GC/MC:\n',field);
  1287. kruskalwallisNdunns(catuneven(reshape(values([1 2],[3 4])',4,1),NaN),...
  1288. labels(1:4)); % Exclude [3] pre, there are no opto in MEC mice
  1290. % Save figure if desired
  1291. if printfig
  1292. pos = get(f1,'Position');
  1293. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1294. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1295. print(f1,'Fig3f_TransmissionProbabilities.pdf','-dpdf','-r0')
  1297. pos = get(f2,'Position');
  1298. set(f2,'PaperPositionMode','Auto','PaperUnits','points',...
  1299. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1300. print(f2,'FigS3f_TransmissionProbabilities_opto.pdf','-dpdf','-r0')
  1301. end
  1303. clear locIDs globIDs presel postsel preinclude postinclude labels isRightPair...
  1304. values f1 post pre printfig pos f1 isRightPair colors labels postinclude...
  1305. preinclude ylab ylims field
  1307. %% 3g, S3g: Synaptic facilitation
  1308. burstSpk = 2; % 2 or 3, normalized to first
  1309. ylims = [-3 3]; %[0 .005],[0 .4]
  1310. ylab = 'Facilitation'; %'Latency (s)','Transm prob'
  1311. labels = {'GC-PV','GC-SST','MC-PV','MC-SST','MEC-PV','MEC-SST'}; % 'o' = opto! --
  1312. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; 1 0 0; 0 0 1];
  1313. printfig = false;
  1315. preinclude = struct(...
  1316. 'putativeCellType',{'Granule Cell','Mossy Cell','Pyramidal Cell'},...
  1317. 'brainRegion',{'DG','DG','ENTm'});
  1319. postinclude = struct(...
  1320. 'putativeCellType',{'PV','SST','',''},...
  1321. 'brainRegion',{'DG','DG','DG','DG'},...
  1322. 'groundTruthClassification',{'','','PV+','SST+'});
  1324. % Sanity check, local ID's must be identical for pairs
  1325. globIDs = class_metrics.putativeConnections.excitatory;
  1326. locIDs = class_metrics.UID(class_metrics.putativeConnections.excitatory);
  1327. if ~all(globIDs == mono_analysis.UIDs,'all')
  1328. warning('Mismatch between batch_metrics and mono_analysis, do not continue!');
  1329. return
  1330. end
  1332. % Get facilitation for desired pair types
  1333. for pre = 1:length(preinclude)
  1334. [~,presel] = select_cells(class_metrics,'include',preinclude(pre));
  1335. for post = 1:length(postinclude)
  1336. [~,postsel] = select_cells(class_metrics,'include',postinclude(post));
  1337. isRightPair = ismember(globIDs(:,1),presel) & ismember(globIDs(:,2),postsel);
  1339. % Calculate facilitation
  1340. rawTransm{pre,post,1} = mono_analysis.peak(isRightPair,1);
  1341. rawTransm{pre,post,2} = mono_analysis.peak(isRightPair,burstSpk);
  1342. theseval = rawTransm{pre,post,2}./rawTransm{pre,post,1};
  1343. theseval = theseval(theseval>0 & ~isinf(theseval)); % Remove negative transm prob
  1344. values{pre,post} = log2(theseval); % log2 transformation
  1345. end
  1346. end
  1348. % FIGURE 3G (SVM)
  1349. f1 = figure('position',[100 50 190 220]);
  1350. plot([0 7],[0 0],':','Color',[.5 .5 .5]);
  1351. plotSpread(mat2cell(catuneven(reshape(values(:,[1 2])',6,1),NaN)',ones(6,1)),...
  1352. 'spreadFcn',{'xp',[]},'spreadWidth',.9,...
  1353. 'distributionColors',[.5 .5 .5],'distributionMarkerSize',1,...
  1354. 'distributionMarkerAlpha',.5,'distributionMarkers','o');
  1356. plotSpread(mat2cell(catuneven(reshape(values(:,[3 4])',6,1),NaN)',ones(6,1)),...
  1357. 'spreadFcn',{'xp',[]},'spreadWidth',.9,...
  1358. 'distributionColors',[.5 .5 .5],...
  1359. 'distributionMarkerSize',4,'binWidth',.05,'distributionMarkers','+');
  1361. customErrorBar(catuneven(reshape(values(:,[1 2])',6,1),NaN), 'errtype','STD',...
  1362. 'colors',colors([3 3 4 4 5 5],:), 'MarkerSize',100, 'alpha',.8, 'LineWidth',1.5);
  1364. xticklabels(labels);
  1365. xtickangle(45);
  1366. ylim(ylims);
  1367. yticks(ylims(1):ylims(2));
  1368. yticklabels(round(2.^(ylims(1):ylims(2)),3));
  1369. ylabel(ylab);
  1371. % FIGURE S3G (Opto)
  1372. f2 = figure('position',[100 50 125 220]);
  1373. plot([0 5],[0 0],':','Color',[.5 .5 .5]);
  1374. plotSpread(mat2cell(catuneven(reshape(values([1 2],[3 4])',4,1),NaN)',ones(4,1)),...
  1375. 'spreadFcn',{'xp',[]},'spreadWidth',.9,...
  1376. 'distributionColors',[.5 .5 .5],...
  1377. 'distributionMarkerSize',4,'binWidth',.05,'distributionMarkers','+');
  1379. customErrorBar(catuneven(reshape(values([1 2],[3 4])',4,1),NaN), 'errtype','STD',...
  1380. 'colors',colors([3 3 4 4],:), 'MarkerSize',100, 'alpha',.8, 'LineWidth',1.5);
  1382. xticklabels(labels(1:4));
  1383. xtickangle(45);
  1384. ylim(ylims);
  1385. yticks(ylims(1):ylims(2));
  1386. yticklabels(round(2.^(ylims(1):ylims(2)),3));
  1387. ylabel(ylab);
  1390. % STATS
  1391. % Separate ANOVA for opto and SVM identified interneurons!
  1392. fprintf('\nDG statistics for facilitation spike no %i, SVM types:',burstSpk);
  1393. kruskalwallisNdunns(catuneven(reshape(values(:,[1 2])',6,1),NaN),labels);
  1395. fprintf('\nDG statistics for facilitation spike no %i, opto vs GC/MC:',burstSpk);
  1396. kruskalwallisNdunns(catuneven(reshape(values([1 2],[3 4])',4,1),NaN),...
  1397. labels(1:4)); % Exclude [3] pre, there are no opto in MEC mice
  1399. % Pairwise comparisons for faciliation degree
  1400. fprintf('\n\nSVM cell types:');
  1401. for pre = 1:3
  1402. for post = [1 2]
  1403. if length(rawTransm{pre,post,1})<2; continue; end
  1404. fprintf('\n%s, pairwise comparison spike 1 vs. %i:\n',...
  1405. labels{2*(pre-1)+post},burstSpk)
  1406. pval = pairedSampleTest(rawTransm{pre,post,1},rawTransm{pre,post,2},...
  1407. 'labels',{'spike1',sprintf('spike%i',burstSpk)},'verbose',true);
  1408. % pvals = 1 - (1 - pvals) .^ N; Dunn-Sidak
  1409. fprintf('Corrected p (Dunn-Sidak): %.4g\n',1 - (1-pval)^5);
  1410. end
  1411. end
  1413. % Pairwise comparisons for faciliation (opto only)
  1414. fprintf('\n\nOpto cell types:');
  1415. for pre = 1:2
  1416. for post = [3 4]
  1417. if length(rawTransm{pre,post,1})<2; continue; end
  1418. fprintf('\n%s, pairwise comparison spike 1 vs. %i (opto):\n',...
  1419. labels{2*(pre-1)+post-2},burstSpk)
  1420. pval = pairedSampleTest(rawTransm{pre,post,1},rawTransm{pre,post,2},...
  1421. 'labels',{'spike1',sprintf('spike%i',burstSpk)},'verbose',true);
  1422. fprintf('Corrected p (Dunn-Sidak): %.4g\n',1 - (1-pval)^4);
  1423. end
  1424. end
  1426. % Save figure if desired
  1427. if printfig
  1428. pos = get(f1,'Position');
  1429. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1430. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1431. print(f1,'Fig3g_SynapticFacilitationLog2.pdf','-dpdf','-r0')
  1433. pos = get(f2,'Position');
  1434. set(f2,'PaperPositionMode','Auto','PaperUnits','points',...
  1435. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1436. print(f2,'FigS3g_SynapticFacilitationLog2_opto.pdf','-dpdf','-r0')
  1437. end
  1439. clear locIDs globIDs presel postsel preinclude postinclude labels isRightPair...
  1440. values f1 post pre ylab ylims pval pp rawTransm theseval
  1442. %% S3a: Synaptic facilitation (CCG)
  1443. prelabels = {'GC','MC','MEC'};
  1444. postlabels = {'PV','SST'};
  1445. printfig = true;
  1447. preinclude = struct('putativeCellType',{'Granule Cell','Mossy Cell','Pyramidal Cell'},...
  1448. 'brainRegion',{'DG','DG','ENTm'});
  1449. postinclude = struct('putativeCellType',{'PV','SST'},'brainRegion',{'DG','DG'}); % Consider adding opto types extra
  1451. % Sanity check, local ID's must be identical for pairs
  1452. globIDs = class_metrics.putativeConnections.excitatory;
  1453. locIDs = class_metrics.UID(class_metrics.putativeConnections.excitatory);
  1454. if ~all(globIDs == mono_analysis.UIDs,'all')
  1455. warning('Mismatch between batch_metrics and mono_analysis, do not continue!');
  1456. return
  1457. end
  1459. f1 = figure('Position',[100 10 500 980]); hold on; sp = 0;
  1460. % Plot PSTH
  1461. for pre = 1:length(preinclude)
  1462. [~,presel] = select_cells(class_metrics,'include',preinclude(pre));
  1463. for post = 1:length(postinclude)
  1465. % Get Proper IDs
  1466. [~,postsel] = select_cells(class_metrics,'include',postinclude(post));
  1467. isRightPair = ismember(globIDs(:,1),presel) & ismember(globIDs(:,2),postsel);
  1468. values = mono_analysis.peak(isRightPair);
  1469. [~,order] = sort(values,'ascend');
  1470. PairIDs = find(isRightPair);
  1472. for bb = 1:3
  1473. sp = sp+1;
  1474. thisccg = (mono_analysis.CCG(PairIDs(order),:,bb) -...
  1475. mono_analysis.pred(PairIDs(order),:,bb)) .* 100; % Convert to percent
  1476. ccgts = mono_analysis.ccgTimeStamps * 1e3; % Convert to ms
  1478. subplot(6,3,sp);
  1479. plim = [-1 3];
  1481. % Plot PSTH
  1482. imagesc(ccgts, linspace(plim(1), plim(2), size(thisccg,1)), thisccg);
  1483. set(gca,'YDir','normal')
  1484. xlim([-2 6]);
  1485. xlabel('Time (ms)');
  1486. colormap(bwyr);
  1487. clim([plim(1) plim(2)]);
  1489. % Overlay mean +/- SEM
  1490. shadedErrorBar(ccgts, nanmean(thisccg,1), nanstd(thisccg,[],1)./sqrt(size(thisccg,1)),...
  1491. 'lineProps',{'-','color',[0 0 0]});
  1493. % Label with N;
  1494. text(prctile(.002,55), plim(1)+ (plim(2)-plim(1))/10,...
  1495. sprintf('N = %i',size(thisccg,1)),'Color','k')
  1496. ylabel('Transm prob (%)')
  1497. drawnow
  1498. end
  1499. end
  1500. end
  1502. % Save figure if desired
  1503. if printfig
  1504. pos = get(f1,'Position');
  1505. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1506. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1507. print(f1,'Burst_transmission_psth.pdf','-dpdf','-r0')
  1508. end
  1510. clear globIDs locIDs pre post presel postsel iRightPair f1 sp values...
  1511. printfig pos f1 ccgts thisccg isRightPair order PairIDs plim...
  1512. postinclude postlabels preinclude prelabels bb
  1514. %% S3b: Connectivity table
  1515. confrct = round(sum(mono_res.nEcon,3)./sum(mono_res.nEtotal,3)*100,2);
  1516. array2table(confrct,'RowNames',labels,'VariableNames',labels)
  1518. % Make heatmap
  1519. figure('position',[100 100 220 175]); heatmap(log10(confrct(1:5,1:5)));
  1520. colormap(bwyr); caxis([-1 .5]); % For color
  1522. figure('position',[100 100 220 175]); heatmap(confrct(1:5,1:5)); % For values
  1524. % Plot raw numbers / table to console for text
  1525. fprintf('Number connected pairs, rows = pre, columns = post\n')
  1526. array2table(sum(mono_res.nEcon,3),'RowNames',labels,'VariableNames',labels)
  1528. fprintf('Total number simult recorded pairs, rows = pre, columns = post\n')
  1529. array2table(sum(mono_res.nEtotal,3),'RowNames',labels,'VariableNames',labels)
  1531. %% S3c: Facilitation over initial strength
  1532. burstSpk = 2; % 2 or 3, normalized to first
  1533. labels = {'PV','SST','GC','MC','MEC PC'}; % 'o' = opto! --
  1534. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; 1 0 0; 0 0 1];
  1535. printfig = true;
  1537. preinclude = struct(...
  1538. 'putativeCellType',{'Granule Cell','Mossy Cell','Pyramidal Cell'},...
  1539. 'brainRegion',{'DG','DG','ENTm'});
  1541. postinclude = struct(...
  1542. 'putativeCellType',{'PV','SST'},...
  1543. 'brainRegion',{'DG','DG'},...
  1544. 'groundTruthClassification',{'',''});
  1546. % Sanity check, local ID's must be identical for pairs
  1547. globIDs = class_metrics.putativeConnections.excitatory;
  1548. locIDs = class_metrics.UID(class_metrics.putativeConnections.excitatory);
  1549. if ~all(globIDs == mono_analysis.UIDs,'all')
  1550. warning('Mismatch between batch_metrics and mono_analysis, do not continue!');
  1551. return
  1552. end
  1554. % Get facilitation for desired pair types
  1555. f1 = figure('position',[100 100 300 500]);
  1556. for pre = 1:length(preinclude)
  1557. [~,presel] = select_cells(class_metrics,'include',preinclude(pre));
  1558. for post = 1:length(postinclude)
  1559. [~,postsel] = select_cells(class_metrics,'include',postinclude(post));
  1560. isRightPair = ismember(globIDs(:,1),presel) & ismember(globIDs(:,2),postsel);
  1562. % Calculate facilitation
  1563. rawTransm{pre,post,1} = mono_analysis.peak(isRightPair,1);
  1564. rawTransm{pre,post,2} = mono_analysis.peak(isRightPair,burstSpk);
  1565. theseval = rawTransm{pre,post,2}./rawTransm{pre,post,1};
  1567. init{pre,post} = rawTransm{pre,post,1}(theseval>0 & ~isinf(theseval));
  1568. theseval = theseval(theseval>0 & ~isinf(theseval)); % Remove negative transm prob
  1569. fac{pre,post} = log2(theseval); % log2 transformation
  1571. % Scatterplot initial strength vs facilitation
  1572. subplot(3,2,post + (pre-1)*2); hold on
  1573. scatter(init{pre,post}, fac{pre,post}, 4, colors(pre+2,:),...
  1574. 'o','filled','MarkerFaceAlpha',.3);
  1575. title(sprintf('%s-%s',labels{pre+2},labels{post}));
  1577. % Pearson's R & Linear regression
  1578. if ~isempty(init{pre,post})
  1580. % Linear regression
  1581. mxx = max(init{pre,post});
  1582. mny = min(fac{pre,post});
  1583. P = polyfit(init{pre,post},fac{pre,post},1); % m=P(1), b=P(2),
  1584. plot([0 mxx], [P(2) P(2)+P(1)*mxx],'k:');
  1586. % Pearson's R
  1587. [R,pval] = corr(init{pre,post}, fac{pre,post});
  1588. text(.5*mxx, .7*mny, sprintf(...
  1589. 'R = %.2f\np = %.2g',R,pval),'FontSize',8);
  1590. fprintf('%s-%s: R = %.2f p = %.3g, N = %i\n',labels{pre+2},...
  1591. labels{post},R,pval,length(init{pre,post}))
  1592. end
  1593. end
  1594. end
  1596. % Save figure if desired
  1597. if printfig
  1598. pos = get(f1,'Position');
  1599. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1600. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1601. print(f1,'Facilitation_vs_initialStrength.pdf','-dpdf','-r0')
  1602. end
  1604. clear locIDs globIDs presel postsel preinclude postinclude labels isRightPair...
  1605. values f1 post pre ylab ylims pval pp rawTransm theseval init fac...
  1606. printfig pos f1
  1608. %% S3d-g: Opto only
  1609. % -- see above for 3d-g
  1612. %% Figure 4
  1614. %% Load place fields:
  1615. linear = load_LinearBatch(batch_metrics);
  1616. openfield = load_OpenfieldBatch(batch_metrics);
  1617. radial = load_RadialBatch(batch_metrics);
  1619. %% 4a(1) Linear place field examples
  1620. [~,UIDs] = select_cells(class_metrics,'include',...
  1621. struct('putativeCellType','SST interneuron','groundTruthClassification',{'SST+'}),...
  1622. 'criteria',struct('SIsec1D',[0 Inf]));
  1623. color = [0 .72 .53]; % [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7]
  1624. printfig = false;
  1625. %UIDs = 5521; printfig = true; % Example on the figure, comment out to cycle through others
  1627. for n = 1:length(UIDs)
  1628. batchID = class_metrics.batchIDs(UIDs(n));
  1629. localUID = class_metrics.UID(UIDs(n));
  1630. basepath = class_metrics.general.path{batchID};
  1631. placefields = get_placeFieldsLinear('basepath',basepath);
  1633. f1 = figure('position',[200 200 180 300]); %f1 = figure('position',[200 200 160 350]);
  1634. subplot(2,1,1);
  1635. plot_ACG(class_metrics,UIDs(n),'color',color);
  1637. subplot(2,1,2);
  1638. plot_placefield1D(localUID, placefields, 'xdim','distance');
  1640. title(sprintf('%s-UID%i',class_metrics.sessionName{UIDs(n)},localUID));
  1641. drawnow
  1642. end
  1644. % Save figure
  1645. if printfig
  1646. pos = get(f1,'Position');
  1647. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1648. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1649. print(f1,sprintf('LinearPFexample-%s-UID%i.pdf',...
  1650. class_metrics.sessionName{UIDs(n)},localUID),'-dpdf','-r0')
  1651. end
  1653. clear basepath radial batchID localUID f1 pos printfig UIDs color placefields n
  1655. %% 4a(2) + S4a-c Linear Spatial tuning parameters
  1656. % Plot parameters
  1657. fields = {'SIspk1D','SIsec1D','SpCoh1D','SesStab1D'}; %3b-c -- CONSIDER ADDING Min peak rate to criteria
  1658. transformations = {'log10','log10','',''};
  1659. printfig = true;
  1661. % Cell selection
  1662. labels = {'PV','SST','GC','MC','PV+','SST+',};
  1663. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  1664. 'Granule Cell','Mossy Cell','',''},'groundTruthClassification',{'','','','','PV+','SST+'});
  1665. exclude = struct('tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  1666. criteria = struct('linearRateMean',repmat({[.5 Inf]},1,6));
  1667. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  1669. fi = figure('position',[125 450 145*length(fields) 125]);
  1671. for ff = 1:length(fields)
  1672. % PLOTTING
  1673. figure(fi);
  1674. subplot(1,length(fields),ff);
  1676. [~, valMtrx, ~] = plot_metricfield(class_metrics,fields{ff},...
  1677. 'include',include,'exclude',exclude,'criteria',criteria,...
  1678. 'transformation',transformations{ff},'tukeyFence',3,'labels',labels,...
  1679. 'colors',colors,'splitgroups',[5 1; 6 2]);
  1680. set(gca,'fontsize',6);
  1682. % ANOVA for SVM classified neurons
  1683. fprintf('\nDG statistics for %s, SVM types:',fields{ff});
  1684. kruskalwallisNdunns(valMtrx(:,1:4),labels(1:4));
  1686. % ANOVA for optotagged cells only
  1687. if ~all(isnan(valMtrx(:,5:6)),'all')
  1688. fprintf('\nDG statistics for %s, opto vs GC/MC:',fields{ff});
  1689. kruskalwallisNdunns(valMtrx(:,[5 6 3 4]),labels([5 6 3 4]));
  1690. end
  1692. % LMM nested data stats (SVM & Opto separately)
  1693. valTbl = get_metricfieldTbl(class_metrics, fields{ff}, 'include',include,...
  1694. 'exclude',exclude, 'criteria',criteria, 'labels',labels,...
  1695. 'transformation',transformations{ff});
  1696. %glme_stats(valTbl)
  1698. fprintf('\nDG LMM statistics for %s, SVM types:',fields{ff});
  1699. isSVM = cellfun(@(x) (any(x=={'PV','SST','GC','MC'})), table2cell(valTbl(:,2)));
  1700. thisTbl = valTbl(isSVM,:);
  1701. thisTbl.Group = removecats(thisTbl.Group); % Remove unused categories
  1702. glme_stats(thisTbl);
  1704. fprintf('\nDG LMM statistics for %s, opto vs GC/MC:',fields{ff});
  1705. isOpto = cellfun(@(x) (any(x=={'PV+','SST+','GC','MC'})), table2cell(valTbl(:,2)));
  1706. thisTbl = valTbl(isOpto,:);
  1707. thisTbl.Group = removecats(thisTbl.Group); % Remove unused categories
  1708. glme_stats(thisTbl);
  1709. end
  1711. % Save figure
  1712. if printfig
  1713. pos = get(fi,'Position');
  1714. set(fi,'PaperPositionMode','Auto','PaperUnits','points',...
  1715. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1716. print(fi,'Fig4a_Linear_spatialparams_svm_opto.pdf','-dpdf','-r0')
  1717. end
  1719. clear include exclude criteria valGrp valMtrx valCell fields transformation...
  1720. sel fi printfig ff colors labels pos transformations valTbl isSVM isOpto thisTbl
  1722. %% 4b(1) Radial place field examples
  1723. [~,UIDs] = select_cells(class_metrics,'include',...
  1724. struct('putativeCellType','PV interneuron'),...%'groundTruthClassification',{'SST+'}),...
  1725. 'criteria',struct('SIsecRadial',[0 Inf]));
  1726. color = [.56 .39 .93]; % [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7]
  1727. printfig = false;
  1728. UIDs = 5985; printfig = true; % Example on the figure, comment out to cycle through others
  1730. for n = 1:length(UIDs)
  1731. batchID = class_metrics.batchIDs(UIDs(n));
  1732. localUID = class_metrics.UID(UIDs(n));
  1733. basepath = class_metrics.general.path{batchID};
  1734. placefields = get_placeFieldsRadial('basepath',basepath);
  1736. f1 = figure('position',[200 200 180 350]);
  1737. subplot(2,1,1);
  1738. plot_ACG(class_metrics,UIDs(n),'color',color);
  1740. subplot(2,1,2);
  1741. plot_placefieldRAD(localUID,placefields);
  1743. title(sprintf('%s-UID%i',class_metrics.sessionName{UIDs(n)},localUID));
  1744. drawnow
  1745. end
  1747. % Save figure
  1748. if printfig
  1749. pos = get(f1,'Position');
  1750. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1751. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1752. print(f1,sprintf('RadialPFexample-%s-UID%i.pdf',...
  1753. class_metrics.sessionName{UIDs(n)},localUID),'-dpdf','-r0')
  1754. end
  1756. clear basepath radial batchID localUID f1 pos printfig UIDs color placefields n
  1758. %% 4b(2) + S4d-f Radial Spatial tuning parameters
  1759. % Plot parameters
  1760. fields = {'SIspkRadial','SIsecRadial','SpCohRadial','SesStabRadial'};
  1761. transformations = {'log10','log10','',''};
  1762. printfig = false;
  1764. % Cell selection
  1765. labels = {'PV','SST','GC','MC','PV+','SST+',};
  1766. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  1767. 'Granule Cell','Mossy Cell','',''},'groundTruthClassification',{'','','','','PV+','SST+'});
  1768. exclude = struct('tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  1769. criteria = struct('radialRateMean',repmat({[.5 Inf]},1,6));
  1770. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  1772. fi = figure('position',[125 450 145*length(fields) 125]);
  1774. for ff = 1:length(fields)
  1775. % PLOTTING
  1776. figure(fi);
  1777. subplot(1,length(fields),ff);
  1779. [~, valMtrx, ~] = plot_metricfield(class_metrics,fields{ff},...
  1780. 'include',include,'exclude',exclude,'criteria',criteria,...
  1781. 'transformation',transformations{ff},'tukeyFence',3,'labels',labels,...
  1782. 'colors',colors,'overlaygroups',[5 1; 6 2]);
  1783. set(gca,'fontsize',6);
  1785. % ANOVA for SVM classified neurons
  1786. fprintf('\nDG statistics for %s, SVM types:',fields{ff});
  1787. kruskalwallisNdunns(valMtrx(:,1:4),labels(1:4));
  1789. % ANOVA for optotagged cells only
  1790. if ~all(isnan(valMtrx(:,5:6)),'all')
  1791. fprintf('\nDG statistics for %s, opto vs GC/MC:',fields{ff});
  1792. kruskalwallisNdunns(valMtrx(:,[5 6 3 4]),labels([5 6 3 4]));
  1793. end
  1795. % LMM nested data stats (SVM & Opto separately)
  1796. valTbl = get_metricfieldTbl(class_metrics, fields{ff}, 'include',include,...
  1797. 'exclude',exclude, 'criteria',criteria, 'labels',labels,...
  1798. 'transformation',transformations{ff});
  1799. %glme_stats(valTbl)
  1801. fprintf('\nDG LMM statistics for %s, SVM types:',fields{ff});
  1802. isSVM = cellfun(@(x) (any(x=={'PV','SST','GC','MC'})), table2cell(valTbl(:,2)));
  1803. thisTbl = valTbl(isSVM,:);
  1804. thisTbl.Group = removecats(thisTbl.Group); % Remove unused categories
  1805. glme_stats(thisTbl);
  1806. end
  1808. % Save figure
  1809. if printfig
  1810. pos = get(fi,'Position');
  1811. set(fi,'PaperPositionMode','Auto','PaperUnits','points',...
  1812. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1813. print(fi,'Fig4b_Radial_spatialparams_svm_opto.pdf','-dpdf','-r0')
  1814. end
  1816. clear include exclude criteria valGrp valMtrx valCell fields transformation...
  1817. sel fi printfig ff colors labels pos transformations isSVM isOpto thisTbl valTbl
  1819. %% 4c(1) Openfield place field examples
  1820. [~,UIDs] = select_cells(class_metrics,'include',...
  1821. struct('putativeCellType','Mossy Cell'),...%'groundTruthClassification',{'SST+'}),...
  1822. 'criteria',struct('openfieldRatePeak',[1 Inf]));
  1823. color = [.9 0 .7]; % [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7]
  1824. printfig = false;
  1825. UIDs = 3292; printfig = true; % Example on the figure, comment out to cycle through others
  1827. for n = 1%:length(UIDs)
  1828. batchID = class_metrics.batchIDs(UIDs(n));
  1829. localUID = class_metrics.UID(UIDs(n));
  1830. basepath = class_metrics.general.path{batchID};
  1831. placefields = get_placeFieldsOF('basepath',basepath);
  1833. f1 = figure('position',[200 200 180 350]);
  1834. subplot(2,1,1);
  1835. plot_ACG(class_metrics,UIDs(n),'color',color);
  1837. subplot(2,1,2);
  1838. plot_placefield2D(localUID,placefields);
  1840. title(sprintf('%s-UID%i',class_metrics.sessionName{UIDs(n)},localUID));
  1841. drawnow
  1842. end
  1844. % Save figure
  1845. if printfig
  1846. pos = get(f1,'Position');
  1847. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1848. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1849. print(f1,sprintf('OpenfieldPFexample-%s-UID%i.pdf',...
  1850. class_metrics.sessionName{UIDs(n)},localUID),'-dpdf','-r0')
  1851. end
  1853. clear basepath radial batchID localUID f1 pos printfig UIDs color placefields n
  1855. %% 4c(2) + S4g-i Openfield Spatial tuning parameters
  1856. % Plot parameters
  1857. fields = {'SIspk2D','SIsec2D','SpCoh2D','SesStab2D'}; %3b-c -- CONSIDER ADDING Min peak rate to criteria
  1858. transformations = {'log10','log10','',''};
  1859. printfig = false;
  1861. % Cell selection
  1862. labels = {'PV','SST','GC','MC','PV+','SST+',};
  1863. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  1864. 'Granule Cell','Mossy Cell','',''},'groundTruthClassification',{'','','','','PV+','SST+'});
  1865. exclude = struct('tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  1866. criteria = struct('openfieldRateMean',repmat({[.5 Inf]},1,6));
  1867. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  1869. fi = figure('position',[125 450 145*length(fields) 125]);
  1871. for ff = 1:length(fields)
  1872. % PLOTTING
  1873. figure(fi);
  1874. subplot(1,length(fields),ff);
  1876. [~, valMtrx, ~] = plot_metricfield(class_metrics,fields{ff},...
  1877. 'include',include,'exclude',exclude,'criteria',criteria,...
  1878. 'transformation',transformations{ff},'tukeyFence',3,'labels',labels,...
  1879. 'colors',colors,'splitgroups',[5 1; 6 2]);
  1880. set(gca,'fontsize',6);
  1882. % ANOVA for SVM classified neurons
  1883. fprintf('\nDG statistics for %s, SVM types:',fields{ff});
  1884. kruskalwallisNdunns(valMtrx(:,1:4),labels(1:4));
  1886. % ANOVA for optotagged cells only
  1887. if ~all(isnan(valMtrx(:,5:6)),'all')
  1888. fprintf('\nDG statistics for %s, opto vs GC/MC:',fields{ff});
  1889. kruskalwallisNdunns(valMtrx(:,[5 6 3 4]),labels([5 6 3 4]));
  1890. end
  1892. % LMM nested data stats (SVM & Opto separately)
  1893. valTbl = get_metricfieldTbl(class_metrics, fields{ff}, 'include',include,...
  1894. 'exclude',exclude, 'criteria',criteria, 'labels',labels,...
  1895. 'transformation',transformations{ff});
  1896. %glme_stats(valTbl)
  1898. fprintf('\nDG LMM statistics for %s, SVM types:',fields{ff});
  1899. isSVM = cellfun(@(x) (any(x=={'PV','SST','GC','MC'})), table2cell(valTbl(:,2)));
  1900. thisTbl = valTbl(isSVM,:);
  1901. thisTbl.Group = removecats(thisTbl.Group); % Remove unused categories
  1902. glme_stats(thisTbl);
  1904. fprintf('\nDG LMM statistics for %s, opto vs GC/MC:',fields{ff});
  1905. isOpto = cellfun(@(x) (any(x=={'PV+','SST+','GC','MC'})), table2cell(valTbl(:,2)));
  1906. thisTbl = valTbl(isOpto,:);
  1907. thisTbl.Group = removecats(thisTbl.Group); % Remove unused categories
  1908. glme_stats(thisTbl);
  1909. end
  1911. % Save figure
  1912. if printfig
  1913. pos = get(fi,'Position');
  1914. set(fi,'PaperPositionMode','Auto','PaperUnits','points',...
  1915. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1916. print(fi,'Fig4c_Openfield_spatialparams_svm_opto.pdf','-dpdf','-r0')
  1917. end
  1919. clear include exclude criteria valGrp valMtrx valCell fields transformation...
  1920. sel fi printfig ff colors labels pos transformations isSVM isOpto thisTbl valTbl
  1922. %% 4d: Place fields of converging pre synaptic partners on same post
  1923. pre = 4; % 3 = GC, 4 = MC, 5 = MEC PC
  1924. post = 1; % 1 = PV, 2 = SST (svm)
  1925. fieldtype = 'linear'; % 'linear','radial','openfield'
  1926. minpre = 8; % minimum number of converging partners
  1927. minrate = 1;
  1928. ratetype = 'firingRatePeak'; % 'firingRatePeak','firingRateMean'
  1929. cell_types = {'PV','SST','Granule','Mossy','Pyramidal'}; % Do not use 3-5 as post!
  1930. printfig = false;
  1932. goodPost = find(contains(class_metrics.putativeCellType,cell_types{post})' &...
  1933. mono_res.nEinInd(:,pre) >= minpre);
  1934. % pairs = class_metrics.putativeConnections.excitatory(...
  1935. % class_metrics.putativeConnections.excitatory(:,2) == post,:);
  1937. for po = 1:length(goodPost)
  1938. clear placefields
  1939. postLocID = class_metrics.UID(goodPost(po));
  1940. pairs = mono_res.EinInd{goodPost(po),pre};
  1941. batchID = class_metrics.batchIDs(goodPost(po));
  1942. metPath = [class_metrics.general.basepaths{batchID},filesep,...
  1943. class_metrics.general.basenames{batchID} '.' fieldtype '.firingRateMap.mat'];
  1945. % Load stuff
  1946. if isfile(metPath); placefields = load(metPath,'linear'); placefields = placefields.linear; end
  1948. % Continue if no placefields exist
  1949. if ~exist('placefields','var'); continue; end
  1951. % Remove pairs with low rates
  1952. preLocIDs = class_metrics.UID(pairs(:,1));
  1953. isGoodPre = any(placefields.(ratetype)(:,preLocIDs) > minrate, 1);
  1954. pairs = pairs(isGoodPre, :);
  1955. preLocIDs = preLocIDs(isGoodPre);
  1957. % Continue if no placefields left
  1958. if size(pairs,1) < minpre; continue; end
  1960. f1 = figure('position',[50 50 300 65*size(pairs,1)]);
  1962. for pr = 1:size(pairs,1)
  1964. subplot(size(pairs,1),2,1 + (pr-1)*2);
  1965. plot_placefield1D(preLocIDs(pr), placefields,'xdim','distance');
  1966. xlabel('');
  1967. set(gca,'FontSize',6);
  1968. title(sprintf('%s %i',class_metrics.putativeCellType{pairs(pr,1)},pairs(pr,1)));
  1969. end
  1971. % Post: Linear PF
  1972. subplot(size(pairs,1),2,2);
  1973. plot_placefield1D(postLocID, placefields, 'xdim', 'distance')
  1974. title(sprintf('%s %i',class_metrics.putativeCellType{goodPost(po)},goodPost(po)));
  1975. xlabel('');
  1976. set(gca,'FontSize',6);
  1978. set(gcf,'Renderer','Painters');
  1979. drawnow
  1981. % Save figure
  1982. if printfig
  1983. pos = get(f1,'Position');
  1984. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  1985. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  1986. print(f1,sprintf('%i-_inputPlacefieldExample.pdf',goodPost(pp)),'-dpdf','-r0')
  1987. end
  1988. end
  1990. clear post pairs fs pos cell_types pre printfig pos f1 pr po thisPre...
  1991. batchID linpath radpath ofpath minrate ratetypes fieldtype goodPost...
  1992. metPath minpre postLocID preLocIDs ratetype placefields isGoodPre
  1994. %% 4e Histogram converging pair numbers per cell
  1995. cell_types = {'PV','SST','Granule','Mossy','Pyramidal'}; % Do not use 3-5 as post!
  1996. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; 1 0 0; 0 0 1];
  1997. printfig = true;
  1999. hct = [];
  2000. f1 = figure('position',[100 100 600 250]);
  2001. for post = [1 2]
  2002. subplot(1,2,post); hold on
  2003. for pre = [4 3 5]
  2004. isGoodPost{post} = find(contains(class_metrics.putativeCellType, cell_types{post}));
  2005. %frct(end+1,:) = 100*histcounts(mono_res.nEinInd(isGoodPost,pre),0:22)./length(find(isGoodPost));
  2006. hct(end+1,:) = histcounts(mono_res.nEinInd(isGoodPost{post},pre),0:22);
  2008. histogram('BinEdges',(0:22), 'BinCounts',hct(end,:),...
  2009. 'FaceColor',colors(pre,:),'EdgeColor',colors(pre,:),'FaceAlpha',.5);
  2010. end
  2011. set(gca,'YScale','log');
  2012. ylabel('N pairs');
  2013. xlabel('Partners per cell');
  2014. set(gca,'FontSize',12);
  2016. if printfig
  2017. pos = get(f1,'Position');
  2018. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2019. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2020. print(f1,sprintf('NpresynPartnerHist.pdf'),'-dpdf','-r0')
  2021. end
  2022. end
  2024. for pre = [3 4 5]
  2025. [~,p] = unpairedSampleTest(mono_res.nEinInd(isGoodPost{1},pre),...
  2026. mono_res.nEinInd(isGoodPost{2},pre),'labels',{'PV','SST'},'verbose',true);
  2027. %fprintf('%s: PV vs. SST (N input per cell), p = %.4g\n',cell_types{pre},p)
  2028. end
  2030. clear hct isGoodPost cell_types pre post f1 pos p pre
  2032. %% 4f: Place-field correlations between converging presynaptic cells
  2033. % Run section 'Get converging...' above (preparations)
  2034. % first! It runs through all sessions, but non-manual mono res are excluded
  2035. % by 'include' criteria for both connected and unconnected pairs!
  2037. % 1=PV(svm), 2=SST(svm), 3=GC, 4=MC, 5=MEC-PYR, 6=PV(opto), 7=SST(opto)
  2038. labels = {'PV','SST','GC','MC','MEC PYR','PVo','SSTo'};
  2039. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; 1 0 0; 0 0 1];
  2040. printfig = false;
  2041. placefields = {linear,radial,openfield};
  2042. pflabels = {'Linear','Radial','Openfield'};
  2043. maxndots = 1000; % Downsample number of dots for distribution plots
  2045. % Use peak rate (i.e. in place field)
  2046. minrate = 1;
  2047. ratetype = 'peakrates'; % 'meanrates','peakrates'
  2048. pvals = NaN(3,5,7);
  2050. for pre = [3 4 5] % 3 = GC, 4 = MC
  2051. fprintf('\n%s:\n',labels{pre}) % For stats
  2053. for post = [1 2 6 7] % 1,2 PV/SST (svm), 6,7 PV/SST (opto)
  2054. for beh = [1 2 3] % Linear, Radial, Openfield
  2055. ConvCorr{beh,post} = get_PlacefieldCorrelations(convg_res.Econv{pre,post},...
  2056. placefields{beh},'minrate',minrate,'ratetype',ratetype);
  2057. NonConvCorr{beh,post} = get_PlacefieldCorrelations(convg_res.Enoconv{pre,post},...
  2058. placefields{beh},'minrate',minrate,'ratetype',ratetype);
  2060. % Print stats
  2061. fprintf('%s - %s, %s:\n',labels{pre},labels{post},pflabels{beh});
  2062. try
  2063. [~,pvals(beh,pre,post)] = unpairedSampleTest(...
  2064. ConvCorr{beh,post},NonConvCorr{beh,post},...
  2065. 'labels',{'convg','not convg'},'verbose',true);
  2066. end
  2067. end
  2068. end
  2070. % Two-half violin plot
  2071. % Downsample dots
  2072. for beh = 1:3
  2073. for post = 1:7
  2074. if length(NonConvCorr{beh,post})>maxndots
  2075. NonConvCorrPl{beh,post} = randsample(NonConvCorr{beh,post},maxndots);
  2076. else
  2077. NonConvCorrPl{beh,post} = NonConvCorr{beh,post};
  2078. end
  2080. if length(ConvCorr{beh,post})>maxndots
  2081. ConvCorrPl{beh,post} = randsample(ConvCorr{beh,post},maxndots);
  2082. else
  2083. ConvCorrPl{beh,post} = ConvCorr{beh,post};
  2084. end
  2085. end
  2086. end
  2088. % Controls (non-connected SVM pairs)
  2089. f1 = figure('position',[200 100 200 200]); hold on;
  2091. % Plot non-connected pair correlations (left half)
  2092. % NoConnCorr{behavior type, cell type (1=PV,2=SST,6=optoPV,7=optoSST)}
  2093. plotSpread(reshape(NonConvCorrPl([1 2 3],[1 2]),1,6),'spreadFcn',{'lin',8},...
  2094. 'spreadWidth',.9,'distributionColors',[.5 .5 .5],'distributionMarkerSize',1,...
  2095. 'spreadOri','left','distributionMarkerAlpha',.5,'distributionMarkers','o');
  2096. distributionPlot(reshape(NonConvCorr([1 2 3],[1 2]),1,6),...
  2097. 'histOri','left','color',[0 0 0],'widthDiv',[2 1],'showMM',6,...
  2098. 'FaceAlpha',.5,'divFactor',2);
  2100. % Connected opto Pairs (add as + on same plot)
  2101. plotSpread(reshape(ConvCorrPl([1 2 3],[6 7]),1,6),'spreadFcn',{'lin',8},... % [sic!], use wider (10) spread here 2/2 lower n datapoints
  2102. 'spreadWidth',.9,'distributionColors',[.5 .5 .5],'distributionMarkerSize',3,...
  2103. 'spreadOri','right','distributionMarkers','+');
  2105. % Connected SVM pairs
  2106. plotSpread(reshape(ConvCorrPl([1 2 3],[1 2]),1,6),'spreadFcn',{'lin',8},... % [sic!], use wider (10) spread here 2/2 lower n datapoints
  2107. 'spreadWidth',.9,'distributionColors',[.5 .5 .5],'distributionMarkerSize',1,...
  2108. 'spreadOri','right','distributionMarkerAlpha',.5,'distributionMarkers','o');
  2109. distributionPlot(reshape(ConvCorr([1 2 3],[1 2]),1,6),...
  2110. 'histOri','right','color',colors(pre,:),'widthDiv',[2 2],'showMM',6,...
  2111. 'FaceAlpha',.5,'divFactor',2);
  2113. ylabel('Placefield corr (R)');
  2114. ylim([-1 1]);
  2115. xticklabels({'PV Lin','PV Rad','PV OF','SST Lin','SST Rad','SST OF'});
  2116. xtickangle(45);
  2118. % Save figure
  2119. if printfig
  2120. pos = get(f1,'Position');
  2121. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2122. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2123. print(f1,sprintf('Fig3-%s-pairs-placefieldcorr.pdf',labels{pre}),'-dpdf','-r0')
  2124. end
  2125. end
  2127. % Dunn-Sidak Corrected p-values
  2128. %pvals = reshape(HolmBonferroniCorr(pvals(:)'),size(pvals));
  2129. pvals = multCompCorr(pvals,'method','Dunn-Sidak');
  2130. for pre = [3 4 5]
  2131. for post = [1 2 6 7]
  2132. for beh = [1 2 3]
  2133. fprintf('%s-%s: %s, p = %.4f\n',labels{pre},labels{post},...
  2134. pflabels{beh},pvals(beh,pre,post));
  2135. end
  2136. end
  2137. end
  2139. clear f1 minrate placefields ratetype pre post labels colors printfig pflabels...
  2140. ConvCorr NonConvCorr NonConvCorrPl ConvCorrPl pvals
  2142. %% S4j-l: Spatial coding properties of GC and MC recruiting PV or SST IN
  2143. % Cell selection etc.
  2144. fields = {'SIspk1D','SIspkRadial','SIspk2D'};%,'SIsecRadial','SpCohRadial','SesStabRadial'};
  2145. critfields = {'linearRateMean','radialRateMean','openfieldRateMean'};
  2146. transformation = 'log10';
  2147. printfig = true;
  2149. % Cell selection
  2150. labels = {'GC->PV','GC->SST','MC->PV','MC->SST'};
  2151. include = struct('putativeCellType',{'Granule Cell','Granule Cell','Mossy Cell','Mossy Cell'},...
  2152. 'groundTruthClassification',{'','','',''},'targets',{'PV','SST','PV','SST'});
  2153. exclude = struct('tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  2154. colors = [.83 .78 .46; .83 .78 .46; .9 0 .7; .9 0 .7];
  2156. % Create plot metrics, add .targetsSST and .targetsPV as metric tags
  2157. M = class_metrics;
  2158. econ = M.putativeConnections.excitatory;
  2159. for n = 1:length(M.UID)
  2160. targets = econ(econ(:,1)==n, 2);
  2161. M.targets{n} = {''};
  2162. if any(contains(M.putativeCellType(targets),'SST'))
  2163. M.targets{n} = cat(2, M.targets{n}, {'SST'});
  2164. end
  2166. if any(contains(M.putativeCellType(targets),'PV'))
  2167. M.targets{n} = cat(2, M.targets{n}, {'PV'});
  2168. %cat(2,metrics.groundTruthClassification{id},{thislabel{id}});
  2169. end
  2170. % M.targetsSST(n) = any(contains(M.putativeCellType(targets),'SST'));
  2171. % M.targetsPV(n) = any(contains(M.putativeCellType(targets),'PV'));
  2172. end
  2174. % PLOT
  2175. fi = figure('position',[125 450 105*length(fields) 125]);
  2176. for ff = 1:length(fields)
  2177. figure(fi);
  2178. subplot(1,length(fields),ff)
  2179. criteria = struct(critfields{ff},[.5 Inf]);
  2180. [~, valMtrx, ~] = plot_metricfield(M, fields{ff},...
  2181. 'include',include,'exclude',exclude, 'criteria',criteria,...
  2182. 'transformation',transformation,'tukeyFence',3,'labels',labels,...
  2183. 'colors',colors);
  2184. set(gca,'fontsize',6);
  2185. set(gca,'YLim',[-2 1]);
  2186. yticks([-2 -1 0 1]);
  2187. yticklabels([.01 .1 1 10]);
  2188. ylabel('SI (bits/spike)');
  2190. % ANOVA for SVM classified neurons
  2191. fprintf('\nDG statistics for %s, SVM types:',fields{ff});
  2192. kruskalwallisNdunns(valMtrx,labels);
  2193. end
  2195. % Save figure
  2196. if printfig
  2197. pos = get(fi,'Position');
  2198. set(fi,'PaperPositionMode','Auto','PaperUnits','points',...
  2199. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2200. print(fi,'SI_spk_PV_SST_targeting_cells.pdf','-dpdf','-r0')
  2201. end
  2203. clear M n targets econ valMtrx fi fields printfig include exclude criteria...
  2204. colors transformation labels ff critfields
  2207. %% Figure 5
  2209. %% 5c DS PSTH including LII PYR
  2210. labels = {'MECii PYR','DG pPV','DG pSST','DG GC','DG MC'};
  2211. colors = [1 0 0; .56 .39 .93; .28 1 .7; .9 .9 .47; .9 0 .7];
  2212. type = 'events'; % Manipulations, events
  2213. field = 'DS2'; % 'DS2','CA3ripples'
  2214. minRate = 0; minPrctile = 50; % Minimum percentile of bins during baseline above minRate
  2215. printfig = false;
  2217. % Create in- and exclude structure
  2218. % include = struct('putativeCellType',{'Pyramidal Cell','Narrow Interneuron',...
  2219. % 'Wide Interneuron','Granule Cell','Mossy Cell'},...
  2220. % 'brainRegion',{'ENTm2','DG','DG','DG','DG'});
  2221. include = struct('putativeCellType',{'Pyramidal Cell','PV interneuron',...
  2222. 'SST interneuron','Granule Cell','Mossy Cell'},...
  2223. 'brainRegion',{'ENTm2','DG','DG','DG','DG'});
  2224. exclude = struct('tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  2226. % DG plot
  2227. %sel = 1:length(include); % 1:4
  2228. f1 = figure('position',[200 20 180 180*length(include)]);
  2230. [psth, psthNorm, ccgts] = plot_psth(class_metrics,field,'type',type,'labels',labels,...
  2231. 'include',include,'exclude',exclude,...%'criteria',criteria(sel)
  2232. 'orientation','column','SDrange',[-3 3],...
  2233. 'colors',[0 0 0],'minRate',minRate,'minPrctile',minPrctile);
  2235. if printfig
  2236. pos = get(f1,'Position');
  2237. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2238. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2239. print(f1,sprintf('DS2_response.pdf'),'-dpdf','-r0')
  2240. end
  2242. f1 = figure('position',[400 20 200 200]);
  2243. for cc = 1:length(include) % Use order of peak delay instead?
  2244. shadedErrorBar(1e3*ccgts, nanmean(psthNorm{cc},1)-1.5*cc, nanstd(psthNorm{cc},[],1)./...
  2245. sqrt(size(psthNorm{cc},1)),'lineProps',{'-','color',colors(cc,:)});
  2246. % shadedErrorBar(ts, [0 diff(nanmean(psthNorm{cc},1))]-1*cc, nanstd(psthNorm{cc},[],1)./...
  2247. % sqrt(size(psthNorm{cc},1)),'lineProps',{'-','color',colors(cc,:)});
  2248. end
  2249. xlim([-50 50]);
  2250. xlabel('Time (ms)');
  2252. if printfig
  2253. pos = get(f1,'Position');
  2254. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2255. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2256. print(f1,sprintf('DS2_response_mean50ms.pdf'),'-dpdf','-r0')
  2257. end
  2259. clear include exclude f1 f2 printfig field pos sel type ds ts psth psthNorm...
  2260. ccgts minPrctile cc colors labels minRate
  2262. %% 5d Incl MEC: DS2_modulationPeakResponseTime
  2263. % Plot parameters
  2264. fields = {'DS2_modulationPeakResponseTime'}; % DS2_modulationSD
  2265. transformations = {''}; % -- set 'divFactor' in distribution plot to "1" for peakResponse time!
  2266. printfig = false;
  2268. % Cell selection
  2269. labels = {'MECii PC','PV','SST','GC','MC','PV+','SST+'};
  2270. include = struct('putativeCellType',{'Pyramidal Cell','PV interneuron','SST interneuron',...
  2271. 'Granule Cell','Mossy Cell','',''},...
  2272. 'groundTruthClassification',{'','','','','','PV+','SST+'},...
  2273. 'brainRegion',{'ENTm2','DG','DG','DG','DG','DG','DG'});
  2274. exclude = struct('tags',{{'Bad','InverseSpike'}});
  2275. criteria = struct('DS2_modulationRatio',[1.5 Inf]);
  2276. colors = [1 0 0; .56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  2279. for ff = 1:length(fields)
  2280. % PLOTTING
  2281. fi{ff} = figure('position',[200*ff 450 150 130]); % Same aspect ratio
  2283. [~, valMtrx, ~] = plot_metricfield(class_metrics,fields{ff},...
  2284. 'include',include,'exclude',exclude,'criteria',criteria,...
  2285. 'transformation',transformations{ff},'tukeyFence',3,'labels',labels,...
  2286. 'colors',colors,'splitgroups',[6 2; 7 3],'plottype','violin',...
  2287. 'divFactor',1);
  2288. set(gca,'fontsize',6);
  2289. xlim([.25 5.75]);
  2291. % Save figure
  2292. if printfig
  2293. pos = get(fi{ff},'Position');
  2294. set(fi{ff},'PaperPositionMode','Auto','PaperUnits','points',...
  2295. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2296. print(fi{ff},sprintf('Fig5d_%s_svm_opto.pdf',fields{ff}),'-dpdf','-r0')
  2297. end
  2299. % ANOVA for SVM classified neurons
  2300. fprintf('\nDG statistics for %s, SVM types:',fields{ff});
  2301. kruskalwallisNdunns(valMtrx(:,1:5), labels(1:5));
  2303. % ANOVA for optotagged cells only
  2304. if ~all(isnan(valMtrx(:,5:6)),'all')
  2305. fprintf('\nDG statistics for %s, opto vs GC/MC:',fields{ff});
  2306. kruskalwallisNdunns(valMtrx(:,[1 6 7 4 5]), labels([1 6 7 4 5]));
  2307. end
  2309. % LMM nested data stats (SVM & Opto separately)
  2310. valTbl = get_metricfieldTbl(class_metrics, fields{ff}, 'include',include,...
  2311. 'exclude',exclude, 'criteria',criteria, 'labels',labels);
  2312. %glme_stats(valTbl)
  2314. fprintf('\nDG LMM statistics for %s, SVM types:',fields{ff});
  2315. isSVM = cellfun(@(x) (any(x=={'PV','SST','MECii PC','GC','MC'})), table2cell(valTbl(:,2)));
  2316. thisTbl = valTbl(isSVM,:);
  2317. thisTbl.Group = removecats(thisTbl.Group); % Remove unused categories
  2318. glme_stats(thisTbl);
  2320. fprintf('\nDG LMM statistics for %s, opto vs GC/MC:',fields{ff});
  2321. isOpto = cellfun(@(x) (any(x=={'PV+','SST+','MECii PC','GC','MC'})), table2cell(valTbl(:,2)));
  2322. thisTbl = valTbl(isOpto,:);
  2323. thisTbl.Group = removecats(thisTbl.Group); % Remove unused categories
  2324. glme_stats(thisTbl);
  2325. end
  2327. % Plot numbers to console
  2328. fprintf('\nLatency to Dentate Spike (ms)\n');
  2329. for ct = 1:length(include)
  2330. % fprintf('%s firing rate: %.2f +/- %.2f (mean +/- STD)\n',labels{ct},...
  2331. % nanmean(valMtrx(:,ct)), nanstd(valMtrx(:,ct)))
  2332. fprintf('%s Latency: Median %.2f (%.2f - %.2f IQR)\n',labels{ct},...
  2333. 1e3*nanmedian(valMtrx(:,ct)), 1e3*prctile(valMtrx(:,ct),25),...
  2334. 1e3*prctile(valMtrx(:,ct),75))
  2335. end
  2337. clear include exclude criteria valGrp valMtrx valCell fields transformation...
  2338. sel fi printfig ff colors labels pos transformations valTbl thisTbl isSVM isOpto
  2340. %% 5e Incl MEC: DS2_modulationRatio
  2341. % Plot parameters
  2342. fields = {'DS2_modulationRatio'}; % DS2_modulationSD
  2343. transformations = {'log2'}; % -- set 'divFactor' in distribution plot to "1" for peakResponse time!
  2344. printfig = true;
  2346. % Cell selection
  2347. labels = {'MECii PC','PV','SST','GC','MC','PV+','SST+'};
  2348. include = struct('putativeCellType',{'Pyramidal Cell','PV interneuron','SST interneuron',...
  2349. 'Granule Cell','Mossy Cell','',''},...
  2350. 'groundTruthClassification',{'','','','','','PV+','SST+'},...
  2351. 'brainRegion',{'ENTm2','DG','DG','DG','DG','DG','DG'});
  2352. exclude = struct('tags',{{'Bad','InverseSpike'}});
  2353. criteria = struct('UID',[0 Inf]);
  2354. %[criteria.radialRatePeak] = deal([1 Inf]); % Specific exclusion criteria here
  2355. colors = [1 0 0; .56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  2358. for ff = 1:length(fields)
  2359. % PLOTTING
  2360. fi{ff} = figure('position',[200*ff 450 150 130]); % Same aspect ratio
  2362. [~, valMtrx, ~] = plot_metricfield(class_metrics,fields{ff},...
  2363. 'include',include,'exclude',exclude,'criteria',criteria,...
  2364. 'transformation',transformations{ff},'tukeyFence',3,'labels',labels,...
  2365. 'colors',colors,'splitgroups',[6 2; 7 3],'plottype','violin',...
  2366. 'divFactor',1);
  2367. set(gca,'fontsize',6);
  2368. xlim([.25 5.75]);
  2370. % Save figure
  2371. if printfig
  2372. pos = get(fi{ff},'Position');
  2373. set(fi{ff},'PaperPositionMode','Auto','PaperUnits','points',...
  2374. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2375. print(fi{ff},sprintf('Fig5e_%s_svm_opto.pdf',fields{ff}),'-dpdf','-r0')
  2376. end
  2378. % ANOVA for SVM classified neurons
  2379. fprintf('\nDG statistics for %s, SVM types:',fields{ff});
  2380. kruskalwallisNdunns(valMtrx(:,1:5), labels(1:5));
  2382. % ANOVA for optotagged cells only
  2383. if ~all(isnan(valMtrx(:,5:6)),'all')
  2384. fprintf('\nDG statistics for %s, opto vs GC/MC:',fields{ff});
  2385. kruskalwallisNdunns(valMtrx(:,[1 6 7 4 5]), labels([1 6 7 4 5]));
  2386. end
  2388. % LMM nested data stats (SVM & Opto separately)
  2389. valTbl = get_metricfieldTbl(class_metrics, fields{ff}, 'include',include,...
  2390. 'exclude',exclude, 'criteria',criteria, 'labels',labels,...
  2391. 'transformation',transformations{ff});
  2392. %glme_stats(valTbl)
  2394. fprintf('\nDG LMM statistics for %s, SVM types:',fields{ff});
  2395. isSVM = cellfun(@(x) (any(x=={'PV','SST','MECii PC','GC','MC'})), table2cell(valTbl(:,2)));
  2396. thisTbl = valTbl(isSVM,:);
  2397. thisTbl.Group = removecats(thisTbl.Group); % Remove unused categories
  2398. glme_stats(thisTbl);
  2400. fprintf('\nDG LMM statistics for %s, opto vs GC/MC:',fields{ff});
  2401. isOpto = cellfun(@(x) (any(x=={'PV+','SST+','MECii PC','GC','MC'})), table2cell(valTbl(:,2)));
  2402. thisTbl = valTbl(isOpto,:);
  2403. thisTbl.Group = removecats(thisTbl.Group); % Remove unused categories
  2404. glme_stats(thisTbl);
  2405. end
  2407. clear include exclude criteria valGrp valMtrx valCell fields transformation...
  2408. sel fi printfig ff colors labels pos transformations
  2410. %% 5c-f DS rate responses to opto manipulations
  2411. % Control in / out of stim DS
  2412. % ts = randsample(DS2.timestamps(:,1),2000);
  2413. % ts2 = DS2.timestamps(InIntervals(DS2.timestamps(:,1),opto_1sec.timestamps),1);
  2414. % ts3 = DS2.peaks(InIntervals(DS2.peaks(:,1),opto_1sec.timestamps));
  2415. % bz_multishankEventCSD(ts,'window',[-.1 .1]);
  2417. ChRmice = {'SST2','SST4','YutaMouse20','YutaMouse21','YutaMouse37','YutaMouse38',...
  2418. 'YutaMouse39','YutaMouse40','PV1','PV3','PV5','PV16','PV18','PV19','PV21','PV29'};
  2419. HRmice = {'PV13','PV22','SST3','SST5'};
  2421. n = 0;
  2422. minPrctile = 10; % Minimum fraction of bins that is >0
  2423. evname = 'DS2'; % 'DS2','CA3ripples','ripples'
  2424. opsins = {'ChR2','eNpHR'};
  2425. lines = {'PV','SST'};
  2427. % GET DATA
  2428. clear results
  2429. for ds = 1:length(handle)
  2430. basename = handle(ds).basename;
  2431. basepath = handle(ds).basepath;
  2433. if isfile(fullfile(basepath,filesep,[basename '.' evname '.events.mat'])) &&...
  2434. isfile(fullfile(basepath,filesep,[basename '.opto_1sec.manipulation.mat']))
  2436. evt = load(fullfile(basepath,filesep,[basename '.' evname '.events.mat']),evname);
  2437. evt = evt.(evname);
  2438. load(fullfile(basepath,filesep,[basename '.opto_1sec.manipulation.mat']))
  2440. [thisCCG, PSTHts] = CCG({opto_1sec.timestamps(:,1),evt.peaks},[],...
  2441. 'binSize',.05,'duration',4,'norm','rate');
  2443. if ~prctile(thisCCG(:,1,2),100-minPrctile) > 0
  2444. % Exclude session if little or no DS detected.
  2445. continue
  2446. end
  2448. n = n+1;
  2449. results(n).name = basename;
  2450. results(n).PSTH = thisCCG(:,1,2);
  2451. results(n).ts = PSTHts;
  2453. results(n).stimrate = mean(thisCCG(InIntervals(PSTHts,[0 1]),1,2));
  2454. results(n).baserate = mean(thisCCG(~InIntervals(PSTHts,[0 1]),1,2));
  2456. results(n).pctchg = results(n).stimrate / results(n).baserate;
  2458. % Get genetic Line
  2459. if contains(handle(ds).basename,'PV')
  2460. results(n).geneticLine = 'PV-Cre';
  2461. else
  2462. results(n).geneticLine = 'SST-Cre';
  2463. end
  2465. % Get opsin
  2466. if any(strcmp(handle(ds).animal, ChRmice))
  2467. results(n).opsin = 'ChR2';
  2468. elseif any(strcmp(handle(ds).animal, HRmice))
  2469. results(n).opsin = 'eNpHR';
  2470. end
  2471. else
  2472. fprintf('%s does not have all files, continue...\n',basename);
  2473. end
  2474. end
  2476. % PLOT FIGURE
  2477. figure('position',[200 50 270 800]);
  2478. for op = 1:2
  2479. for li = 1:2
  2480. isSes = contains({results.geneticLine},lines{li}) &...
  2481. contains({results.opsin},opsins{op});
  2482. allPSTH = cat(2,results(isSes).PSTH);
  2484. plID = 1 + 3*(li-1) + 6*(op-1);
  2485. subplot(4, 3, plID:plID+1)
  2486. hold on;
  2488. % Baseline normalization / zscoring
  2489. isBsl = InIntervals(results(1).ts,[-1 0]); % Assume uniform timestamps
  2490. allPSTHz = (allPSTH - mean(allPSTH(isBsl,:),1)) ./ std(allPSTH(isBsl,:),[],1);
  2492. % PSTH (BSL-ZSCORED) & TIMECOURSE (HZ)
  2493. imagesc(results(1).ts, linspace(0,1,size(allPSTHz,1)), allPSTHz'); % Keeping y-range simple 0-1 Hz is a good range
  2494. colormap(bwyr);%'hot');
  2495. caxis([-4 4]);
  2496. set(gca,'YDir','normal')
  2498. % Overlay mean (Hz)
  2499. %plot(results(1).ts, mean(allPSTH,2), 'k', 'lineWidth',1);
  2500. shadedErrorBar(results(1).ts, mean(allPSTH,2), std(allPSTH,[],2)./sqrt(length(find(isSes))));
  2502. % Label with N;
  2503. unitstr = sprintf('N = %i',length(find(isSes)));
  2504. text(prctile(results(1).ts,70), .1, unitstr,'Color','k') % gn-gn/15
  2506. title(sprintf('%s-%2',lines{li},opsins{op}));
  2507. xlabel('Time from Stim (s)');
  2508. xlim([-1 2]);
  2509. ylabel(sprintf('%s rate (Hz)',evname));
  2511. % BOXPLOTS
  2512. subplot(4,3,plID+2);
  2513. hold on;
  2514. plot(cat(1,[results(isSes).baserate],[results(isSes).stimrate]),'color',[0 0 0 .5],'LineWidth',1);
  2515. boxplot(cat(1,[results(isSes).baserate],[results(isSes).stimrate])','symbol', '');
  2516. xticklabels({'Pre','Stim'});
  2517. ylim([0 1]);
  2519. % STATISTICS
  2520. fprintf('%s - %s, opto 1 sec DS rate\n',lines{li},opsins{op});
  2521. pairedSampleTest([results(isSes).baserate],[results(isSes).stimrate],...
  2522. 'labels',{'Baseline','Stim'},'verbose',true);
  2523. changes{op,li} = 100*[results(isSes).pctchg];
  2524. end
  2525. end
  2527. % ChR reduction numbers & comparison (text)
  2528. fprintf('\nPV-ChR DS rate stim/bsl: %.2f +/- %.2f; SST-ChR DS rate stim/bsl: %.2f +/- %.2f\n',...
  2529. nanmean(changes{1,1}), nanstd(changes{1,1}), nanmean(changes{1,2}), nanstd(changes{1,2}));
  2530. unpairedSampleTest(changes{1,1},changes{1,2},'verbose',true);
  2532. clear PSTHts n thisCCG Epath sess DS2 pulses mouse a animals Ebase sf subf...
  2533. allPSTH ChRmice HRmice li op lines opsins allPSTHz basename basepath...
  2534. ds evname evt isBsl isSes minPrctile opto_1sec plID unitstr changes
  2536. %% Supplementary 5a(1) theta CSD
  2537. % DS and Theta CSD, see LFP_analysis/LFP_scripts and Analysis/LFP folder in
  2538. % each dataset for code and figures, respectively.
  2540. %% Supplementary 5a(2) Theta-gamma CFC
  2541. % see LFP_analysis/LFP_scripts
  2542. % Load session.mat and Analysis/LFP/.THetaGammaCFC.mat e.g. from
  2543. % TH15_220706
  2544. spkg = [5 11 12];
  2546. figure('position',[100 100 1400 300]); hold on
  2547. for sh = 1:length(spkg)
  2548. cfcg = CFC.comod(:,:,session.extracellular.spikeGroups.channels{spkg(sh)});
  2549. cfcg = reshape(cfcg,size(cfcg,1),[]);
  2550. subplot(1,length(spkg),sh);
  2551. contourf(CFC.amp_bincenters, 1:size(cfcg,2), flipud(cfcg'), 40, 'LineColor','none');
  2552. colormap(jet);
  2553. clim([0 1e-3]);
  2554. end
  2555. clear shanks sh
  2557. %% Supplementary 5b: DS2 modulation Lii vs. Liii MEC
  2558. type = 'events'; % Manipulations, events
  2559. field = 'DS2'; % 'DS2','CA3ripples'
  2560. metrics = class_metrics; % batch_metrics;
  2561. minRate = 0; minPrctile = 50; % Minimum percentile of bins during baseline above minRate
  2562. printfig = false;
  2564. % Create in- and exclude structure
  2565. include = struct('brainRegion',{'ENTm2','ENTm2','ENTm3','ENTm3'},...
  2566. 'putativeCellType',{'Pyramidal','Interneuron','Pyramidal','Interneuron'});
  2567. labels = {'Lii PYR','Lii IN','Liii PYR','Liii IN'};
  2568. colors = [1 0 0; 0 0 1; 1 1 0; 0 1 1];
  2569. sel = 1:length(include); % 1:4
  2570. % include = struct('putativeCellType',cell_types,'groundTruthClassification',incl_opto);
  2571. % exclude = struct('groundTruthClassification',excl_opto);
  2573. % PSTH
  2574. f1 = figure('position',[200 20 150 180*length(sel)]);
  2575. plot_psth(metrics,field,'type',type,'labels',labels(sel),...
  2576. 'include',include(sel),...%'exclude',exclude(sel),'criteria',criteria(sel),...
  2577. 'orientation','column','SDrange',[-3 3],'overlay',true,...
  2578. 'colors',[0 0 0],'minRate',minRate,'minPrctile',minPrctile); %colors(sel,:)
  2580. if printfig
  2581. pos = get(f1,'Position');
  2582. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2583. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2584. print(f1,sprintf('DS2_response_MEC.pdf'),'-dpdf','-r0')
  2585. end
  2587. % PIE CHARTS
  2588. f2 = figure('position',[400 20 200 180*length(sel)]);
  2589. plot_response_piechart(metrics,'field','DS2','labels',labels(sel),'alpha',.001,...
  2590. 'include',include(sel));%,'exclude',exclude(sel),'criteria',criteria(sel));
  2592. if printfig
  2593. pos = get(f2,'Position');
  2594. set(f2,'PaperPositionMode','Auto','PaperUnits','points',...
  2595. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2596. print(f2,sprintf('DS2_response_MECpie.pdf'),'-dpdf','-r0')
  2597. end
  2598. clear include exclude f1 f2 printfig field pos sel type ds ts psth psthNorm metrics
  2600. %% Supplementary 5c: MEC brain-state-dependent firing rates
  2601. printfig = false;
  2603. % Create in- and exclude structure
  2604. labels = {'MECii PC','MECii IN','MECiii PC','MECiii IN'};
  2605. %limrate = [8 40 8 40]; % Hard coding here, no easy fix...
  2606. %limrate = [-.5 1.2; -.5 2; -.5 1.2; -.5 2]; Log10
  2607. limrate = [0 8; 0 40; 0 8; 0 40]; % Linear
  2608. include = struct('brainRegion',{'ENTm2','ENTm2','ENTm3','ENTm3'},...
  2609. 'putativeCellType',{'Pyramidal','Interneuron','Pyramidal','Interneuron'});
  2610. exclude = struct('putativeCellType',repmat({'Unknown'},1,4),...
  2611. 'tags',repmat({{'InverseSpike','Bad'}},1,4));
  2612. colors = [1 0 0; 0 0 1; 1 0 0; 0 0 1];
  2614. f1 = figure('position',[100 100 1250 200]);
  2616. for ct = 1:4
  2617. [~,cellIDs] = select_cells(class_metrics,'include',include(ct),'exclude',exclude(ct));
  2619. rates = cat(1, class_metrics.firingRateWake(cellIDs),...
  2620. class_metrics.firingRateNREM(cellIDs), class_metrics.firingRateREM(cellIDs));
  2621. % rates = log10(rates);
  2622. rates = rates(:,~any(isnan(rates),1));
  2623. figure(f1)
  2624. ax(ct) = subplot(1,4,ct); hold(ax(ct),'on');
  2626. % Make rates heatmap in background
  2627. [~,order] = sort(rates(1,:));
  2628. imagesc(1:3, linspace(limrate(ct,1), limrate(ct,2), size(rates,2)), rates(:,order)');
  2629. colormap(bwyr);
  2630. colorbar
  2631. %clim([0 prctile(rates(:),98)]);
  2632. clim([limrate(ct,1) limrate(ct,2)]);
  2633. % % Plot lines
  2634. % plot(rates,'color',[0 0 0 .1],'LineWidth',.3);
  2636. % Plot boxes
  2637. boxplot(rates','symbol', '');
  2638. h = findobj('LineStyle','--'); set(h, 'LineStyle','-');
  2639. h = findobj(gca,'Tag','Box');
  2640. for j=1:length(h)
  2641. patch(get(h(j),'XData'),get(h(j),'YData'),colors(ct,:),'FaceAlpha',.7);
  2642. end
  2644. % Markup
  2645. title(labels{ct});
  2646. ylabel('Firing Rate (Hz)');
  2647. % yticklabels('');
  2648. xticklabels({'Wake','NREM','REM'});
  2649. xtickangle(45);
  2650. ylim([limrate(ct,1) limrate(ct,2)]);
  2651. % else
  2652. % % subplot(1,4,ct-4);
  2653. % % Just add lines for opto-positive cells
  2654. % plot(ax(ct-4),rates,'color',[colors(ct,:) .25],'LineWidth',.5);
  2655. % end
  2657. % Stats (repeated measures ANOVA on ranks)
  2658. fprintf([labels{ct} '\n'])
  2659. kruskalwallisNdunns(rates',{'Wake','NREM','REM'});
  2661. % TODO: Log transform and repeated measures
  2663. %rates = rates(:,~any(rates == 0,1)); % Remove zeros to avoid prob w/ log transform
  2664. %rates = log10(rates);
  2666. %rates = array2table(rates','VariableNames',{'Wake','NREM','REM'});
  2667. %anovan(rates);
  2668. end
  2670. if printfig
  2671. pos = get(f1,'Position');
  2672. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2673. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2674. print(f1,sprintf('StateRates_psth.pdf'),'-dpdf','-r0')
  2675. end
  2677. clear ct h field j rates f1 wi order pos printfig ax cellIDs colors exclude...
  2678. include labels limrate
  2681. %% Figure 6
  2683. %% 6a,c PSTH 1 sec pulses - row-wise, all, include direct-activated cells
  2684. % TODO: Run class_metrics = reclassifyOptoResponses2(class_metrics); prior
  2685. % to this! -- make sure addAmbiguousResponse was run
  2686. type = 'manipulations'; % Manipulations, events
  2687. field = 'opto_1sec'; %'opto_chirp_freq','opto_1sec'
  2688. sortCriterion = '_modulationSD'; % '_modulationSD' (default) or '_peakFreq' for chirp
  2689. baseline = [-Inf -.005]; % [-Inf 0] standard, [4 40] for freq
  2690. lines = {'PV-Cre','SST-Cre'};
  2691. opsins = {'eNpHR','ChR2'};%'eNpHR' 'ChR2'
  2692. printfig = true;
  2693. minRate = 0; minPrctile = 1; % Minimum percentile of bins during baseline above minRate
  2695. % Create in- and exclude structure
  2696. %labels = {'PV','SST','GC','MC'};
  2697. include = struct('putativeCellType',{'','PV interneuron','SST interneuron',...
  2698. 'Granule Cell','Mossy Cell'}, 'brainRegion','DG',...
  2699. 'groundTruthClassification',cat(2,{'+'},repmat({''},1,4)));
  2700. exclude = struct('groundTruthClassification',cat(2,{''},repmat({{'+','~'}},1,4)),... % {{'+'}},{{'+','~'}}
  2701. 'tags',repmat({{'Bad','InverseSpike'}},1,5),'putativeCellType',repmat({'Unknown'},1,5));
  2703. % Plot
  2704. f1 = figure('position',[200 200 600 550]);
  2706. for op = 1:2
  2707. [include.labels] = deal(opsins{op});
  2708. for li = 1:2
  2709. [include.geneticLine] = deal(lines{li});
  2710. for ct = 1:5
  2711. subplot(4,5,ct + 5*(li-1) + 10*(op-1));
  2712. plot_psth(class_metrics,field,'type',type,'labels',{''},...%labels(ct),...
  2713. 'include',include(ct),'exclude',exclude(ct),...
  2714. 'colors',[0 0 0],'minRate',minRate,'minPrctile',minPrctile,... %'colors',colors(sel,:)
  2715. 'sortCriterion',sortCriterion,'baseline',baseline,'shadedErrorBar',false);
  2716. if ~(op == 2 && li == 2); xlabel(''); end
  2717. if ~(ct == 1); ylabel(''); end
  2718. end
  2719. end
  2720. end
  2722. % Save figure
  2723. if printfig
  2724. pos = get(f1,'Position');
  2725. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2726. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2727. print(f1, '1secPSTH_all.pdf','-dpdf','-r0')
  2728. end
  2730. clear include exclude f1 f2 printfig field pos sel type ds ts psth psthNorm ct li op...
  2731. lines opsins include exclude minRate minPrctile sortCriterion
  2733. %% 6b,d Firing rate ratio during / pre 1 sec pulses (PV vs. SST)
  2734. fields = {'opto_1sec_modulationRatio'}; % 4c,d 'DS2_modulationSD','opto_1sec_modulationSD','opto_1sec_modulationIndex'
  2735. transformations = {'log2'}; % 4c,d -- set 'divFactor' in distribution plot to "1" for peakResponse time!
  2736. opsin = 'eNpHR'; % 'eNpHR' for b), 'ChR' for d)
  2737. printfig = false;
  2739. % Create in- and exclude structure
  2740. labels = {'PV-Cre','SST-Cre'};
  2741. statlabels = cat(1,strcat({'PV','SST','GC','MC'},'-',labels{1}),...
  2742. strcat({'PV','SST','GC','MC'},'-',labels{2}));
  2743. include = struct('putativeCellType',{'PV','PV','SST','SST',...
  2744. 'Granule','Granule','Mossy','Mossy'});
  2745. [include(1:2:end).geneticLine] = deal('PV-Cre');
  2746. [include(2:2:end).geneticLine] = deal('SST-Cre');
  2747. [include.labels] = deal(opsin);
  2748. exclude = struct('groundTruthClassification',{{'+','~'}},...
  2749. 'tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  2750. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  2752. for ff = 1:length(fields)
  2753. % Plotting
  2754. f1 = figure('position',[200*ff 450 200 230]); % Same aspect ratio
  2755. hold on
  2757. [~, valMtrx, ~] = plot_metricfield(class_metrics,fields{ff},...
  2758. 'include',include,'exclude',exclude,...
  2759. 'transformation',transformations{ff},'tukeyFence',3,'labels',repmat(labels,1,4),...
  2760. 'colors',colors([1 1 2 2 3 3 4 4],:),'plottype','violin',...
  2761. 'divFactor',1);
  2763. % Zero-line
  2764. plot([0 length(include)+1],[0 0],':','color',[.5 .5 .5]);
  2766. % Statistics
  2767. fprintf('\nDG statistics for %s, SVM types:',fields{ff});
  2768. kruskalwallisNdunns(valMtrx,strcat(repmat(...
  2769. {'PV-Cre:','SST-Cre:'},1,4),{include.putativeCellType}));
  2771. % LMM nested data stats (SVM & Opto separately)
  2772. valTbl = get_metricfieldTbl(class_metrics, fields{ff}, 'include',include,...
  2773. 'exclude',exclude, 'labels',statlabels(:), 'transformation',transformations{ff});
  2774. %glme_stats(valTbl)
  2776. fprintf('\nDG LMM statistics for %s:',fields{ff});
  2777. glme_stats(valTbl);
  2778. end
  2780. if printfig
  2781. pos = get(f1,'Position');
  2782. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2783. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2784. print(f1,sprintf('Fig6b+d_%s_1sec_ratioViolin.pdf',opsin),'-dpdf','-r0')
  2785. end
  2787. clear include exclude thiscriteria values fields transformations valCell...
  2788. valGrp valMtrx ff opsin colors labels f1 pos
  2790. %% 6e,f + S8a-d: 1 sec single-trial Lorenz Curves, gini-diff & CV
  2791. % Create in- and exclude structure
  2792. printfig = false;
  2793. labels = {'PV-HR','SST-HR','PV-ChR','SST-ChR'};
  2794. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  2795. 'Granule Cell','Mossy Cell'});
  2796. %exclude = struct('groundTruthClassification',{{'+'},{'+'},{'+'},{'+'}});
  2797. exclude = struct('groundTruthClassification',{{'+','~'}},...
  2798. 'tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  2799. lines = {'PV-Cre','SST-Cre'};
  2800. opsins = {'eNpHR','ChR2'};
  2801. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  2802. fields = {'gini','CV'}; % For errorbar plots (diff stim vs bsl)
  2804. % Get Lorenz curves, gini, CV
  2805. if ~exist('LorRes','var') % Takes long to load
  2806. for li = [1 2] % PV-Cre / SST-Cre
  2807. [include.geneticLine] = deal(lines{li});
  2809. for op = [1 2] % eNpHR / ChR
  2810. [include.labels] = deal(opsins{op});
  2812. fprintf('%s, %s\n',lines{li},opsins{op});
  2813. LorRes{li,op} = get_1sec_Lorenz_Curves(class_metrics,...
  2814. 'include',include,'exclude',exclude,'show_results',false,...
  2815. 'minNspikes',1, 'minNcells',4);
  2816. end
  2817. end
  2818. end
  2820. % LORENZ CURVES
  2821. f1= figure('position',[100 100 500 450]); hold on;
  2822. for li = 1:2
  2823. for op = 1:2
  2824. for ct = 1:length(include)
  2825. subplot(4,length(include),ct + (op-1)*length(include) + (li-1)*2*length(include))
  2827. % Plot eNpHR results
  2828. plot([0 100],[0 1],':','color',[.5 .5 .5]);
  2829. nSess = length(find(LorRes{li,op}.nCells(:,ct)>0));
  2830. shadedErrorBar(1:100, nanmean(LorRes{li,op}.MeanLorenzCurveNorm(:,:,ct,1),1),...
  2831. nanstd(LorRes{li,1}.MeanLorenzCurveNorm(:,:,ct,1),1)./sqrt(nSess),...
  2832. 'lineprops',{'color',colors(ct,:)});
  2833. shadedErrorBar(1:100, nanmean(LorRes{li,op}.MeanLorenzCurveNorm(:,:,ct,2),1),...
  2834. nanstd(LorRes{li,1}.MeanLorenzCurveNorm(:,:,ct,2),1)./sqrt(nSess),...
  2835. 'lineprops',{'color',[.5 .5 .5]});
  2836. text(10,.8,sprintf('%s\n%s',lines{li},opsins{op}),'FontSize',8);
  2837. drawnow
  2838. end
  2839. end
  2840. end
  2842. % Save figure
  2843. if printfig
  2844. pos = get(f1,'Position');
  2845. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2846. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2847. print(f1,'Fig_6e_S7a_LorenzCurves.pdf','-dpdf','-r0')
  2848. end
  2850. % PLOT GINI, CV, ETC
  2851. for ff = 1:2
  2852. % GET VALUES
  2853. stvals = cat(1, LorRes{1,1}.(fields{ff})(:,:,1), LorRes{2,1}.(fields{ff})(:,:,1),...
  2854. LorRes{1,2}.(fields{ff})(:,:,1), LorRes{2,2}.(fields{ff})(:,:,1)); % To sort by condition, cat(2,... and omit reshape
  2855. blvals = cat(1, LorRes{1,1}.(fields{ff})(:,:,2), LorRes{2,1}.(fields{ff})(:,:,2),...
  2856. LorRes{1,2}.(fields{ff})(:,:,2), LorRes{2,2}.(fields{ff})(:,:,2));
  2857. stvals = reshape(stvals, size(LorRes{1,1}.gini,1), length(include)*length(lines)*length(opsins));
  2858. blvals = reshape(blvals, size(LorRes{1,1}.gini,1), length(include)*length(lines)*length(opsins));
  2860. % RAW VALUES (DUAL VIOLIN PLOTS)
  2861. figure('position',[100 100 500 150]); hold on;
  2863. % baseline (left, grey)
  2864. plotSpread(blvals,'spreadFcn',{'lin',10},...
  2865. 'spreadWidth',.9,'distributionColors',[.5 .5 .5],'distributionMarkerSize',3,...
  2866. 'spreadOri','left','distributionMarkerAlpha',.5,'distributionMarkers','o');
  2867. distributionPlot(blvals,'histOri','left','color',[0 0 0],...
  2868. 'widthDiv',[2 1],'showMM',6,'FaceAlpha',.5);
  2870. % Stim (right)
  2871. plotSpread(stvals,'spreadFcn',{'lin',10},... % [sic!], use wider (10) spread here 2/2 lower n datapoints
  2872. 'spreadWidth',.9,'distributionColors',[.5 .5 .5],'distributionMarkerSize',3,...
  2873. 'spreadOri','right','distributionMarkerAlpha',.5,'distributionMarkers','o');
  2874. distributionPlot(stvals,'histOri','right','color',num2cell(colors([1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4],:),2),...
  2875. 'widthDiv',[2 2],'showMM',6,'FaceAlpha',.5);
  2876. %ylim([.4 1.1]);
  2877. ylabel(fields{ff});
  2880. % STIM-BSL DELTA PLOT
  2881. f1 = figure('position',[100 100 400 250]); hold on;
  2882. % distributionPlot(stvals-blvals,'showMM',6,'FaceAlpha',.5,'histOpt',1,'divFactor',1,...
  2883. % 'color',num2cell(colors([1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4],:),2));
  2884. plot([0 size(stvals,2)+1],[0 0],'--','Color',[.5 .5 .5]); % Zero line
  2885. plotSpread(stvals-blvals,'spreadFcn',{'lin',5},...
  2886. 'spreadWidth',.9,'distributionColors',[0 0 0],'distributionMarkerSize',4,...
  2887. 'distributionMarkerAlpha',1,'distributionMarkers','o');
  2888. customErrorBar(stvals-blvals, 'errtype','STD', 'alpha',.7,...
  2889. 'markerSize',100, 'LineWidth',1.2,...
  2890. 'colors',colors([1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4],:));
  2891. xlim([0 size(stvals,2)+1]);
  2892. %ylim([-.3 .4]);
  2893. ylabel(sprintf('%s (stim-bsl)',fields{ff}));
  2894. xticklabels(repmat(labels,1,4));
  2895. xtickangle(45);
  2897. % Save figure
  2898. if printfig
  2899. pos = get(f1,'Position');
  2900. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  2901. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  2902. print(f1,sprintf('Opto_1sec_%s_change.pdf',fields{ff}),'-dpdf','-r0')
  2903. end
  2905. % STATISTICS
  2906. for ct = 1:length(include)
  2907. clear blvals stvals groupVals
  2909. % Statistics @console
  2910. fprintf('\n%s:',include(ct).putativeCellType)
  2911. fprintf('\n%s (raw p):\n',fields{ff}) % For statistics in console
  2913. for li = 1:2
  2914. for op = 1:2
  2915. blvals(:,li,op) = LorRes{li,op}.(fields{ff})(:,ct,2); % nSess x nGroup(ct) x [1 (stim) 2 (bsl)]
  2916. stvals(:,li,op) = LorRes{li,op}.(fields{ff})(:,ct,1);
  2918. try
  2919. pvals(ct,li,op) = pairedSampleTest(blvals(:,li,op), stvals(:,li,op),...
  2920. 'labels',{'baseline','stim'},'verbose',true);
  2921. catch
  2922. pvals(ct,li,op) = NaN;
  2923. fprintf('\n');
  2924. end
  2925. end
  2926. end
  2928. % ANOVA
  2929. groupVals = reshape(stvals-blvals, size(stvals,1), size(stvals,2)*size(stvals,3));
  2930. fprintf('%s ANOVA:\n',fields{ff})
  2931. kruskalwallisNdunns(groupVals, labels);
  2932. end
  2934. % Multi-Comparison Corrected p-values
  2935. padj = multCompCorr(pvals,'method','Dunn-Sidak'); % 'Holm-Bonferroni', 'Dunn-Sidak'
  2936. fprintf('%s (adjusted p-values):\n',fields{ff});
  2937. for op = [1 2]
  2938. for li = [1 2]
  2939. for ct = 1:length(include)
  2940. fprintf('%s-%s: %s (stim vs. bsl), p = %.4g\n',lines{li}, opsins{op},...
  2941. include(ct).putativeCellType, padj(ct,li,op));
  2942. end
  2943. end
  2944. end
  2945. end
  2947. clear labels include exclude lines opsins li ct op f1 f2 colors color...
  2948. limrate pvals ncolors labels baseMLD stimMLD isGoodID baserates blvals...
  2949. stimrates stvals f1 fields ff pos nSess printfig
  2951. %% 6g + S6k: Scatter baseline vs. stim rates & Correlation
  2952. % Create in- and exclude structure
  2953. labels = {'PV','SST','GC','MC'};
  2954. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  2955. 'Granule Cell','Mossy Cell'});
  2956. %exclude = struct('groundTruthClassification',{{'+'},{'+'},{'+'},{'+'}});
  2957. exclude = struct('groundTruthClassification',{{'+','~'}},...
  2958. 'tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  2959. lines = {'PV-Cre','SST-Cre'};
  2960. opsins = {'eNpHR','ChR2'};
  2961. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  2962. printfig = true;
  2964. f1 = figure('position',[100 150 600 600]);
  2966. % Plot Lorenz curves
  2967. for li = [1 2] % PV-Cre / SST-Cre
  2968. [include.geneticLine] = deal(lines{li});
  2970. for op = [1 2] % eNpHR / ChR
  2971. [include.labels] = deal(opsins{op});
  2973. for ct = 1:length(include)
  2974. [~,baserates{ct,li,op},~] = get_metricfield(class_metrics,...
  2975. 'firingRate',... % 'firingRate', opto_1sec_baseRate
  2976. 'include',include(ct),'exclude',exclude,'omitnan',false,...
  2977. 'tukeyFence',Inf);
  2978. [~,stimrates{ct,li,op},~] = get_metricfield(class_metrics,...
  2979. 'opto_1sec_modulationRatio','include',include(ct),...
  2980. 'exclude',exclude,'omitnan',false,'tukeyFence',Inf);
  2981. stimrates{ct,li,op} = stimrates{ct,li,op}.* baserates{ct,li,op};
  2983. baserates{ct,li,op} = baserates{ct,li,op}(...
  2984. ~isnan(stimrates{ct,li,op}) & ~isinf(stimrates{ct,li,op}));
  2985. stimrates{ct,li,op} = stimrates{ct,li,op}(...
  2986. ~isnan(stimrates{ct,li,op}) & ~isinf(stimrates{ct,li,op}));
  2988. % PLOT Rate scatter
  2989. figure(f1);
  2990. subplot(4,4,op+(ct-1)*4+(li-1)*2); hold on
  2991. title(sprintf('%s %s, %s',labels{ct},lines{li},opsins{op}))
  2992. scatter(baserates{ct,li,op},stimrates{ct,li,op},...
  2993. 3, colors(ct,:), 'o', 'filled', 'MarkerFaceAlpha',.5);
  2994. set(gca,'XScale','log');
  2995. set(gca,'YScale','log');
  2996. xlim([1e-3 1e2])
  2997. ylim([1e-3 1e2])
  2998. if ct == 4; xlabel('Bsl Rate (log)'); end
  2999. if li==1 & op == 1; ylabel('Stim Rate (log)'); end
  3000. set(gca,'fontsize',10);
  3001. drawnow
  3002. clear plotrates baseplot stimplot
  3004. if ~isempty(baserates{ct,li,op})
  3005. [R,pval] = corr(baserates{ct,li,op}, stimrates{ct,li,op});
  3006. text(1e-2,1e-2,sprintf(...
  3007. 'R = %.2f\np = %.2g',R,pval),'FontSize',10);
  3008. fprintf('%s-%s, %s: R = %.2f p = %.4g, N = %i\n',lines{li},...
  3009. opsins{op},labels{ct},R,pval,length(baserates{ct,li,op}))
  3010. end
  3011. end
  3012. end
  3013. end
  3015. % Save figure
  3016. if printfig
  3017. pos = get(f1,'Position');
  3018. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  3019. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  3020. print(f1,'Fig6g_FigS8g_BslStimRateScatter.pdf','-dpdf','-r0')
  3021. end
  3023. clear labels include exclude lines opsins li ct op f1 f2 colors color...
  3024. limrate pvals ncolors labels baseMLD stimMLD isGoodID baserates blvals...
  3025. stimrates stvals
  3027. %% Suppl 6a Examples of direct and indirect light-responsive cells
  3028. % direct: PV3_230502 UID 108 (pPV, PV+); indirect: PV1_230426, UID 4 (MC)
  3029. % Load cell_metrics, spikes, opto_1sec
  3030. color = colors(1,:); % 4=MC, 1 = pPV
  3031. UID = 108;
  3032. printfig = true;
  3034. % ACG
  3035. acgT = cell_metrics.general.ccg_time;
  3036. acgBinSz = mean(diff(acgT));
  3037. acgEdg = [acgT-acgBinSz/2; acgT(end)+acgBinSz/2]';
  3039. % Plot ACG
  3040. f1 = figure('position',[200 200 400 215]);
  3041. subplot(1,2,1); hold on;
  3042. histogram('BinEdges',acgEdg,'BinCounts',...
  3043. cell_metrics.general.ccg(:,UID,UID),...
  3044. 'FaceColor',color,'FaceAlpha',1,'LineStyle','none');
  3045. title(sprintf('%s (%i)',cell_metrics.putativeCellType{UID},UID));
  3046. xlabel('Time (s)');
  3048. % Plot Raster
  3049. subplot(1,2,2);
  3050. bz_eventRasterplot(spikes,opto_1sec.timestamps(1:200,:),UID,'periEvtWdw',.5,...
  3051. 'color',color,'subsample',5);
  3053. sgtitle(cell_metrics.general.basename);
  3055. % Save figure if desired
  3056. if printfig
  3057. pos = get(f1,'Position');
  3058. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  3059. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  3060. print(f1,sprintf('%s-UID %i',cell_metrics.general.basename,UID),'-dpdf','-r0')
  3061. end
  3063. clear acgT acgBinSz acgEdg color UID printfig pos f1
  3065. %% Suppl 6b: 10 ms - 1 sec side by side w/ correlations
  3066. geneticLine = '';%'PV-Cre';
  3067. opsin = 'ChR2'; % 'eNpHR','ChR2'
  3068. field = 'opto_10ms';
  3069. labels = {'rapid','delayed pPV','delayed pSST','delayed pGC','delayed pMC'};
  3070. nGr = 5;
  3071. printfig = true;
  3073. % EVENT SELECTION
  3074. type = 'manipulations'; % Manipulations, events
  3075. sortCriterion = '_modulationIndex'; % '_modulationSD' '_modulationIndex'(default) or '_peakFreq' for chirp
  3077. % CELL SELECTION
  3078. include = struct('putativeCellType',{'','PV interneuron','SST interneuron',...
  3079. 'Granule Cell','Mossy Cell'}, 'brainRegion',repmat({'DG'},1,nGr),...
  3080. 'groundTruthClassification',{{'+'},{''},{''},{''},{''}},...
  3081. 'labels',repmat({opsin},1,nGr),'geneticLine',repmat({geneticLine},1,nGr));
  3082. exclude = struct('putativeCellType',repmat({'Unknown'},1,nGr),...
  3083. 'tags',repmat({{'InverseSpike','Bad'}},1,nGr), 'groundTruthClassification',...
  3084. {{},{'+','~'},{'+','~'},{'+','~'},{'+','~'}});
  3085. % {{},{'+'},{'+'},{'+'},{'+'}});
  3087. criteria = struct('UID',repmat({[0 Inf]},1,nGr),...
  3088. 'opto_1sec_modulationSignificanceLevelevt',repmat({[0 .01]},1,nGr),...
  3089. 'opto_1sec_modulationRatio',repmat({[1.1 Inf]},1,nGr));
  3091. % Get timestamps
  3092. for b = 1:length(class_metrics.general.batch)
  3093. if isfield(class_metrics.general.batch{b},'manipulations') &...
  3094. isfield(class_metrics.general.batch{b}.manipulations,field)
  3095. %hasfield(b) = isfield(metrics.general.batch{b}.(type),field);
  3096. tss = class_metrics.general.batch{b}.manipulations.(field).x_bins;
  3097. tsl = class_metrics.general.batch{b}.manipulations.opto_1sec.x_bins; % ASSUME 1 sec is in all opto recordings
  3098. end
  3099. end
  3101. % PLOT
  3102. f1 = figure('position',[200 20 650 145*nGr]); % Same abs height
  3103. for gr = 1:nGr
  3104. % Get PSTHs
  3105. [short_psth, ~, sUIDs] = extract_psth(class_metrics, field, 'type',type,...
  3106. 'include',include(gr),'exclude',exclude(gr),'criteria',criteria(gr),...
  3107. 'sortCriterion',sortCriterion);
  3108. [long_psth, ~, lUIDs] = extract_psth(class_metrics, 'opto_1sec', 'type',type,...
  3109. 'include',include(gr),'exclude',exclude(gr),'criteria',criteria(gr),...
  3110. 'sortCriterion',sortCriterion);
  3112. % Match IDs between plots
  3113. overlap = intersect(sUIDs, lUIDs);
  3114. [~,lLoc] = ismember(sUIDs,lUIDs);
  3115. lLoc = lLoc(lLoc>0);
  3117. % Plot Short PSTH
  3118. subplot(5,4,[4*(gr-1)+1 4*(gr-1)+2]);
  3119. display_psth(short_psth(ismember(sUIDs,overlap),:),'ts',tss*1e3,...
  3120. 'minPrctile',0,'x_units','ms','minSD',.5);
  3121. xlim([-10 30]);
  3123. % Plot Long PSTH
  3124. subplot(5,4,4*(gr-1)+3);
  3125. display_psth(long_psth(lLoc,:),'ts',tsl,'x_units','s',...
  3126. 'minPrctile',0,'minSD',.5); % Take out long-only entries
  3128. % Plot Modulation-Ratio correlations
  3129. subplot(5,4,4*(gr-1)+4);
  3131. % Rank-order correlations
  3132. sranks = tiedrank(class_metrics.([field sortCriterion])(overlap));
  3133. lranks = tiedrank(class_metrics.(['opto_1sec' sortCriterion])(overlap));
  3134. isvalid = ~isnan(sranks) & ~isnan(lranks);
  3135. N(gr) = length(find(isvalid));
  3137. scatterWreg(sranks(isvalid)',lranks(isvalid)');
  3138. if gr == nGr; xlabel('Rank 5 ms'); end
  3139. ylabel('Rank 1 sec');
  3141. [R(gr),pRank(gr)] = corr(sranks(isvalid)',lranks(isvalid)');
  3142. end
  3144. % Save figure if desired
  3145. if printfig
  3146. pos = get(f1,'Position');
  3147. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  3148. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  3149. print(f1,'FigS6b_Opto_shortVSlong.pdf','-dpdf','-r0')
  3150. end
  3152. % Multi-comparison correction for correlation p-values to console
  3153. pRanks = multCompCorr(pRank,'method','Dunn-Sidak');
  3154. for gr = 1:nGr
  3155. fprintf('%s: N = %i, R = %.2f, p(raw) = %.4g, p(Dunn-Sidak) = %.4g\n',...
  3156. labels{gr}, N(gr), R(gr), pRank(gr), pRanks(gr));
  3157. end
  3159. clear include exclude nGr gr field type labels minRate minPrctile baseline...
  3160. sortCriterion printfig f1 pos colors R labels pRank pRanks lranks sranks...
  3161. long_psth short_psth tsl tss N
  3163. %% Suppl 6c Pie-Chart 1 sec pulses - row-wise, all
  3164. lines = {'PV-Cre','SST-Cre'};
  3165. opsins = {'eNpHR','ChR2'};%'eNpHR' 'ChR2'
  3166. printfig = true;
  3168. % Create in- and exclude structure
  3169. %labels = {'PV','SST','GC','MC'};
  3170. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  3171. 'Granule Cell','Mossy Cell'},'brainRegion','DG');
  3172. exclude = struct('groundTruthClassification',{{'+','~'}},...
  3173. 'tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  3175. % Plot
  3176. f1 = figure('position',[200 200 700 700]);
  3178. for op = 1:2
  3179. [include.labels] = deal(opsins{op});
  3180. for li = 1:2
  3181. [include.geneticLine] = deal(lines{li});
  3182. for ct = 1:4
  3183. subplot(4,4,ct + 4*(li-1) + 8*(op-1));
  3184. plot_response_piechart(class_metrics,'labels',{''},'alpha',.01,...
  3185. 'signfield','_modulationSignificanceLevelevt','minchange',.1,...
  3186. 'include',include(ct),'exclude',exclude);
  3187. drawnow
  3188. end
  3189. end
  3190. end
  3192. % Save figure
  3193. if printfig
  3194. pos = get(f1,'Position');
  3195. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  3196. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  3197. print(f1,'FigS6c_ChR_eNpHR_1sec_piechart.pdf','-dpdf','-r0')
  3198. end
  3200. clear include exclude f1 f2 printfig field pos sel type ds ts psth psthNorm ct li op...
  3201. lines opsins include exclude
  3203. %% Suppl 6d,f: Classification of direct- and indirect responding cells
  3204. % For S6d, run this with metrics classified on all PV and SST data
  3205. % (see below for instructions for eNpHR trained classifier control).
  3207. % General
  3208. labels = {'PV','SST','GC','MC'};
  3209. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7];
  3210. printfig = true;
  3212. % CELL SELECTION
  3213. classinclude = struct('putativeCellType',{'PV interneuron','SST interneuron','Granule Cell','Mossy Cell'},...
  3214. 'labels',repmat({'ChR2'},1,4),'brainRegion',repmat({'DG'},1,4));
  3216. trueinclude = struct('putativeCellType',{'','','Granule Cell','Mossy Cell'},...
  3217. 'groundTruthClassification',{{'PV+'},{'SST+'},{''},{''}},...
  3218. 'labels',repmat({'ChR2'},1,4),'brainRegion',repmat({'DG'},1,4));
  3220. exclude = struct('putativeCellType',{'Unknown'},'tags',{{'InverseSpike','Bad'}});
  3222. % GET TRAINING AND CLASSIFIER LABELS
  3223. trueLbl = NaN(1,length(batch_metrics.UID));
  3224. classLbl = NaN(1,length(class_metrics.UID));
  3226. for ii = 4:-1:1 % Do reverse to make sure opto+ can overwrite classifier results (1 MC opto+ in PV mice).
  3227. [~,selected] = select_cells(batch_metrics,'include',trueinclude(ii),...
  3228. 'exclude',exclude);
  3229. trueLbl(selected) = ii;
  3230. end
  3232. for ii = 4:-1:1
  3233. [~,selected] = select_cells(class_metrics,'include',classinclude(ii),...
  3234. 'exclude',exclude);
  3235. classLbl(selected) = ii;
  3236. end
  3238. % Make confusion matrix
  3239. figure; cm = confusionchart(confusionmat(...
  3240. trueLbl,classLbl),{'1 PV','2 SST','3 GC','4 MC'});
  3241. cm.RowSummary = 'row-normalized';
  3242. cm.ColumnSummary = 'column-normalized';
  3245. % PIE CHART FOR 10ms AND 1 sec RESPONSIVE CELLS
  3246. [~,ChRpv] = select_cells(batch_metrics,'include',trueinclude(1),'exclude',exclude);
  3247. [~,ChRsst] = select_cells(batch_metrics,'include',trueinclude(2),'exclude',exclude);
  3249. % Cells w/ pos response during 1 sec stimulus (PV- and SST-ChR2 mice, respectively)
  3250. [~,pv1secResp] = select_cells(class_metrics,...
  3251. 'include',struct('brainRegion',{'DG'},'labels',{'ChR2'},'geneticLine',{'PV-Cre'}),...
  3252. 'exclude',struct('groundTruthClassification',{{'+','~'}},'putativeCellType',{'Unknown'},...
  3253. 'tags',{{'InverseSpike','Bad'}}),...
  3254. 'criteria',struct('opto_1sec_modulationSignificanceLevel',[0 .01],...
  3255. 'opto_1sec_modulationRatio',[1.1 Inf]));
  3256. [~,sst1secResp] = select_cells(class_metrics,...
  3257. 'include',struct('brainRegion',{'DG'},'labels',{'ChR2'},'geneticLine',{'SST-Cre'}),...
  3258. 'exclude',struct('groundTruthClassification',{{'+','~'}},...
  3259. 'putativeCellType',{'Unknown'},'tags',{{'InverseSpike','Bad'}}),...
  3260. 'criteria',struct('opto_1sec_modulationSignificanceLevel',[0 .01],...
  3261. 'opto_1sec_modulationRatio',[1.1 Inf]));
  3263. % Make pie chart w/ labels
  3264. f1 = figure('position',[200 200 300 300]);
  3265. groups = {ChRpv,ChRsst,pv1secResp,sst1secResp};
  3266. titles = {'PV-Cre direct','SST-Cre direct','PV-Cre indirect','SST-Cre indirect'};
  3268. for gr = 1:4
  3269. subplot(2,2,gr);
  3270. theselbl = classLbl(groups{gr});
  3271. theselbl = theselbl(~isnan(theselbl));
  3272. Ncells = histcounts(theselbl,[.5 1.5 2.5 3.5 4.5]);
  3273. pie(Ncells);%,labels);
  3274. ax = gca; ax.Colormap = colors;
  3275. title(titles{gr},'FontSize',8);
  3276. fprintf('%s: %s: %i/%i, %s: %i/%i, %s: %i/%i, %s: %i/%i\n',titles{gr},...
  3277. labels{1},Ncells(1),sum(Ncells),labels{2},Ncells(2),sum(Ncells),...
  3278. labels{3},Ncells(3),sum(Ncells),labels{4},Ncells(4),sum(Ncells))
  3279. end
  3281. % Save figure
  3282. if printfig
  3283. pos = get(f1,'Position');
  3284. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  3285. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  3286. print(f1,'1sec_direct_indirect_Pie.pdf','-dpdf','-r0')
  3287. end
  3289. clear colors labels classinclude ChRpv ChRsst sst1secResp pv1secResp classLbl trueLbl...
  3290. groups title pos printfig f1 theselbl Ncells ax cellIDs chrIDs sstIDs
  3292. %% Suppl 6e: eNpHR -> ChR cross validation
  3293. % Re-run classifier training (DG_IN_Scripts/Make table for classifier
  3294. % learning and include 'labels',{'eNpHR'} in include criteria for PV+ and
  3295. % SST+ cells. Apply SVM across full dataset (include all cells).
  3296. % Then run the section above (Fig 6d,f, 'Classification of direct and indirect
  3297. % responding cells').
  3299. %% Suppl 6g,h: opto_1sec_modulationPeakResponseTime, spikeJitter
  3300. field = 'opto_1sec_modulationPeakResponseTime'; % 'opto_1sec_modulationPeakResponseTime','opto_1sec_spikeJitter' 'opto_1sec_peak_ratio'
  3301. thisYlabel = 'Peak Response Time (s)'; % 'Peak Response Time (s)', 'Spike jitter (s)'
  3302. labels = {'rapid','PV','SST','GC','MC'};
  3303. transformation = ''; % '', 'log2', 'log10'
  3304. printfig = true;
  3306. nGr = 5;
  3307. include = struct('putativeCellType',{'','PV interneuron','SST interneuron',...
  3308. 'Granule Cell','Mossy Cell'}, 'brainRegion',repmat({'DG'},1,nGr),...
  3309. 'groundTruthClassification',{{'+'},{''},{''},{''},{''}},...
  3310. 'labels',repmat({'ChR2'},1,5));
  3311. exclude = struct('groundTruthClassification',{{},{'+','~'},{'+','~'},{'+','~'},{'+','~'}},...
  3312. 'putativeCellType',repmat({'Unknown'},1,5),...
  3313. 'tags',repmat({{'InverseSpike','Bad'}},1,nGr));
  3314. criteria = struct('opto_1sec_modulationSignificanceLevelevt',...
  3315. {[0 1],[0 .01],[0 .01],[0 .01],[0 .01]},...
  3316. 'opto_1sec_modulationRatio',{[-Inf Inf],[1.1 Inf],[1.1 Inf],[1.1 Inf],[1.1 Inf]});
  3317. colors = {[1 .2 .2],[.56 .39 .93], [0 .72 .53], [.83 .78 .46], [.9 0 .7]};
  3319. [valGrp, valMtrx, valCell] = get_metricfield(class_metrics,field,...
  3320. 'include',include, 'exclude',exclude, 'criteria',criteria,...
  3321. 'transformation',transformation, 'tukeyFence',3);
  3323. % Plotting
  3324. f1 = figure('position',[200 450 140 170]); % Same aspect ratio
  3325. hold on
  3327. % Plot spread for SVM classified cell types
  3328. plotSpread(valGrp(:,1),'distributionIdx',...
  3329. valGrp(:,2),'spreadFcn',{'xp',[]},'spreadWidth',.9,...
  3330. 'distributionColors',[0 0 0],...
  3331. 'distributionMarkerSize',1,'binWidth',.05);
  3333. % Plot only svm classified cell types as violin plots
  3334. distributionPlot(valCell,'color',colors,...
  3335. 'showMM',6,'distWidth',.9,'addSpread',0,'histOpt',1,'divFactor',1,...
  3336. 'FaceAlpha',.5);
  3338. xticklabels(labels);
  3339. xtickangle(45);
  3340. ylabel(thisYlabel); % 'Peak Response Time (s)', 'Spike jitter (s)'
  3341. ylim([-.5 1.5]); % [-.5 1.5] For peak response time, [-.1 .4] for jitter
  3343. % % Stats
  3344. fprintf('Direct vs. indirect excited cells ChR2\n');
  3345. kruskalwallisNdunns(valMtrx,labels);
  3347. % Save figure
  3348. if printfig
  3349. pos = get(f1,'Position');
  3350. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  3351. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  3352. print(f1,sprintf('%s_direct_vs_indirect.pdf',field),'-dpdf','-r0')
  3353. end
  3355. clear valCell valGrp include exclude criteria field printfig pos f1 nGr colors...
  3356. labels thisYlabel transformation valMtrx
  3358. %% Suppl 6i,j: Peer-prediction opto_1sec
  3359. % CELL SELECTION
  3360. nGr = 5;
  3361. labels = {'rapid','PV','SST','GC','MC'};
  3362. include = struct('putativeCellType',{'','PV interneuron','SST interneuron',...
  3363. 'Granule Cell','Mossy Cell'}, 'brainRegion',repmat({'DG'},1,nGr),...
  3364. 'groundTruthClassification',{{'+'},{''},{''},{''},{''}},...
  3365. 'labels',repmat({'ChR2'},1,5));
  3366. exclude = struct('groundTruthClassification',{{},{'+','~'},{'+','~'},{'+','~'},{'+','~'}},...
  3367. 'putativeCellType',repmat({'Unknown'},1,5),...
  3368. 'tags',repmat({{'InverseSpike','Bad'}},1,nGr));
  3369. criteria = struct('opto_1sec_modulationSignificanceLevelevt',...
  3370. {[0 1],[0 .01],[0 .01],[0 .01],[0 .01]},...
  3371. 'opto_1sec_modulationRatio',{[-Inf Inf],[1.1 Inf],[1.1 Inf],[1.1 Inf],[1.1 Inf]});
  3372. %'opto_1sec_modulationRatio',{[-Inf Inf],[1.1 Inf],[1.1 Inf],[1.1 Inf],[1.1 Inf]});
  3373. RsqThr = 0; % Min adjusted Rsq on training data for inclusion -- was 0.25 on 1/31/2026
  3374. colors = {[1 .2 .2],[.56 .39 .93], [0 .72 .53], [.83 .78 .46], [.9 0 .7]};
  3376. % PEER PREDICTION
  3377. if ~exist('peerpred','var')
  3378. peerpred = peerPredictOpto1sec(class_metrics, 'include',include,...
  3379. 'exclude',exclude, 'criteria',criteria, 'binsz',.25); %.02
  3380. end
  3382. % Extract fit quality values
  3383. errfield = 'MSE'; % 'R', 'Rsq', 'MSE'
  3384. valMtrx = peerpred.(['val' errfield]);
  3385. stimMtrx = peerpred.(['stim' errfield]);
  3387. % Get training Rsq and exclude below threshold
  3388. RsqTrng = -Inf(size(peerpred.Glm));
  3389. for gr = 1:5
  3390. for n = 1:size(peerpred.Glm,2)
  3391. if ~isempty(peerpred.Glm{gr,n})
  3392. RsqTrng(gr,n) = peerpred.Glm{gr,n}.Rsquared.Adjusted;
  3393. end
  3394. end
  3395. valMtrx{gr} = valMtrx{gr}(RsqTrng(gr,:) > RsqThr);
  3396. stimMtrx{gr} = stimMtrx{gr}(RsqTrng(gr,:) > RsqThr);
  3397. end
  3399. valMtrx = cellfun(@sqrt, valMtrx, 'UniformOutput',false);
  3400. stimMtrx = cellfun(@sqrt, stimMtrx, 'UniformOutput',false);
  3401. valMtrx = cellfun(@log10, valMtrx, 'UniformOutput',false);
  3402. stimMtrx = cellfun(@log10, stimMtrx, 'UniformOutput',false);
  3404. for gr = 1:5
  3405. [~,isoutlierVal] = tukeyOutlierRemoval(valMtrx{gr},3);
  3406. [~,isoutlierStim] = tukeyOutlierRemoval(stimMtrx{gr},3);
  3408. % Remove major outliers
  3409. valMtrx{gr} = valMtrx{gr}(~isoutlierVal & ~isoutlierStim);
  3410. stimMtrx{gr} = stimMtrx{gr}(~isoutlierVal & ~isoutlierStim);
  3411. end
  3413. % PLOT LOG MSE (VIOLIN PLOTS)
  3414. f1 = figure('position',[200 450 280 220]); % Same aspect ratio
  3415. hold on
  3417. valGrp = cell2groupedvector(valMtrx);
  3418. stimGrp = cell2groupedvector(stimMtrx);
  3420. plotSpread(valGrp(:,1),'distributionIdx',...
  3421. (valGrp(:,2)-1)*2+1,'spreadFcn',{'lin',10},'spreadWidth',.9,...
  3422. 'distributionColors',[0 0 0],...
  3423. 'distributionMarkerSize',1,'binWidth',.05);
  3425. plotSpread(stimGrp(:,1),'distributionIdx',...
  3426. stimGrp(:,2)*2,'spreadFcn',{'lin',10},'spreadWidth',.9,...
  3427. 'distributionColors',[0 0 0],...
  3428. 'distributionMarkerSize',1,'binWidth',.05);
  3430. % Plot direct and indirect responding cell's peer prediction quality
  3431. combMtrx = cat(1,valMtrx,stimMtrx);
  3432. distributionPlot(combMtrx,'color',{[1 .2 .2],[1 .2 .2],[.56 .39 .93],[.56 .39 .93],...
  3433. [0 .72 .53],[0 .72 .53],[.83 .78 .46],[.83 .78 .46],[.9 0 .7],[.9 0 .7]},...
  3434. 'showMM',6,'distWidth',.9,'addSpread',0,'histOpt',1,'divFactor',1,...
  3435. 'FaceAlpha',.5);
  3437. xticks([1.5 3.5 5.5 7.5 9.5]);
  3438. xticklabels({'rapid','PV','SST','GC','MC'});
  3439. xtickangle(45);
  3440. ylabel(['Log' errfield ' validation']);
  3441. ylim([-1 3]);
  3443. % PLOT LOG MSE DIFFERENCE VAL TO STIM
  3444. f4 = figure('position',[200 450 140 170]); % Same aspect ratio
  3445. hold on
  3447. diffMtrx = cellfun(@minus, stimMtrx, valMtrx, 'UniformOutput',false);
  3448. diffGrp = cell2groupedvector(diffMtrx);
  3450. plotSpread(diffGrp(:,1),'distributionIdx',...
  3451. diffGrp(:,2),'spreadFcn',{'lin',10},'spreadWidth',.9,...
  3452. 'distributionColors',[0 0 0],...
  3453. 'distributionMarkerSize',1,'binWidth',.05);
  3455. % Plot direct and indirect responding cell's peer prediction quality
  3456. distributionPlot(diffMtrx,'color',{[1 .2 .2],[.56 .39 .93],[0 .72 .53],[.83 .78 .46],[.9 0 .7]},...
  3457. 'showMM',6,'distWidth',.9,'addSpread',0,'histOpt',1,'divFactor',1,...
  3458. 'FaceAlpha',.5);
  3460. xticklabels({'rapid','PV','SST','GC','MC'});
  3461. xtickangle(45);
  3462. ylabel(['Log' errfield ' difference (Stim-Val)']);
  3463. %ylim([-1 3]);
  3465. % Stats (error difference)
  3466. fprintf('Direct vs. indirect excited cells ChR2, prediction error difference');
  3467. kruskalwallisNdunns(catuneven(diffMtrx,NaN),labels);
  3469. % CORRELATION PLOTS (Actual rate / predicted rate during stim and
  3470. % validation).
  3471. f2 = figure('position',[100 100 950 350]); hold on;
  3472. for gr = 1:5
  3473. %Plot Stim data
  3474. subplot(2, 5, gr);
  3475. stimActual = real(log10(peerpred.stimActual{gr}(RsqTrng(gr,:) > RsqThr)));
  3476. stimPred = real(log10(peerpred.stimPred{gr}(RsqTrng(gr,:) > RsqThr)));
  3478. scatterWreg(stimActual,stimPred,'MarkerSz',8,'MarkerSymbol','o',...
  3479. 'MarkerFaceAlpha',.5,'MarkerFaceColor',colors{gr});
  3480. xlabel('Stim Observed'); ylabel('Stim Predicted');
  3481. xlim([-3 2]); ylim([-3 2]);
  3482. xticks([-3:2]); yticks([-3:2]);
  3483. xticklabels([.001 .01 .1 1 10 100]);
  3484. yticklabels([.001 .01 .1 1 10 100]);
  3485. xtickangle(45);
  3487. [R(gr,1),p(gr,1)] = corr(stimActual,stimPred);
  3488. N(gr,1) = length(stimActual);
  3490. % Plot Validation data
  3491. subplot(2, 5, gr+5);
  3492. valActual = real(log10(peerpred.valActual{gr}(RsqTrng(gr,:) > RsqThr)));
  3493. valPred = real(log10(peerpred.valPred{gr}(RsqTrng(gr,:) > RsqThr)));
  3495. scatterWreg(valActual,valPred,'MarkerSz',8,'MarkerSymbol','o',...
  3496. 'MarkerFaceAlpha',.5,'MarkerFaceColor',colors{gr});
  3497. xlabel('Validation Observed'); ylabel('Validation Predicted');
  3498. xlim([-3 2]); ylim([-3 2]);
  3499. xticks([-3:2]); yticks([-3:2]);
  3500. xticklabels([.001 .01 .1 1 10 100]);
  3501. yticklabels([.001 .01 .1 1 10 100]);
  3502. xtickangle(45);
  3504. [R(gr,2),p(gr,2)] = corr(valActual,valPred);
  3505. N(gr,2) = length(valActual);
  3506. end
  3508. % Plot corrected p-values:
  3509. pcorr = multCompCorr(p,'method','Dunn-Sidak');
  3510. fprintf('\nActivity correlation bsl-stim\n');
  3511. for gr = 1:5
  3512. fprintf('%s, validation: N = %i, R = %.2f, p(raw) = %.4g, p(corr) = %.4g\n',...
  3513. labels{gr},N(gr,2),R(gr,2),p(gr,2),pcorr(gr,2))
  3514. fprintf('%s, stim: N = %i, R = %.2f, p(raw) = %.4g, p(corr) = %.4g\n',...
  3515. labels{gr},N(gr,1),R(gr,1),p(gr,1),pcorr(gr,1))
  3516. end
  3518. clear include exclude criteria valGrp stimGrp valMtrx stimMtrx gr R R2 p n...
  3519. colors errfield f1 f2 f3 f4 nGr RsqThr RsqTrng stimActual stimPred valActual valPred...
  3520. actualRatio predRatio combMtrx diffMtrx diffGrp labels isoutlierStim isoutlierVal...
  3521. pcorr transformation N
  3524. %% Supplementary Figure 8
  3526. %% Suppl 8a-c: Lorenz curves, gini differences, CV differences
  3527. % Run Fig 6e,f code above
  3529. %% Suppl 8d-f: Spike-subsample control for Lorenz curves, gini differences, CV differences
  3530. % Create in- and exclude structure
  3531. printfig = true;
  3532. labels = {'PV-HR','SST-HR','PV-ChR','SST-ChR'};
  3533. include = struct('putativeCellType',{'PV interneuron','SST interneuron',...
  3534. 'Granule Cell','Mossy Cell'});
  3535. %exclude = struct('groundTruthClassification',{{'+'},{'+'},{'+'},{'+'}});
  3536. exclude = struct('groundTruthClassification',{{'+','~'}},...
  3537. 'tags',{{'Bad','InverseSpike'}},'putativeCellType',{'Unknown'});
  3538. lines = {'PV-Cre','SST-Cre'};
  3539. opsins = {'eNpHR','ChR2'};
  3540. colors = [.56 .39 .93; 0 .72 .53; .83 .78 .46; .9 0 .7; .56 .39 .93; 0 .72 .53];
  3541. fields = {'gini','CV'}; % For errorbar plots (diff stim vs bsl)
  3543. % Get Lorenz curves, gini, CV
  3544. if ~exist('LorRes','var') % Takes long to load
  3545. for li = [1 2] % PV-Cre / SST-Cre
  3546. [include.geneticLine] = deal(lines{li});
  3548. for op = [1 2] % eNpHR / ChR
  3549. [include.labels] = deal(opsins{op});
  3551. fprintf('%s, %s\n',lines{li},opsins{op});
  3552. LorRes{li,op} = get_1sec_Lorenz_Curves_rateMatch(class_metrics,...
  3553. 'include',include,'exclude',exclude,'show_results',false,...
  3554. 'minNspikes',1, 'minNcells',4);
  3555. end
  3556. end
  3557. end
  3559. % LORENZ CURVES
  3560. f1= figure('position',[100 100 500 450]); hold on;
  3561. for li = 1:2
  3562. for op = 1:2
  3563. for ct = 1:length(include)
  3564. subplot(4,length(include),ct + (op-1)*length(include) + (li-1)*2*length(include))
  3566. % Plot eNpHR results
  3567. plot([0 100],[0 1],':','color',[.5 .5 .5]);
  3568. nSess = length(find(LorRes{li,op}.nCells(:,ct)>0));
  3569. shadedErrorBar(1:100, nanmean(LorRes{li,op}.MeanLorenzCurveNorm(:,:,ct,1),1),...
  3570. nanstd(LorRes{li,1}.MeanLorenzCurveNorm(:,:,ct,1),1)./sqrt(nSess),...
  3571. 'lineprops',{'color',colors(ct,:)});
  3572. shadedErrorBar(1:100, nanmean(LorRes{li,op}.MeanLorenzCurveNorm(:,:,ct,2),1),...
  3573. nanstd(LorRes{li,1}.MeanLorenzCurveNorm(:,:,ct,2),1)./sqrt(nSess),...
  3574. 'lineprops',{'color',[.5 .5 .5]});
  3575. text(10,.8,sprintf('%s\n%s',lines{li},opsins{op}),'FontSize',8);
  3576. drawnow
  3577. end
  3578. end
  3579. end
  3581. % Save figure
  3582. if printfig
  3583. pos = get(f1,'Position');
  3584. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  3585. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  3586. print(f1,'SpikeNoControl_LorenzCurves.pdf','-dpdf','-r0')
  3587. end
  3589. % PLOT GINI, CV, ETC
  3590. for ff = 1:2
  3591. % GET VALUES
  3592. stvals = cat(1, LorRes{1,1}.(fields{ff})(:,:,1), LorRes{2,1}.(fields{ff})(:,:,1),...
  3593. LorRes{1,2}.(fields{ff})(:,:,1), LorRes{2,2}.(fields{ff})(:,:,1)); % To sort by condition, cat(2,... and omit reshape
  3594. blvals = cat(1, LorRes{1,1}.(fields{ff})(:,:,2), LorRes{2,1}.(fields{ff})(:,:,2),...
  3595. LorRes{1,2}.(fields{ff})(:,:,2), LorRes{2,2}.(fields{ff})(:,:,2));
  3596. stvals = reshape(stvals, size(LorRes{1,1}.gini,1), length(include)*length(lines)*length(opsins));
  3597. blvals = reshape(blvals, size(LorRes{1,1}.gini,1), length(include)*length(lines)*length(opsins));
  3599. % RAW VALUES (DUAL VIOLIN PLOTS)
  3600. figure('position',[100 100 500 150]); hold on;
  3602. % baseline (left, grey)
  3603. plotSpread(blvals,'spreadFcn',{'lin',10},...
  3604. 'spreadWidth',.9,'distributionColors',[.5 .5 .5],'distributionMarkerSize',3,...
  3605. 'spreadOri','left','distributionMarkerAlpha',.5,'distributionMarkers','o');
  3606. distributionPlot(blvals,'histOri','left','color',[0 0 0],...
  3607. 'widthDiv',[2 1],'showMM',6,'FaceAlpha',.5);
  3609. % Stim (right)
  3610. plotSpread(stvals,'spreadFcn',{'lin',10},... % [sic!], use wider (10) spread here 2/2 lower n datapoints
  3611. 'spreadWidth',.9,'distributionColors',[.5 .5 .5],'distributionMarkerSize',3,...
  3612. 'spreadOri','right','distributionMarkerAlpha',.5,'distributionMarkers','o');
  3613. distributionPlot(stvals,'histOri','right','color',num2cell(colors([1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4],:),2),...
  3614. 'widthDiv',[2 2],'showMM',6,'FaceAlpha',.5);
  3615. %ylim([.4 1.1]);
  3616. ylabel(fields{ff});
  3619. % STIM-BSL DELTA PLOT
  3620. f1 = figure('position',[100 100 400 250]); hold on;
  3621. % distributionPlot(stvals-blvals,'showMM',6,'FaceAlpha',.5,'histOpt',1,'divFactor',1,...
  3622. % 'color',num2cell(colors([1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4],:),2));
  3623. plot([0 size(stvals,2)+1],[0 0],'--','Color',[.5 .5 .5]); % Zero line
  3624. plotSpread(stvals-blvals,'spreadFcn',{'lin',5},...
  3625. 'spreadWidth',.9,'distributionColors',[0 0 0],'distributionMarkerSize',4,...
  3626. 'distributionMarkerAlpha',1,'distributionMarkers','o');
  3627. customErrorBar(stvals-blvals, 'errtype','STD', 'alpha',.7,...
  3628. 'markerSize',100, 'LineWidth',1.2,...
  3629. 'colors',colors([1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4],:));
  3630. xlim([0 size(stvals,2)+1]);
  3631. %ylim([-.3 .4]);
  3632. ylabel(sprintf('%s (stim-bsl)',fields{ff}));
  3633. xticklabels(repmat(labels,1,4));
  3634. xtickangle(45);
  3636. % Save figure
  3637. if printfig
  3638. pos = get(f1,'Position');
  3639. set(f1,'PaperPositionMode','Auto','PaperUnits','points',...
  3640. 'PaperSize',[pos(3), pos(4)],'Renderer','Painters')
  3641. print(f1,sprintf('SpikeNoControl_Opto_1sec_%s_change.pdf',fields{ff}),'-dpdf','-r0')
  3642. end
  3644. % STATISTICS
  3645. for ct = 1:length(include)
  3646. clear blvals stvals groupVals
  3648. % Statistics @console
  3649. fprintf('\n%s:',include(ct).putativeCellType)
  3650. fprintf('\n%s (raw p):\n',fields{ff}) % For statistics in console
  3652. for li = 1:2
  3653. for op = 1:2
  3654. blvals(:,li,op) = LorRes{li,op}.(fields{ff})(:,ct,2); % nSess x nGroup(ct) x [1 (stim) 2 (bsl)]
  3655. stvals(:,li,op) = LorRes{li,op}.(fields{ff})(:,ct,1);
  3657. try
  3658. pvals(ct,li,op) = pairedSampleTest(blvals(:,li,op), stvals(:,li,op),...
  3659. 'labels',{'baseline','stim'},'verbose',true);
  3660. catch
  3661. pvals(ct,li,op) = NaN;
  3662. fprintf('\n');
  3663. end
  3664. end
  3665. end
  3667. % ANOVA
  3668. groupVals = reshape(stvals-blvals, size(stvals,1), size(stvals,2)*size(stvals,3));
  3669. fprintf('%s ANOVA:\n',fields{ff})
  3670. kruskalwallisNdunns(groupVals, labels);
  3671. end
  3673. % Multi-Comparison Corrected p-values
  3674. padj = multCompCorr(pvals,'method','Dunn-Sidak'); % 'Holm-Bonferroni', 'Dunn-Sidak'
  3675. fprintf('%s (adjusted p-values):\n',fields{ff});
  3676. for op = [1 2]
  3677. for li = [1 2]
  3678. for ct = 1:length(include)
  3679. fprintf('%s-%s: %s (stim vs. bsl), p = %.4g\n',lines{li}, opsins{op},...
  3680. include(ct).putativeCellType, padj(ct,li,op));
  3681. end
  3682. end
  3683. end
  3684. end
  3686. clear labels include exclude lines opsins li ct op f1 f2 colors color...
  3687. limrate pvals ncolors labels baseMLD stimMLD isGoodID baserates blvals...
  3688. stimrates stvals f1 fields ff pos nSess printfig statlabels valTbl groupVals padj
  3690. %% Suppl 8d: Baseline vs. Stim firing rate correlations
  3691. % Runf FIg 6g code above

DG_IN_Figures_final.m at commit 1566f35, no license · at the source

Overview

Authors: Thomas Hainmueller1,2,3, Elisabeth A. Heynold1,4, Nicholas Paleologos1, Ivan G. Raikov4, Ivan Soltesz4, György Buzsáki1,5,6
ORCID iDs: Thomas Hainmueller, Nicholas Paleologos, Ivan G. Raikov, György Buzsáki
  1. Neuroscience Institute, New York University Langone Medical Center, New York, NY 10016, USA
  2. Department of Psychiatry, New York University Langone Medical Center, New York, NY 10016, USA
  3. Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Palo Alto, CA 94305, USA
  4. Department of Neurosurgery, Stanford University School of Medicine, Palo Alto, CA 94305, USA
  5. Department of Neurology, New York University Langone Medical Center, New York, NY 10016, USA
  6. Lead contact
Institutions: NYU Langone Health (United States); Stanford University (United States); Stanford Medicine (United States)
Journal: Neuron, volume 114, issue 17, pages 3280-3294.e7
Dates: published online 20 April 2026; in print 2 September 2026
Type: Research article · Language: English
License: CC BY-NC-ND
Identifiers: DOI 10.1016/j.neuron.2026.03.034 · PMID 42013854 · PMCID PMC13186253 · OpenAlex W4411053555
Open access: green, a free copy (OpenAlex)
Preprint: doi:10.2139/ssrn.5282602
Status: code verified
Categories: mouse (organism), systems (subfield)
Methods: Spectral & time-frequency, Statistics, Smoothing, state filtering, decompositions, Machine learning, Connectivity, fMRI & imaging, Single-unit activity, calcium imaging, Physiology & signal measures
Keywords: Somatostatin, Hippocampus, Sleep, Parvalbumin, Interneuron, dentate gyrus, Theta, Optogenetics, Dentate Spikes
MeSH: Dentate Gyrus*, Interneurons*, Nerve Net*, Action Potentials, Animals, Male, Mice, Mice, Inbred C57BL, Mice, Transgenic, Mossy Fibers, Hippocampal, Optogenetics, Parvalbumins, Somatostatin (* major topic)
Topic: Neural dynamics and brain function (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: NIMH NIH HHS (R01 MH139216, R01 MH122391); NINDS NIH HHS (U19 NS107616, RF1 NS133381, R01 NS142069); Brain &amp;amp; Behavior Research Foundation (32940); German Research Foundation; Finding A Cure for Epilepsy and Seizures; National Institute of Neurological Disorders and Stroke; National Alliance for Research on Schizophrenia and Depression; National Institute of Mental Health
Citations: cited by 3 papers (Europe PMC); 93 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.

Repositories

Its files are read in the Code ↔ Paper reader above, with 40 matches between paragraphs and lines of code.

ThomasHainmueller/HainmuellerHeynold_et_al_2025

License: none: the authors keep all their rights
State: the link answers, verified on 29 September 2026
Evidence: files inventoried
Commit: 1566f35fc1522f1cefd7a3f59b85e76a9925df3d, 22 March 2026
Languages: MATLAB (595), C (1)
Size: 662 files, 596 scripts
Software Heritage: not archived
Found in: “Data and code availability”
Holds: README, tests
Not found: license file, CITATION.cff, environment file, continuous integration, documentation
Tools: Statistics and Machine Learning Toolbox (122 files), Signal Processing Toolbox (67 files), Image Processing Toolbox (19 files), Optimization Toolbox (5 files), shadedErrorBar (5 files), Parallel Computing Toolbox (4 files), CircStat (3 files), boundedline (2 files), EEGLAB (2 files), Chronux (1 file), Curve Fitting Toolbox (1 file)
Availability: 1 check, the latest on 29 September 2026: the link answers
  • 29 September 2026: the link answers
597 files
At the source: github.com/ThomasHainmueller/HainmuellerHeynold_et_al_2025

soltesz-lab/DG-Interneurons

License: GPL-3.0
State: the link answers, verified on 29 September 2026
Evidence: files inventoried
Commit: 160ca41868460bbf0770ff6e842b2da10d2b8eed, 16 March 2026
Languages: Python (28)
Size: 31 files, 28 scripts
Software Heritage: not archived
Found in: “Data and code availability”
Holds: README, license file, tests
Not found: CITATION.cff, environment file, continuous integration, documentation
Tools: NumPy (25 files), PyTorch (23 files), Matplotlib (18 files), SciPy (15 files), h5py (6 files), seaborn (5 files), pandas (4 files), NetworkX (1 file), scikit-learn (1 file)
Availability: 1 check, the latest on 29 September 2026: the link answers
  • 29 September 2026: the link answers
30 files
At the source: github.com/soltesz-lab/DG-Interneurons

The paper's code and data availability statement is in the Data section.

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:

  • 2 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 624 scripts, each with its path and the digest of its content;
  • 40 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.

Code and data availability statement

The paper has a code and data 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/10.1016/j.neuron.2026.03.034.

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, 29 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 6 authors, 9 keywords, 13 MeSH terms, 8 funders, 91 references.

Cite

This paper

Hainmueller, T., Heynold, E. A., Paleologos, N., Raikov, I. G., Soltesz, I., & Buzsáki, G. (2026). Dentate gyrus interneurons modulate winner-take-all network dynamics in freely behaving mice. Neuron, 114(17), 3280-3294.e7. https://doi.org/10.1016/j.neuron.2026.03.034

BibTeX

@article{hainmueller2026dentate,
author = {Hainmueller, Thomas and Heynold, Elisabeth A. and Paleologos, Nicholas and Raikov, Ivan G. and Soltesz, Ivan and Buzsáki, György},
title = {{Dentate gyrus interneurons modulate winner-take-all network dynamics in freely behaving mice}},
journal = {Neuron},
year = {2026},
month = apr,
volume = {114},
number = {17},
pages = {3280--3294.e7},
publisher = {Cell Press},
issn = {0896-6273},
doi = {10.1016/j.neuron.2026.03.034},
url = {https://doi.org/10.1016/j.neuron.2026.03.034},
pmid = {42013854},
pmcid = {PMC13186253}
}

RIS

TY - JOUR
AU - Hainmueller, Thomas
AU - Heynold, Elisabeth A.
AU - Paleologos, Nicholas
AU - Raikov, Ivan G.
AU - Soltesz, Ivan
AU - Buzsáki, György
TI - Dentate gyrus interneurons modulate winner-take-all network dynamics in freely behaving mice
T2 - Neuron
J2 - Neuron
PY - 2026
DA - 2026/04/20
VL - 114
IS - 17
SP - 3280
EP - 3294.e7
SN - 0896-6273
PB - Cell Press
DO - 10.1016/j.neuron.2026.03.034
UR - https://doi.org/10.1016/j.neuron.2026.03.034
LA - en
ER -

CSL-JSON

{
"id": "10.1016/j.neuron.2026.03.034",
"type": "article-journal",
"title": "Dentate gyrus interneurons modulate winner-take-all network dynamics in freely behaving mice",
"container-title": "Neuron",
"author": [
{
"family": "Hainmueller",
"given": "Thomas"
},
{
"family": "Heynold",
"given": "Elisabeth A."
},
{
"family": "Paleologos",
"given": "Nicholas"
},
{
"family": "Raikov",
"given": "Ivan G."
},
{
"family": "Soltesz",
"given": "Ivan"
},
{
"family": "Buzsáki",
"given": "György"
}
],
"container-title-short": "Neuron",
"volume": "114",
"issue": "17",
"page": "3280-3294.e7",
"DOI": "10.1016/j.neuron.2026.03.034",
"PMID": "42013854",
"PMCID": "PMC13186253",
"ISSN": "0896-6273",
"publisher": "Cell Press",
"URL": "https://doi.org/10.1016/j.neuron.2026.03.034",
"language": "en",
"issued": {
"date-parts": [
[
2026,
4,
20
]
]
}
}

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:10.1523/jneurosci.2001-25.2026 [code]
Dynamics of Dentate Gyrus Place Cells and Dentate Spikes during Spatial and Nonspatial Changes in Environments.
Journal: The Journal of neuroscience : the official journal of the Society for Neuroscience
In common: Chronux, CircStat, Curve Fitting Toolbox, 9 other tools, systems, 11 references
[2] doi:10.1016/j.celrep.2026.117646 [code]
Medial entorhinal-hippocampal desynchronization parallels the emergence of memory impairment in a mouse model of Alzheimer's disease pathology.
Journal: Cell reports
In common: Chronux, boundedline, CircStat, 11 other tools, systems, mouse, 4 references
[3] doi:10.1038/s41593-026-02357-2 [code]
Experience reorganizes content-specific memory traces in macaques.
Journal: Nature neuroscience
In common: Chronux, boundedline, CircStat, 13 other tools, 3 references
[4] doi:10.1038/s41467-026-75347-4 [code]
Sleep reveals dynamics integrating and segregating movement and stimulus representations in V1.
Journal: Nature communications
In common: Chronux, boundedline, CircStat, 13 other tools, systems, mouse, 2 references
[5] doi:10.1038/s41467-026-73106-z [code]
Respiratory pauses highlight sleep architecture in mice.
Journal: Nature communications
In common: Chronux, boundedline, CircStat, 11 other tools, mouse, 2 references
[6] doi:10.64898/2026.03.12.710517 [code]
Cortical excitability inversely modulates fMRI connectivity via low-frequency neuronal coupling
Journal: bioRxiv (preprint)
In common: Chronux, shadedErrorBar, CircStat, 11 other tools, systems, 1 reference
[7] doi:10.1016/j.patter.2026.101590 [code]
Density-based longitudinal neuron tracking in high-density electrophysiological recordings.
Journal: Patterns (New York, N.Y.)
In common: shadedErrorBar, CircStat, Optimization Toolbox, 11 other tools, 2 references
[8] doi:10.1038/s41593-026-02388-9 [code]
Hippocampal CA3 connectomics reveals a gradient of mossy fiber inputs and selective feedforward inhibition onto pyramidal cells.
Journal: Nature neuroscience
In common: NetworkX, PyTorch, seaborn, 5 other tools, systems, mouse, 7 references
[9] doi:10.1126/sciadv.aeh7220 [code]
Central complex representations of self-movement are sufficient to compute wind direction in flight.
Journal: Science advances
In common: shadedErrorBar, CircStat, Curve Fitting Toolbox, 11 other tools
[10] doi:10.1038/s41467-026-76581-6 [code]
Thalamocortical bursts encode reward contingencies and drive associative learning.
Journal: Nature communications
In common: CircStat, Curve Fitting Toolbox, h5py, 8 other tools, systems, mouse, 3 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.