update

DataJoint/+ANLI/FPSTHMatrix.m

@@ -0,0 +1,66 @@
+%{
+#
+-> EXP2.Session
+---
+psth_roi_tr_fr            : longblob        # psth matrix (ROIs x Trials  X Frames)
+typical_psth_timestamps   : longblob        # timestamps of each frame relative to go cue, typical refers to the case in which trial times would differ across sessions (e.g. when delay duration is different for different trials). In this this timestamps are only representative for the session, but may differ for indvidual trials.
+typical_time_sample_start : double          # time of sample start relative to go cue
+typical_time_sample_end   : double          # time of sample end relative to go cue
+%}
+
+
+classdef FPSTHMatrix < dj.Computed
+        properties
+        keySource = EXP2.Session  & IMG.ROI;
+    end
+    methods(Access=protected)
+        function makeTuples(self, key)
+            roi_list=fetchn(IMG.ROI & key,'roi_number','ORDER BY roi_number');
+
+
+            bEvent_all=fetch(  EXP2.BehaviorTrialEvent*EXP2.BehaviorTrial & key & 'early_lick="no early"','*','ORDER BY trial');
+
+            trial_event_time_all = [bEvent_all.trial_event_time];
+            typical_time_go = mode(trial_event_time_all (contains({bEvent_all.trial_event_type},'go')));
+            typical_time_sample_start =mode(trial_event_time_all (contains({bEvent_all.trial_event_type},'sound sample start')))-typical_time_go;
+            typical_time_sample_end = mode(trial_event_time_all (contains({bEvent_all.trial_event_type},'sound sample end')))-typical_time_go;
+
+%          
+            k=key;
+            k.roi_number=1;
+            frames_per_trial=fetchn(IMG.ROITrial & k,'frames_per_trial','ORDER BY trial');
+            typical_num_frames_per_trial=mode(frames_per_trial);
+            f_trace_timestamps=fetchn(IMG.ROITrial & k,'f_trace_timestamps','ORDER BY trial');
+            [~,longest_trial_idx] = max(frames_per_trial);
+            typical_psth_timestamps = f_trace_timestamps{longest_trial_idx}(1:typical_num_frames_per_trial);
+
+%             frames_per_trial = frames_per_trial(b_trials);
+%             f_trace_timestamps = f_trace_timestamps(b_trials);
+
+            numberTrials = numel(frames_per_trial);
+
+            FPSTH_roi_tr_fr = zeros(numel(roi_list),numberTrials,max(frames_per_trial)) + NaN;
+            for iROI = 1:numel(roi_list)
+
+                k.roi_number=roi_list(iROI);
+                baseline_fl_trials=cell2mat(fetchn(IMG.ROI & k,'baseline_fl_trials'));
+                smooth_b=smooth(baseline_fl_trials,10,'rlowess',1);
+
+                ROITrial=fetchn(IMG.ROITrial & k,'f_trace','ORDER BY trial');
+
+                for iTr=1:1:numberTrials
+                    f_trial=(ROITrial{iTr} - smooth_b(iTr))/smooth_b(iTr); %deltaF/F
+                    FPSTH_roi_tr_fr(iROI , iTr, 1:frames_per_trial(iTr))=f_trial;
+                end
+
+            end
+
+            key.psth_roi_tr_fr = FPSTH_roi_tr_fr;
+            key.typical_psth_timestamps =  typical_psth_timestamps -typical_time_go;
+            key.typical_time_sample_start = typical_time_sample_start;
+            key.typical_time_sample_end = typical_time_sample_end;
+            insert(ANLI.FPSTHMatrix, key);
+
+        end
+    end
+end

DataJoint/+ANLI/FPSTHaverage.m

@@ -0,0 +1,117 @@
+%{
+#
+-> IMG.ROI
+-> EXP2.TrialNameType
+-> EXP2.Outcome
+---
+num_trials_averaged             : int         # number of trials averaged per condition, for this ROI
+psth_avg                        : longblob    # trial-type averaged PSTH aligned to go cue time, expressed as deltaF/F, where F is the baseline flourescene in the beginning of each trial
+psth_stem                       : longblob    # standard error of the mean of the above, expressed as deltaF/F
+psth_timestamps                 : longblob    # timestamps of each frame relative to go cue
+time_sample_start               : double      # time of sample start relative to go cue
+time_sample_end                 : double      # time of sample end relative to go cue
+%}
+
+
+classdef FPSTHaverage < dj.Computed
+    properties
+        keySource = (IMG.FOV  & IMG.ROI) * (EXP2.TrialNameType)  * EXP2.Outcome;
+    end
+    methods(Access=protected)
+        function makeTuples(self, key)
+            roi_list=fetchn(IMG.ROI & key,'roi_number');
+
+            b=fetch((EXP2.BehaviorTrial*EXP2.TrialName) & key,'*','ORDER BY trial');
+            if isempty(b)
+                return;
+            end
+            b=struct2table(b); b_trials=[b.trial];
+            b_trials=b_trials(contains(b.early_lick,'no early'));
+            if isempty(b_trials)
+                return;
+            end
+
+
+            % for all conditions and trial-types
+            kk=key;
+            kk=rmfield(kk,'trial_type_name');
+            kk=rmfield(kk,'outcome');
+
+            bEvent_all=fetch(  EXP2.BehaviorTrialEvent*EXP2.BehaviorTrial & kk & 'early_lick="no early"','*','ORDER BY trial');
+
+            trial_event_time_all = [bEvent_all.trial_event_time];
+            typical_time_go = mode(trial_event_time_all (contains({bEvent_all.trial_event_type},'go')));
+            typical_time_sample_start =mode(trial_event_time_all (contains({bEvent_all.trial_event_type},'sound sample start')))-typical_time_go;
+            typical_time_sample_end = mode(trial_event_time_all (contains({bEvent_all.trial_event_type},'sound sample end')))-typical_time_go;
+
+
+            % for this conditions and trial-type combination
+            bEvent=fetch(  EXP2.BehaviorTrialEvent&((EXP2.BehaviorTrial*EXP2.TrialName)& key),'*','ORDER BY trial');
+            trial_event_time = [bEvent.trial_event_time];
+            time_go = trial_event_time (contains({bEvent.trial_event_type},'go'));
+            time_sample_start =trial_event_time (contains({bEvent.trial_event_type},'sound sample start'));
+            time_sample_end = trial_event_time (contains({bEvent.trial_event_type},'sound sample end'));
+
+
+            key.roi_number=1;
+            frames_per_trial=fetchn(IMG.ROITrial & key,'frames_per_trial','ORDER BY trial');
+            typical_num_frames_per_trial=mode(frames_per_trial);
+            f_trace_timestamps=fetchn(IMG.ROITrial & key,'f_trace_timestamps','ORDER BY trial');
+            [~,longest_trial_idx] = max(frames_per_trial);
+            typical_psth_timestamps = f_trace_timestamps{longest_trial_idx}(1:typical_num_frames_per_trial);
+
+            frames_per_trial = frames_per_trial(b_trials);
+            f_trace_timestamps = f_trace_timestamps(b_trials);
+
+            numberTrials = numel(frames_per_trial);
+
+            for iROI = 1:numel(roi_list)
+                F=zeros(numberTrials,max(frames_per_trial)) + NaN;
+
+                key.roi_number=roi_list(iROI);
+                baseline_fl_trials=cell2mat(fetchn(IMG.ROI & key,'baseline_fl_trials'));
+                %                 smooth_b=smooth(baseline_fl_trials,10,'sgolay',1);
+                smooth_b=smooth(baseline_fl_trials,10,'rlowess',1);
+
+                smooth_b=smooth_b(b_trials)';
+
+                ROITrial=fetchn(IMG.ROITrial & key,'f_trace','ORDER BY trial');
+                ROITrial = ROITrial (b_trials);
+
+                key_FPSTHtrial = key;
+                key_FPSTHtrial = repmat(key_FPSTHtrial,1,numberTrials);
+
+                for iTr=1:1:numberTrials
+                    f_trial=(ROITrial{iTr} - smooth_b(iTr))/smooth_b(iTr); %deltaF/F
+                    F(iTr,1:frames_per_trial(iTr))=f_trial;
+
+                    key_FPSTHtrial(iTr).trial=b_trials(iTr);
+                    key_FPSTHtrial(iTr).psth_trial = f_trial;
+                    this_trial_time_go = time_go (iTr);
+                    key_FPSTHtrial(iTr).psth_timestamps = f_trace_timestamps{iTr} - this_trial_time_go;
+                    key_FPSTHtrial(iTr).time_sample_start   = time_sample_start(iTr) - this_trial_time_go;
+                    key_FPSTHtrial(iTr).time_sample_end = time_sample_end(iTr)- this_trial_time_go;
+
+                end
+
+
+                F=F(:,1:typical_num_frames_per_trial);
+                psth_avg = nanmean(F,1);
+                psth_stem = nanstd(F,1)/sqrt(numberTrials);
+
+                key_FPSTHaverage=key;
+                key_FPSTHaverage.num_trials_averaged=numberTrials;
+                key_FPSTHaverage.psth_avg=psth_avg;
+                key_FPSTHaverage.psth_stem=psth_stem;
+                key_FPSTHaverage.psth_timestamps= typical_psth_timestamps -typical_time_go;
+                key_FPSTHaverage.time_sample_start = typical_time_sample_start;
+                key_FPSTHaverage.time_sample_end = typical_time_sample_end;
+
+                insert(ANLI.FPSTHaverage, key_FPSTHaverage)
+                insert(ANLI.FPSTHtrial, key_FPSTHtrial)
+
+
+            end
+        end
+    end
+end

DataJoint/+ANLI/FPSTHaverageClusterbased.m

@@ -0,0 +1,68 @@
+%{
+#
+-> IMG.ROI
+-> EXP2.EpochName
+-> EXP2.TrialNameType
+trial_cluster_group             : int
+---
+num_trials_averaged             : int         # number of trials averaged per condition, for this ROI
+psth_avg                        : longblob    # trial-type averaged PSTH aligned to go cue time, expressed as deltaF/F, where F is the baseline flourescene in the beginning of each trial
+psth_stem                       : longblob    # standard error of the mean of the above, expressed as deltaF/F
+psth_timestamps                 : longblob    # timestamps of each frame relative to go cue
+time_sample_start               : double      # time of sample start relative to go cue
+time_sample_end                 : double      # time of sample end relative to go cue
+%}
+
+
+classdef FPSTHaverageClusterbased < dj.Computed
+    properties
+        keySource = (IMG.FOV  & IMG.ROI) * (EXP2.TrialNameType & 'task="sound"') * EXP2.EpochName;
+    end
+    methods(Access=protected)
+        function makeTuples(self, key)
+            trial_cluster_group=[1,2];
+
+            roi_list=fetchn(IMG.ROI & key,'roi_number');
+
+            key.time_sample_start =fetch1(ANLI.FPSTHMatrix & key,'typical_time_sample_start');
+            key.time_sample_end =fetch1(ANLI.FPSTHMatrix & key,'typical_time_sample_end');
+            psth_timestamps =fetch1(ANLI.FPSTHMatrix & key,'typical_psth_timestamps');
+
+
+            for i_g=1:1:numel(trial_cluster_group)
+                key_group.trial_cluster_group = trial_cluster_group(i_g);
+                rel=(ANLI.TrialCluster*EXP2.TrialName & key & key_group);
+                trials_group = fetchn(rel,'trial','ORDER BY trial');
+                numberTrials = numel(trials_group);
+                frames_per_trial=fetchn(IMG.ROITrial & key & 'roi_number=1' & rel,'frames_per_trial','ORDER BY trial');
+
+                for iROI = 1:numel(roi_list)
+                    key_FPSTHaverage=[];
+                    F=zeros(numberTrials,max([frames_per_trial;numel(psth_timestamps)])) + NaN;
+                    key_roi.roi_number=roi_list(iROI);
+
+                    FPSTHtrial=fetchn((ANLI.FPSTHtrial & rel) & key_roi,'psth_trial','ORDER BY trial');
+
+                    for iTr=1:1:numberTrials
+                        F(iTr,1:frames_per_trial(iTr))=FPSTHtrial{iTr};
+                    end
+
+                    F=F(:,1:numel(psth_timestamps));
+                    psth_avg = nanmean(F,1);
+                    psth_stem = nanstd(F,1)/sqrt(numberTrials);
+
+                    key_FPSTHaverage=key;
+                    key_FPSTHaverage.roi_number = roi_list(iROI);
+                    key_FPSTHaverage.trial_cluster_group =  trial_cluster_group(i_g);
+                    key_FPSTHaverage.psth_timestamps=psth_timestamps;
+                    key_FPSTHaverage.num_trials_averaged=numberTrials;
+                    key_FPSTHaverage.psth_avg=psth_avg;
+                    key_FPSTHaverage.psth_stem=psth_stem;
+
+                    insert(self, key_FPSTHaverage)
+
+                end
+            end
+        end
+    end
+end

DataJoint/+ANLI/FPSTHtrial.m

@@ -0,0 +1,21 @@
+%{
+#
+-> IMG.ROI
+-> EXP2.TrialNameType
+-> EXP2.Outcome
+-> EXP2.SessionTrial
+---
+psth_trial                      : longblob    # PSTH aligned to go cue time, expressed as deltaF/F, where F is the baseline flourescene in the beginning of each trial
+psth_timestamps                 : longblob    # timestamps of each frame relative to go cue
+time_sample_start               : double      # time of sample start relative to go cue
+time_sample_end               : double      # time of sample end relative to go cue
+%}
+
+
+classdef FPSTHtrial < dj.Computed
+    methods(Access=protected)
+        function makeTuples(self, key)
+
+        end
+    end
+end