Skip to content

Commit

Permalink
Merge pull request #66 from bahanonu/bahanonu/functionUpdates
Browse files Browse the repository at this point in the history 
New load/save cell-extraction outputs, EXTRACT support, and misc. updates
  • Loading branch information
@bahanonu
bahanonu authored Mar 21, 2021
2 parents afd1405 + 298f223 commit 03d2cf1
Show file tree
Hide file tree
Showing 31 changed files with 1,092 additions and 128 deletions.
17 changes: 14 additions & 3 deletions +ciapkg/+io/loadDependencies.m
View file
Original file line number Diff line number Diff line change
Expand Up @@ -15,18 +15,19 @@ function loadDependencies(varargin)
% 2021.01.22 [13:42:36] - NWB from specific release to reduce compatibility errors.
% 2021.02.01 [15:10:41] - Separated into non-class function for use in more functions without needing to load CIAtah class.
% 2021.02.01 [‏‎15:19:40] - Update `_external_programs` to call ciapkg.getDirExternalPrograms() to standardize call across all functions.
% 2021.03.20 [18:12:20] - Added EXTRACT support to list of functions to download.
% TODO
% Verify all dependencies download and if not ask user to download again.

%========================
% DESCRIPTION
options.externalProgramsDir = ciapkg.getDirExternalPrograms();
options.guiEnabled = 1;
options.dependencyStr = {'downloadMiji','downloadCnmfGithubRepositories','example_downloadTestData','loadMiji','downloadNeuroDataWithoutBorders'};
options.dependencyStr = {'downloadMiji','downloadCnmfGithubRepositories','example_downloadTestData','loadMiji','downloadNeuroDataWithoutBorders','downloadEXTRACT'};

options.dispStr = {'Download Fiji (to run Miji)','Download CNMF, CNMF-E, and CVX code.','Download test one- and two photon datasets.','Load Fiji/Miji into MATLAB path.','Download NWB (NeuroDataWithoutBorders)'};
options.dispStr = {'Download Fiji (to run Miji)','Download CNMF, CNMF-E, and CVX code.','Download test one- and two photon datasets.','Load Fiji/Miji into MATLAB path.','Download NWB (NeuroDataWithoutBorders)','Download EXTRACT'};
% Int vector: index of options.dependencyStr to run by default with no GUI
options.depIdxArray = [1 2 3 5];
options.depIdxArray = [1 2 3 5 6];
% Binary: 1 = force update even if already downloaded. 0 = skip if already downloaded
options.forceUpdate = 0;
% get options
Expand Down Expand Up @@ -95,6 +96,16 @@ function loadDependencies(varargin)
optionsH.outputDir = optionsH.gitNameDisp;
optionsH.gitName = cellfun(@(x) [x '-master'],optionsH.gitNameDisp,'UniformOutput',false);
[success] = downloadGithubRepositories('options',optionsH);
case 'downloadEXTRACT'
optionsH.forceUpdate = forceUpdate;
optionsH.signalExtractionDir = options.externalProgramsDir;
optionsH.gitNameDisp = {'extract'};
optionsH.gitRepos = {'https://github.com/schnitzer-lab/EXTRACT-public'};
optionsH.gitRepos = cellfun(@(x) [x '/archive/master.zip'],optionsH.gitRepos,'UniformOutput',false);
optionsH.outputDir = optionsH.gitNameDisp;
% optionsH.gitName = cellfun(@(x) [x '-master'],optionsH.gitNameDisp,'UniformOutput',false);
optionsH.gitName = {'EXTRACT-public-master'};
[success] = downloadGithubRepositories('options',optionsH);
case 'downloadNeuroDataWithoutBorders'
optionsH.forceUpdate = forceUpdate;
optionsH.signalExtractionDir = options.externalProgramsDir;
Expand Down
288 changes: 288 additions & 0 deletions +ciapkg/+io/loadSignalExtraction.m
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,288 @@
function [inputImages,inputSignals,infoStruct,algorithmStr,inputSignals2] = loadSignalExtraction(inputFilePath,varargin)
% Loads CIAtah-style MAT or NWB files containing signal extraction results.
% Biafra Ahanonu
% started: 2021.02.03 [10:53:11]
% inputs
% inputFilePath - path to signal extraction output
% outputs
% inputImages - 3D or 4D matrix containing cells and their spatial information, format: [x y nCells].
% inputSignals - 2D matrix containing activity traces in [nCells nFrames] format.
% infoStruct - contains information about the file, e.g. the 'description' property that can contain information about the algorithm.
% algorithmStr - String of the algorithm name.
% inputSignals2 - same as inputSignals but for secondary traces an algorithm outputs.

% changelog
% 2021.03.10 [18:50:48] - Updated to add support for initial set of cell-extraction algorithms.
% TODO
%

% ========================
% Struct: name of structures for CIAtah-style outputs.
options.extractionMethodStructVarname = struct(...
'PCAICA', 'pcaicaAnalysisOutput',...
'EM', 'emAnalysisOutput',...
'CELLMax','cellmaxAnalysisOutput',...
'EXTRACT', 'extractAnalysisOutput',...
'CNMF', 'cnmfAnalysisOutput',...
'CNMFE', 'cnmfeAnalysisOutput',...
'ROI', 'roiAnalysisOutput'...
);

options.extractionMethodStructSaveStr = struct(...
'pcaicaAnalysis','PCAICA',...
'emAnalysis','EM',...
'cellmaxAnalysis','CELLMax',...
'extractAnalysis','EXTRACT',...
'cnmfAnalysis','CNMF',...
'cnmfeAnalysis','CNMFE',...
'roiAnalysis','ROI'...
);
% Name of H5 group for images and signal series in NWB files
options.nwbGroupImages = '/processing/ophys/ImageSegmentation/PlaneSegmentation';
options.nwbGroupSignalSeries = '/processing/ophys/Fluorescence/RoiResponseSeries';
% Int: indicates which trace output to use from 1 = primary trace, 2 = secondary trace, etc.
options.signalExtractionTraceOutputType = 1;
% get options
options = getOptions(options,varargin);
% display(options)
% unpack options into current workspace
% fn=fieldnames(options);
% for i=1:length(fn)
% eval([fn{i} '=options.' fn{i} ';']);
% end
% ========================

try
inputImages = [];
inputSignals = [];
inputSignals2 = [];
algorithmStr = '';
infoStruct = struct;

if iscell(inputFilePath)
inputFilePath = inputFilePath{1};
end
% Return if not a string, return as not appropriate input
if ~ischar(inputFilePath)
return;
end

[~,~,inputEXT] = fileparts(inputFilePath);
switch inputEXT
case '.nwb'
nwbOpts.groupImages = options.nwbGroupImages;
nwbOpts.groupSignalSeries = options.nwbGroupSignalSeries;
[inputImages,inputSignals,infoStruct,algorithmStr] = loadNeurodataWithoutBorders(inputFilePath,'options',nwbOpts);
return;
case '.mat'
% Do nothing, go to the next steps below.
otherwise
% Do nothing
end

% Get list of variables in the MAT file, use this to determine which algorithm to load.
variableInfo = who('-file', inputFilePath);

algList = fieldnames(options.extractionMethodStructVarname);
algorithm = '';
for i = 1:length(algList)
algName = algList{i};
algMatch = options.extractionMethodStructVarname.(algName);
matchAlg = any(strcmp(variableInfo,algMatch));
if matchAlg==1
algorithm = algName;
end
end
algorithmStr = algorithm;

switch algorithm
case 'ROI'
[inputImages,inputSignals,infoStruct] = subfxnROI(inputFilePath,options);
case 'PCAICA'
[inputImages,inputSignals,infoStruct] = subfxnPCAICA(inputFilePath,options);
case {'EM','CELLMax'}
[inputImages,inputSignals,infoStruct,inputSignals2] = subfxnCELLMax(inputFilePath,options);
case 'EXTRACT'
[inputImages,inputSignals,infoStruct] = subfxnEXTRACT(inputFilePath,options);
case 'CNMF'
[inputImages,inputSignals,infoStruct,inputSignals2] = subfxnCNMF(inputFilePath,options);
case 'CNMFE'
[inputImages,inputSignals,infoStruct,inputSignals2] = subfxnCNMFE(inputFilePath,options);
otherwise
disp('Input MAT file is not recognized or not in CIAtah-style format.')
end

catch err
disp(repmat('@',1,7))
disp(getReport(err,'extended','hyperlinks','on'));
disp(repmat('@',1,7))
end
end
function [inputImages,inputSignals,infoStruct] = subfxnPCAICA(inputFilePath,options)
infoStruct = struct;
load(inputFilePath);

if exist('pcaicaAnalysisOutput','var')
inputSignals = double(pcaicaAnalysisOutput.IcaTraces);
if strcmp(pcaicaAnalysisOutput.imageSaveDimOrder,'xyz')
% inputImages = permute(double(pcaicaAnalysisOutput.IcaFilters),[3 1 2])
inputImages = double(pcaicaAnalysisOutput.IcaFilters);
elseif strcmp(pcaicaAnalysisOutput.imageSaveDimOrder,'zxy')
inputImages = permute(double(pcaicaAnalysisOutput.IcaFilters),[2 3 1]);
% inputImages = pcaicaAnalysisOutput.IcaFilters;
else
% inputImages = permute(double(pcaicaAnalysisOutput.IcaFilters));
inputImages = pcaicaAnalysisOutput.IcaFilters;
end
end
if exist('IcaTraces','var')
inputSignals = IcaTraces;
end
if exist('IcaFilters','var')
% inputImages = IcaFilters;
inputImages = permute(IcaFilters,[2 3 1]);
end
end
function [inputImages,inputSignals,infoStruct,inputSignals2] = subfxnCELLMax(inputFilePath,options)
infoStruct = struct;
load(inputFilePath);

% if exist(options.structEMVarname,'var')
if exist('emAnalysisOutput','var')|exist('cellmaxAnalysisOutput','var')
if exist('cellmaxAnalysisOutput','var')
emAnalysisOutput = cellmaxAnalysisOutput;
end
inputImages = emAnalysisOutput.cellImages;

emOutputType = {'scaledProbability','dsScaledProbability','cellTraces','dsCellTraces'};
for emI = 1:length(emOutputType)
if isfield(emAnalysisOutput,emOutputType{emI})
disp(['Using ' emOutputType{emI} '...'])
inputSignals = double(emAnalysisOutput.(emOutputType{emI}));
break;
end
end

if 0
if isfield(emAnalysisOutput,'scaledProbability')
disp('Using scaledProbability...')
inputSignals = double(emAnalysisOutput.scaledProbability);
elseif isfield(emAnalysisOutput,'dsScaledProbability')
disp('Using dsScaledProbability...')
inputSignals = double(emAnalysisOutput.dsScaledProbability);
elseif isfield(emAnalysisOutput,'cellTraces')
disp('Using cellTraces...')
inputSignals = double(emAnalysisOutput.cellTraces);
elseif isfield(emAnalysisOutput,'dsCellTraces')
disp('Using dsCellTraces...')
if length(emAnalysisOutput.dsCellTraces)==1
inputSignals = emAnalysisOutput.cellTraces;
else
inputSignals = emAnalysisOutput.dsCellTraces;
end
else
inputSignals = emAnalysisOutput.cellTraces;
end
end

inputSignals2 = double(emAnalysisOutput.cellTraces);

switch options.signalExtractionTraceOutputType
case 1
%
case 2
inputSignals2Tmp = inputSignals2;
inputSignals2 = inputSignals;
inputSignals = inputSignals2Tmp;
otherwise
%
end

% Convert to dF/F values
inputSignals = normalizeSignalExtractionActivityTraces(inputSignals,inputImages);
inputSignals2 = normalizeSignalExtractionActivityTraces(inputSignals2,inputImages);

end
end
function [inputImages,inputSignals,infoStruct] = subfxnEXTRACT(inputFilePath,options)
infoStruct = struct;
load(inputFilePath);

% if exist(options.structEXTRACTVarname,'var')
if exist('extractAnalysisOutput','var')
% inputImages = double(permute(extractAnalysisOutput.filters,[3 1 2]));
inputImages = double(extractAnalysisOutput.filters);
% inputSignals = double(permute(extractAnalysisOutput.traces, [2 1]));
inputSignals = double(extractAnalysisOutput.traces);
end
end
function [inputImages,inputSignals,infoStruct,inputSignals2] = subfxnCNMF(inputFilePath,options)
infoStruct = struct;
load(inputFilePath);

% if exist(options.structCNMRVarname,'var')
if exist('cnmfAnalysisOutput','var')
% inputImages = double(permute(cnmfAnalysisOutput.extractedImages,[3 1 2]));
inputImages = double(cnmfAnalysisOutput.extractedImages);
% inputSignals = double(permute(extractAnalysisOutput.traces, [2 1]));

switch options.signalExtractionTraceOutputType
case 1
inputSignals = double(cnmfAnalysisOutput.extractedSignals);
inputSignals2 = double(cnmfAnalysisOutput.extractedSignalsEst);
case 2
inputSignals = double(cnmfAnalysisOutput.extractedSignalsEst);
inputSignals2 = double(cnmfAnalysisOutput.extractedSignals);
otherwise
inputSignals = double(cnmfAnalysisOutput.extractedSignalsEst);
inputSignals2 = double(cnmfAnalysisOutput.extractedSignals);
end

% Convert units to relative dF/F for later analysis
inputSignals = normalizeSignalExtractionActivityTraces(inputSignals,inputImages);
inputSignals2 = normalizeSignalExtractionActivityTraces(inputSignals2,inputImages);
end
end
function [inputImages,inputSignals,infoStruct,inputSignals2] = subfxnCNMFE(inputFilePath,options)
infoStruct = struct;
load(inputFilePath);

if exist('cnmfeAnalysisOutput','var')
inputImages = double(cnmfeAnalysisOutput.extractedImages);

switch options.signalExtractionTraceOutputType
case 1
inputSignals = double(cnmfeAnalysisOutput.extractedSignals);
inputSignals2 = double(cnmfeAnalysisOutput.extractedSignalsEst);
case 2
inputSignals = double(cnmfeAnalysisOutput.extractedSignalsEst);
inputSignals2 = double(cnmfeAnalysisOutput.extractedSignals);
otherwise
inputSignals = double(cnmfeAnalysisOutput.extractedSignalsEst);
inputSignals2 = double(cnmfeAnalysisOutput.extractedSignals);
end

% Convert units to relative dF/F for later analysis
inputSignals = normalizeSignalExtractionActivityTraces(inputSignals,inputImages);
inputSignals2 = normalizeSignalExtractionActivityTraces(inputSignals2,inputImages);
end
end
function [inputImages,inputSignals,infoStruct] = subfxnROI(inputFilePath,options)
infoStruct = struct;
load(inputFilePath);

if exist('roiAnalysisOutput','var')
% inputImages = double(permute(extractAnalysisOutput.filters,[3 1 2]));
inputImages = double(roiAnalysisOutput.filters);
% inputSignals = double(permute(extractAnalysisOutput.traces, [2 1]));
if isfield(roiAnalysisOutput,'tracesDemix')
display('Using demixed ROI traces!')
inputSignals = double(roiAnalysisOutput.tracesDemix);
else
inputSignals = double(roiAnalysisOutput.traces);
end
end
if exist('ROItraces','var')
inputSignals = ROItraces;
end
end
Loading

0 comments on commit 03d2cf1

Please sign in to comment.