Skip to content

nigeLab_Block

Jump to bottom
Max-Home-Tower edited this page Jul 1, 2020 · 1 revision

Block

"Building-Block" for managing data from single recordings.

Methods

Constructor doMethods
Block Auto Cluster Spikes Behavior Sync Event Detection Event Header
LFP Decimation Raw Data Conversion Re-Reference Spike Detection
Unit Bandpass Filter Video Info Extraction Video Sync
Utility
Get (Completion) Status Link Data Set (Completion) Status Update Parameters

Constructor

nigeLab.Block

Creates "building-Block" for managing data from single recordings.

  • Associated defaults are stored in nigelObj.Pars.Block (at each hierarchical level)
    • Defaults can be configured in ~/+nigeLab/+defaults/Block.m
      • Critically, all name parsing is configured here. Incorrect parsing configuration is the most-common source of errors in constructing/initializing a Block.
      • There are substantial comments and examples in the code comments, look there for more information.
    • Other +defaults may influence the Block constructor:
      • ~/+nigeLab/+defaults/Animal.m (since Block is constructed during Animal constructor)
      • ~/+nigeLab/+defaults/Tank.m (since Block is constructed during Tank constructor)
      • ~/+nigeLab/+defaults/Video.m (since VidStreams flags can affect Block.init during automated parsing/association of Videos and Video-related Streams with the Block)
      • ~/+nigeLab/+defaults/Queue.m (depending on flags set for parallel or remote processing in combination with specifications of Matlab installation on your local machine)

Example Creation of Block Object

% Get second Block from third Animal of the Tank
animalIndex = 3;
blockIndex = 2;
blockObj = tankObj{animalIndex,blockIndex};
% Construct a Block using two UI prompts: 
% 1) Input block file
% 2) Output (Animal) folder that contains the Block
blockObj = nigeLab.Block();
blockObj = nigeLab.Block([]);
% Construct a Block, specifying save location, but select the file from UI
saveLoc = fullfile('savepath','experimentName','tankName','animalName');
blockObj = nigeLab.Block([],saveLoc);
% Construct a Block, skipping the UI
% Note that if a file path produces parsing errors,
% a UI may pop up to request the correct path.
inputFile = fullfile('inpath','tankName','animalName','blockName.rhd'); % Intan RHD
inputFile = fullfile('inpath','tankName','animalName','blockName.rhs'); % Intan RHS
inputFile = fullfile('inpath','tankName','animalName','blockName','blockName.sev'); % TDT
inputFile = fullfile('inpath','tankName','blockName','blockName.sev'); % TDT - also works
saveLoc = fullfile('savepath','experimentName','tankName','animalName');
blockObj = nigeLab.Block(inputFile,saveLoc);

doMethods

Block.doAutoClustering

Cluster (SPC, K-Means) spikes based on extracted waveform features (PCs, Wavelet Coefficients)

  • Overloads superclass method inherited from nigeLab.nigelObj
    • Superclass just iterates on "connected" nigeLab.Tank and nigeLab.Animal objects.
  • The following .Fields must have a true value for .Status:
    • Raw (doRawExtraction)
      • Required (indirectly)
    • Filt (doUnitFilter)
      • Required (indirectly)
    • CAR (doReReference)
      • Required (indirectly)
    • Spikes (doSD)
      • Required (indirectly)
    • SpikeFeatures (doSD)
      • Required directly
    • These requirements can be changed in ~/+nigeLab/+defaults/doActions.m
  • Associated defaults are in blockObj.Pars.AutoClustering
    • Defaults can be configured in ~/+nigeLab/+defaults/AutoClustering.m
    • For specific clustering algorithms, such as K-Means or SPC, some configurations are located in ``~/+nigeLab/+defaults/+AutoClustering/[methodname].m`

Example Auto-Clustering

blockObj = tankObj{1,2}; % second Block of first Animal in Tank
doAutoClustering(blockObj); % does auto-clustering and saves to diskfiles
linkToData(blockObj); % connect clusters to blockObj in Matlab
% linkToData(blockObj,'Clusters','Sorted'); % should also work, potentially faster

Block.doBehaviorSync

WIP: Synchronize neural data with behavioral events

Block.doEventDetection

WIP: Perform automated detection of events determined by .Videos or .Streams

Block.doEventHeaderExtraction

WIP: Extract Header for use with .scoreVideo method

Block.doLFPExtraction

Decimate raw signal to facilitate frequency-domain analyses

  • Overloads superclass method inherited from nigeLab.nigelObj
    • Superclass just iterates on "connected" nigeLab.Tank and nigeLab.Animal objects.
  • The following .Fields to have a true value for .Status:
    • Raw (doRawExtraction)
      • Required (directly)
    • These requirements can be changed in ~/+nigeLab/+defaults/doActions.m
  • Associated defaults are in blockObj.Pars.LFP
    • Defaults can be configured in ~/+nigeLab/+defaults/LFP.m

Example LFP Decimation

blockObj = tankObj{2,:}; % All Blocks from the second Animal of the Tank
doLFPExtraction(blockObj); % extracts LFP and saves to diskfiles
linkToData(blockObj); % connect clusters to blockObj in Matlab
% linkToData(blockObj,'LFP'); % should also work, potentially faster

Block.doRawExtraction

Extract "raw" recorded amplifier and digital stream signals from recording binaries

  • Overloads superclass method inherited from nigeLab.nigelObj
    • Superclass just iterates on "connected" nigeLab.Tank and nigeLab.Animal objects.
  • Does not require any other fields to be completed before running (by default)
    • This requirement can be changed in ~/+nigeLab/+defaults/doActions.m
  • Associated defaults are somewhat disperse; there are not any explicit +defaults elements corresponding to the Raw field per se; however, the main thing is that parsing parameters must be set up in several places. See Block Constructor for more details.
    • Defaults can be configured in ~/+nigeLab/+defaults/Raw.m

Example Raw Data Extraction

animalObj = tankObj{1}; % Returns 1st Animal in Tank
blockObj = animalObj{[2,5]}; % Returns 2nd and 5th block of 1st Animal in Tank
doRawExtraction(blockObj); % extracts Raw data and saves to diskfiles for both Blocks
linkToData(blockObj); % connect raw data to blockObj in Matlab
% linkToData(blockObj,'Raw','DigIO','AnalogIO','Stim','Time',); % should also work

Block.doReReference

Apply "virtual" common-average subtraction to each probe separately

  • Overloads superclass method inherited from nigeLab.nigelObj

    • Superclass just iterates on "connected" nigeLab.Tank and nigeLab.Animal objects.

  • The following .Fields must have a true value for .Status:

    • Raw (doRawExtraction)
      • Required (indirectly)
    • Filt (doUnitFilter)
      • Required (directly)
    • These requirements can be changed in ~/+nigeLab/+defaults/doActions.m

  • There are currently no associated +defaults for the .CAR field.

Example Re-Referencing

animalObj = tankObj{[1:3]}; % Returns first three Animals in Tank
blockObj = animalObj{:,[2,5]}; % Returns 2nd and 5th block of each Animal in animalObj array (first 3 animals of Tank)
doReReference(blockObj); % extracts Raw data and saves to diskfiles for both Blocks
linkToData(blockObj); % connect raw data to blockObj in Matlab
% linkToData(blockObj,'CAR',); % should also work

Block.doSD

Detect spikes in the extracellular field potentials and associate compressed sample features

  • Overloads superclass method inherited from nigeLab.nigelObj

    • Superclass just iterates on "connected" nigeLab.Tank and nigeLab.Animal objects.

  • The following .Fields must have a true value for .Status:

  • Associated defaults are in blockObj.Pars.SD

    • Note that these parameters are somewhat confusing due to being associated with a much older version of nigeLab, but will be ported to a less-confusing naming convention at some point.

Example Spike Detection

blockObj = tankObj{:,:}; % Return all blocks from the Tank 
linkToData(blockObj); % connect clusters to blockObj in Matlab
% linkToData(blockObj,'Spikes','SpikeFeatures'); % should also work, potentially faster

Block.doUnitFilter

Apply "spike-unit" bandpass filter to recorded extracellular field potentials

  • Overloads superclass method inherited from nigeLab.nigelObj
    • Superclass just iterates on "connected" nigeLab.Tank and nigeLab.Animal objects.
  • The following .Fields to have a true value for .Status:
    • Raw (doRawExtraction)
      • Required (directly)
    • These requirements can be changed in ~/+nigeLab/+defaults/doActions.m
  • Associated defaults are in blockObj.Pars.Filt
    • Defaults can be configured in ~/+nigeLab/+defaults/Filt.m

Example Spike Bandpass Filtering

blockObj = tankObj{:,2}; % Second block from all animals in Tank
doUnitFilter(blockObj); % extracts LFP and saves to diskfiles
linkToData(blockObj); % connect clusters to blockObj in Matlab
% linkToData(blockObj,'Filt'); % should also work, potentially faster

Block.doVidInfoExtraction

WIP: Get basic video-related information prior to performing manual video curation

Block.doVidSyncExtraction

WIP: Synchronize videos with neurophysiological data record

Curation

Block.scoreVideo

WIP: Curate and append metadata to individual "trials" or video frames

Utility

Block.getStatus

Return the completion status of a given .Field

  • superclass method inherited from nigeLab.nigelObj

Block.linkToData

Connect the data saved on the disk to the hierarchical structure

  • superclass method inherited from nigeLab.nigelObj

Block.updateStatus

Update the completion status of a specific .Field

  • superclass method inherited from nigeLab.nigelObj

Block.updateParams

Update some or all of the sub-fields of .Pars

  • superclass method inherited from nigeLab.nigelObj
Overview Getting Started User Interfaces Acknowledgements For Developers Updates

Home

Getting Started

User Interfaces

Acknowledgements

For developers

Updates

Clone this wiki locally