******************************************************************** S Q U E A K Y S P K ******************************************************************** Name: SqueakySpk Authors: Jon Newman <jnewman6 at gatech dot edu> Riley Zeller-Townson ******************************************************************** NOTE: This class is a work in progress. It is licensed under GPLv3. If you wish to contribute to the source, report a bug etc, please request write permission to our google-code repository at http://code.google.com/p/squeakyspk/. DESCRIPTION: Squeaky Spk is a MATLAB(R) class that is useful for artifact removal, spike-sorting, basic analysis, and storage of extracellular neural recordings. It is useful for handling data of the form T, C, and W where T is an NX1 vector of spike times, C is an NX1 vector of channel indicies that correspond to the spike times in T and W is an NXM matrix of spike waveforms ('snips') that correspond to the spike times in T. ******************************************************************** INSTALLATION: 1. Place the SqueakySpk directory somewhere on your hard-drive. For PC users I recommend: C:Users/<your account>/MATLAB 2. Open Matlab. Add SqueakySpk to your matlab path using File > Set Path 3. Click 'Add with Subfolders' and navigate to the SqueakySpk main directory. Hit OK and then click save. ******************************************************************** USAGE: SqueakySpk is a MATLAB(R) class. This means it has properties (which which will store your data) along with methods that can act on that data. It is important to recognize that SqueakySpk never deletes raw data, it only modifies the way in which this data is accessed. For example, if I had my data structure spkdat with fields 'time', 'channel' and 'waveform' as mentioned above, a typical workflow for using Squeaky spikes would be as follows: % Object ID fid = '20100604_16269_SS'; This example uses the loadspike method that works with the output of John Rolston's NeuroRighter recording and stimulation system. SqueakySpk can be used with any recording system so long as data is entered as a struct with 'time', 'channel', and 'waveform' data fields. In addition to storage of spiking data, there are overload inputs that can store stimulus data by inputing a stimulus struct with fields 'time' and 'channel' where 'time' is a WX1 vector of stimulus times and 'channel' is a WX1 vector of corresponding stimulus channels. % Load data into data structs spkdat = loadspike('20100604_16269_spont.spk'); stimdat = loadstim('20100604_16269_stim.stim'); Here spkdat is a struct with fields spkdat.time, spkdat.channel, and spkdat.waveform as mentioned above. % Instantiate a SqueakySpk Object SS = SqueakySpk('20100604_16269_test',25000,1/1200,spkdat,stimdat); % Remove meaningless channels (default is for MCS arrays) SS.RemoveChannel(); % Remove spikes with blanks due to artifact suppression SS.RemoveSpkWithBlank(); % Perform a hard p2p cut at 175 uV SS.HardThreshold(175); % Clustering SS.WaveClus(3,20,'wav',1); SS.RemoveUnit(0); % remove unsorted data % Weed units by average waveforms SS.WeedUnitByWaveform(); % Examime some data to make sure results of sorting and cleaning look good SS.RasterWave([200 210],'both'); % Save the SS data object SS.Save %% Clear everything and just load the SS object you just saved clear all load('20100604_16269_test.SS','-mat') % File ID's SSfid = '20100622_poisstim_wdeletes30_100_16269_SS'; spkfid = '20100622_poisstim_wdeletes30_100_16269.spk'; % Load the data spkdat = loadspike(spkfid); For help on the methods available once you have created a SqueakySpk object, at the command line, enter the name of your SS object: >> SSobj This will provide information on the properties and methods avaiable to you. Next type >> help <method you are interested in> at the command line to get further help on a given method. For an example of SqueakySpk in action, run the example script that is provided in the main SqueakySpk directory, SSTestScript.m ********************************************************************