MatClust

Spike sorting program for extracellular recordings. Version 1.4

Description

GUI program that runs in Matlab and allows the user to draw polygons or boxes around multidimensional data points. This 'clustering' of data is used often for multi-electrode recordings of neural data. The program has been tested on Windows, Linux, and Mac platforms.

Features:
-Manual 2D clustering with polygons
-Interface for custom semi-automated sorting code
-3D rotation of feature space
-Filtering of points to precisely define each polygon’s scope
-Easy to use interface (context menus, undo states, zooming and panning)

Requires a mex compiler (free) for setup. If you don't have one, see here for easy instructions:

http://www.mathworks.com/support/compilers/R2012b/win64.html

Also, please view readme.txt for setup instructions

Acknowledgements

Decimate Polygon inspired this file.

• MATLAB release: MATLAB 6.5 (R13)
• Other requirements: Mex compiler (http://www.mathworks.com/support/compilers/R2012b/win64.html)

[Rob Campbell]

Nice, thanks!

[Mattias Karlsson]

Hi John, matclust takes a matlab matrix as input, where each column is a spike feature and each row is one spike. Check out the readme.txt file for all input options. You'll need to convert the .ntt file into such a matrix before you can start. Luckily, neuralynx provides a function to import .ntt files to matlab (Nlx2MatSpike.m). It can be downloaded here: http://neuralynx.com/research_software/file_converters_and_utilities/

Then, you'll need to compute which features you want to use. The easiest feature to pull out is the peak amplitude of each spike. For tetrodes, you would want an n by 5 matrix, where the first column is time (in seconds), and the next four columns are the peak amplitudes of each spike.

[John Thompson]

Can your function handle Neurlynx (.ntt) data files?