MatArray toolbox
What is it?
The MatArray toolbox is a set of Matlab tools designed for the analysis
of microarray data. Its main features are:
- Advanced normalization schemes.
- Very efficient implementation of hierarchical clustering,
with leaf ordering and export to TreeView.
- Efficient implementation of K-means clustering.
The toolbox is geared towards seasonned Matlab users, who want to treat
microarray data with efficient command-line algorithms.
For more information, you can take a look at the
start-up guide or at the
reference. Both can be downloaded
as indicated below.
The toolbox has been described in an
application note published in Bioinformatics, 19, 659-660 (there are
limitations imposed by Bioinformatics on what you can do with that paper).
License information
The toolbox is free for any use, commercial or otherwise, in its
original or in a modified form, as long as proper credit is given.
Installation
Note
For efficiency, some algorithms are coded in C. This means that
some files should be compiled using the mex function.
For commodity, a win32 version
with pre-compiled binaries is also available.
Downloads
And then...
For the toolbox, unzip in a directory and change your path
so that it encompasses that directory.
To compile the mex-files, move to the directory cfiles
in the
matarray directory and type MatArrayCompile.
Everything should go just
fine if the compiler is properly set up.
A C++ compiler is needed for the leaf ordering routine.
For the help files, unzip to a directory and open "matoc.html"
in you web browser. In order for the links to the Matlab
help files to work, the
files should be unzipped to the directory
MATLAB/help/toolbox/matarray.
Update
The toolbox and the help files have been updated (march 3rd 2003).
- Parts of the documentation have been improved/corrected.
- The
ClustTV routine has been slightly modified.
- The default in
normgen have been slightly changed.
Who is guilty for this?
The toolbox was written for the most part by
David Venet, except for the leaf ordering routine which is based
on a code from Ziv Bar-Joseph.
Bug report, suggestions, improvements,...
Please drop an e-mail to
David Venet.