Download the software here.
**** E-Predict v. 1.0 README, June 2005 ****
This version of E-Predict is an open-source implementation of
the algorithm described in
Urisman A, Fischer KF, Chiu CY, Kistler AL, Beck S, Wang D, DeRisi JL.
E-Predict: A Computational Strategy for Species Identification Based
on Observed DNA Microarray Hybridization Patterns.
Genome Biology, 2005.
The algorithm is implemented as a web application intended for
installation on a web server to be run remotely via a web browser.
The main program (epredict.pl) is a Perl CGI script, which requires
Perl 5.0 or later to run. The implementation uses three C++
executables (metrics, normidist, and vecnormViro) for number crunching
tasks requiring speed.
The software is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES
OR CONDITIONS OF ANY KIND (see license info below).
Running server such as Apache v. 1.3 or later,
Perl 5.0 or later with CGI.pm
1. Uncompress E-Predict_v1.0.tar.gz
2. Make the uncompressed E-Predict_v1.0 directory and subdirectories
readable and executable by the server. An easy way to do this
is to place the entire E-Predict_v1.0 directory in the
cgi-bin directory already used by your server.
3. Edit paths to the E-Predict and cgi-bin directories ($root and
$cgi_bin variables) in epredict.pl
4. Edit path to the csnips directory ($snips variable) in Csnips.pm
5. This should do it on most UNIX/LINUX machines. To run the
application, simply point your web browser to epredict.pl.
For example, the URL might look like this --
6. Executables of the C++ components (metrics, normdist, and
vecnormViro) are included and should work without recompiling
on most UNIX/LINUX machines. If needed, these components
can be recompiled using the makefile in the src/ subdirectory.
The makefile will automatically move the executables into
the E-Predict_v1.0/cgi-bin subdirectory. Note - a C++ compiler
like G++ is required.
Run epredict.pl by pointing your web browser to the script's URL.
A settings page will be displayed. Select a desired dataset from
the four provided examples (described in Urisman et. al, 2005).
Press 'Submit' to run the algorithm.
E-Predict v. 1.0 is licensed under the Apache License, Version 2.0.
A copy of the license is included in this distribution or can be
viewed at http://www.apache.org/licenses/LICENSE-2.0