GenExplore is a Microarray data analysis tool which can help you easily find out relationships among different genes. Results can be validated by experts or domain information, or suggest new experiments.

Related Link: http://www.cs.uncc.edu/~xwu/bio/GenExplore.html

2. Installation

2.0 Note

This instruction is currently for redhat linux platform only, RH linux 8.0 or higher version is recommanded. Moreover, you must use "administrator" or "root" as your login id, because during installation u need rights to write to some directory such as "/usr".

2.1 FLTK

-->Get FLTK source code from ftp://ftp2.easysw.com/pub/fltk/1.1.3/fltk-1.1.3-source.zip
-->Note: You can also download 1.1.4 or latest version and install it
-->use "unzip fltk-1.1.3-source.zip" command under your own directory(e.g. "/usr/app/")
-->after unzip, type "cd fltk-1.1.3" to enter fltk directory.
-->type "./configure"
-->type "./make"
-->type "./make install"

2.2 Graphviz

-->get Graphviz from http://www.graphviz.org/pub/graphviz/ARCHIVE/graphviz-1.10.tar.gz
-->Note: You can also download latest version and install it
-->type "gzip -d graphviz-1.10.tar.gz" command
-->type "tar -xf graphviz-1.10.tar" command
-->type "cd graphviz-1.10" to enter its directory
-->type "./configure"
-->type "./make"
-->type "./make install"

2.3 GenExplore

-->get GenExplore package
-->type"gunzip GenExplore0.98b.tar.gz"

-->type "tar -xvf GenExplore0.98b.tar"
-->type "cd GenExplore0.98b"
-->type "./install.sh"
-->type "ln -b /usr/app/graphviz-1.10/dotneato/neato neato", here we assume "/usr/app/graphviz-1.10/" is where your graphviz was installed, you can replace that by using your own path. If you have already ran "make install" command in Graphviz directory during 2.2, you do not need to do this step.

2.4 apriori

-->an excutable "apriori" program is included in GenExplore0.98b root directory. So no further step needed.
-->you can also download the linux excutable file in http://fuzzy.cs.uni-magdeburg.de/~borgelt/src/apriori
-->you can also download the source file in http://fuzzy.cs.uni-magdeburg.de/~borgelt/src/apriori.zip
-->check this page for the latest news of apriori--http://fuzzy.cs.uni-magdeburg.de/~borgelt/doc/apriori/apriori.html

3. Execution
------------------------------

3.1 load program
-->under your GenExplore0.98b root directory, type "./GenExplore" to start the program.

3.2 load data file
-->After the GUI is loaded. Go to "File" menu to open a micro array data file.

       Note: data file format
       ******Important!!!******
       ******Please make sure your have a data file with correct file format:******

GenExplore now only support "Format 1" array data file. (will support Format 2 in the next  version)

3.2.1
-->Format 1: support data files like "data_expts1-300_ratios.txt", you could find this file in "http://www.rii.com/register/cell2000102Hughes/EULA.htm".
      Note:The first row of this file is just a comment, so before you load this file into GenExplore, please delete the first row by using any text editor.

-->The first row of the data file should be:
"UNIQID NAME GWEIGHT sample1 sample2 ... sampleN"

-->from the second row, the data is like:
"YHR007c ERG11,CYP51 0.60 0.001 0.033 ..."

--> GenExplore program will automatically get rid of NAME and GWEIGHT cols to proceed.

3.2.2
-->Format 2: support a simplified file format.

-->The first row of the data file should be:
"UNIQID sample1 sample2 ... sampleN"

-->from the second row, the data is like:
"YHR007c 0.001 0.033 ..."

--> GenExplore program will deal with this kind of data file in future.

3.3 Preprocess

3.3.1 Preprocess/ Apriori
-->after data file was loaded, go to "Preprocess" menu and choose "Apriori", you must perform this operation before you do any further operation. During this step the raw data will be preprocessed (some temporary file will be generated) for further operations. and frequent large item set  of the raw data will also be generated by "Apriori " program.

3.3.2 Preprocess / OpenGTRFile
--> Before you perform this operation, you need use some clustering software, such as http://rana.lbl.gov/downloads/Cluster/Cluster_vers_2_20.exe, to generate hiearchical cluster(*.GTR) file based on the macroarray data you opened. Related software can be found in http://genome-www.stanford.edu/software/

-->A demo.gtr file was provided with the GenExplore package, it was generated from data_expts1-300_ratios.txt, before load this file. Please perform 3.3.1 first.

--> After you perform this operation, a dialog box will pop up for you to select GTR file to open.

--> After file is selected, a new window with menu will pop up.

--> From the menu of this window you can choose use "new" or "old" method to show hierarchical graph of the data.

3.4 Analysis

3.4.1 Analysis / Graphic Gauss Model
-->after you performed 3.3.1, you can see some clusters are listed in the clusters list. However, you can not see the final result graph unless you perform Graphic Gaussian Model analysis. After analysis, select one cluster from the list then you can see the picture of that cluster. Moreover, you can Add genes by using the "Add Gene" button.

3.4.2 Analysis / Loglinear Model
-->after you performed 3.4.1, you can pick up any cluster in the clusters list. And the member genes of that cluster will shown in the Cluster Member browser. Then you can perform Loglinear Model analysis by press the "Loglinear MD" button or click menu item "Analysis/Loglinear Model". The program will analysis current highlighted cluster in the cluster list.

In the Cluster Member browser, each member gene has a check box. You can also select several genes by click their check box, and you must press "New Cluster" to generate a new cluster which only include the genes you selected, then you can perform loglinear model analysis for this new cluster.

3.5 Help / About
-->The help document is under developing now. I think by the next revision, I can finish it and add it to this program. It will be a HTML style document.

4. bug report
------------------------------

If you find any bug in this program, please send the detail info to me, really appreciate your help. My email address is yye@uncc.edu.

I found that while install this program on RH linux 7.x platform, sometimes it can not compile successfully. However RH 8.0/9.0 looks fine for this program, so if you meet any trouble during installation, try a higher version of linux or try the latest GCC or G++, GCC 3.2 or higher version is recommended.

GenExplore is developed and maintainded by Yong Ye & Dr. Xintao Wu & Dr. K.R. Subramnian. Matt Miller contributed on hierarchical cluster visualization part.