EPIG-Seq Files

Installation instructions for the EPIGSeqGUI Program

EPIGSeqGUI is a Java application developed for Red Hat Linux Linux 6 or Ubuntu 12.04 LTS x86-64 bit operating systems with multicores for parallel processing.

Unzip the EPIGSeqGUI folder to a directory of your choice.
The folder must contain the following files and folder:
	The bin folder: 
		EPIGSeqGUIJavaLinuxv01.jar:The Java Jar file
		log.txt: A log file of the processisng
		MyAppInstaller_web.install:  The MATLAB Compiler (version 8.3/R2014a) Runtime (MCR) application binary web installation executable (this consists of the installation for the math and graphics libs)
		EPIGSeqGUILaunch.java:  The Java wrapper to launch the EPIGSeqGUI
		EPIGSeqLogov01.jpg:  The logo for EPIGSeqGUI
	Sample_data folder: 
		simulated_data.txt: A test tab-delimited file with sample data (n_samples = 201, n_groups = 4, n_patterns = 5, n_genes_per_pattern  = 200).  
				The 1st row must contain the labels of the groups that the samples belong to.
				The 2nd row must contain the total mapped reads for each sample.
				The 1st column must contain the unique gene IDs

				**** The data for processing must be tab-delimited.  No other delimiter is supported at this time. ****
					
	A folder with sample output: 
		categorized_genes.txt: categorized genes
		pattern_pvalues.txt: p-values for the patterns
		outfile.txt: Dispersion for each gene
		outfilePValues.txt: p-values for the categorized genes
		EPIG_pattern_data.txt: data to estimate dispersion from the model
		EPIG_pattern_data2.txt: data to compute the p-value from the model
		

	A folder with the R code: R_code containing two files: quasipoisson.R and quasipoisson2.R.  The R_code folder must be located in the directory one level above where EPIGSeqGUIJavaLinuxv01.jar is located.
	
	This readme file: readme.txt

Verify that the Matlab Runtime Compiler (MCR) is installed and ensure you have version 8.3 (R2014a) 
If the MCR is not installed, do the following:


Execute the MyAppInstaller_web.install file to extract and install Matlab library files via the web.  You will need internet access and may need to be an administrator to do so.  


The MCR Installer will launch.  Click next, then accept agreement, then indicate the path where you want the runtime environment installed.  This will be your <mcr_root>
The install may take a couple of hours depending on your network speed.

Or manually download and install the Linux x86-64-bit version of the MCR for R2014a 
from the MathWorks Web site by navigating to

   http://www.mathworks.com/products/compiler/mcr/index.html
   

Once the Matlab libraries are installed, the workstation is ready to run the Java standalone application. 

Ensure you have version 1.7 of the Java Development Kit (JDK) installed, operable and in your user environment to compile and launch the EPIGSeqGUI components. 

EPIGSeqGUIJavaLinuxv01.jar must be included in your CLASSPATH.

javabuilder.jar must be included in your CLASSPATH. javabuilder.jar can be found in: 
  
  <mcr_root>/toolbox/javabuilder/jar/javabuilder.jar

Finally, you must have a Linux x86-64 installation of R version 3.0.2 or higher installed and accessible in the user's environment path.  


Also, on the target machine/workstation, add the MCR directory to the following environment variable by issuing the following commands:

        NOTE: <mcr_root> is the directory where MCR is installed
              on the target machine/workstation.         

            setenv LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<mcr_root>/v83/runtime/glnxa64:<mcr_root>/v83/bin/glnxa64:<mcr_root>/v83/sys/os/glnxa64
            setenv XAPPLRESDIR=$XAPPLRESDIR:<mcr_root>/v83/X11/app-defaults

	On some systems you may need to set the environment variables this way:

		export LD_LIBRARY_PATH=<mcr_root>/v83/runtime/glnxa64:<mcr_root>/v83/bin/glnxa64:<mcr_root>/v83/sys/os/glnxa64
 
		export PATH=$PATH:$LD_LIBRARY_PATH

	To make the environement variables permanent in order to load each time you log into your shell, add them to your bash profile.

Navigate to the EPIGSeqGUI bin folder.


Execute the following command to compile the jar file:
>javac -classpath "<mcr_root>/toolbox/javabuilder/jar/javabuilder.jar":./EPIGSeqGUIJavaLinuxv01 ./EPIGSeqGUILaunch.java

Execute the following command to launch the EPIGSeqGUI application:
>java -classpath .:"<mcr_root>/toolbox/javabuilder/jar/javabuilder.jar":./EPIGSeqGUIJavaLinuxv01 EPIGSeqGUILaunch

The GUI splash screen may take a while to appear. 

It is estimated that to run EPIGSeqGUI, 450MB of available RAM is needed per 1 MB of data (sized on the Windows file system) per CPU.  
For example, if the data file on Linuxis 8.2MB, 3.7 GB of available RAM is needed per CPU.
It is best that EPIG-Seq run on a Linux x86-64 bit workstation or server with the recommended available RAM.   
The fewer CPUs used, the longer the procesisng time.


Email Pierre R. Bushel at bushel@niehs.nih.gov if you encounter problems.