Birdsuite: Install

• Limitations
• Prerequisites
• Prerequisites for building executables from sources
• Files
• Installation
• Testing

Limitations

• Currently, this Birdsuite release only runs on 32- or 64-bit Linux systems, although it may be possible to make it work on other systems, particularly if you have Matlab for the target system.
• Currently, Birdsuite metadata exists only for Affymetrix Genome-Wide Human SNP Array 6.0.
• Some files are not distributed by Broad Institute, and must be obtained from other parties.

Prerequisites

• Java 1.5 or higher http://www.java.com
• Python 2.5.2 or higher http://www.python.org
• numpy Python package, version 1.2 or higher http://numpy.scipy.org
• R 2.4 or higher http://www.r-project.org
• mclust R package, version 3 or higher http://lib.stat.cmu.edu/R/CRAN/web/packages/mclust/index.html
• Affymetrix Power Tools (APT) 1.8.6 or 1.10.2, binaries http://www.affymetrix.com/support/developer/powertools/index.affx
• Affymetrix library bundle for Genome-Wide Human SNP Array 6.0 http://www.affymetrix.com/Auth/support/downloads/library_files/genomewidesnp6_libraryfile.zip

Prerequisites for building executables from sources

• For birdseye, Matlab compiler is required. http://www.mathworks.com
• For birdseed and larry_bird, the following are required:
  • G++ 4.1.1 or higher http://gcc.gnu.org
  • Affymetrix Power Tools (APT) 1.8.6 or 1.10.2, sources and libraries http://www.affymetrix.com/support/developer/powertools/index.affx

Files to download

The following files are available for download:

File	Description
birdsuite_executables_<version>.tgz	Scripts and executables that are already built for Linux
birdsuite_metadata_<version>.tgz	Metadata required by Birdsuite
MCRInstaller.75.glnxa64.bin.gz	Matlab 7.5 compiled runtime for 64-bit Linux
MCRInstaller.75.glnx86.bin.gz	Matlab 7.5 compiled runtime for 32-bit Linux
birdsuite_inputs_<version>.tgz	Test inputs for birdsuite (optional)
birdsuite_outputs_<version>.tgz	Output from test inputs (optional)
birdsuite_sources_<version>.tgz	Source code (optional)

Click here to download files.

Installation

1. Create a directory where all the Birdsuite executables will go. We'll call this EXEDIR.
2. Unpack birdsuite_executables_<version>.tgz into EXEDIR.
3. Install python packages. Please follow (CLICK THIS LINK) python install documentation
4. For 64-bit, copy (or symlink) the 64-bit version of apt-probeset-summarize from Affymetrix Power Tools into EXEDIR, with the name apt-probeset-summarize.64 . For 32-bit, copy or symlink the 32-bit version of apt-probeset-summarize into EXEDIR, with the name apt-probeset-summarize . Note that you may do both of the above to run on both 32- and 64-bit systems.
5. Run the following commands:
   
   R CMD INSTALL -l EXEDIR EXEDIR/broadgap.utils_<version>.tar.gz
R CMD INSTALL -l EXEDIR EXEDIR/broadgap.cnputils_<version>.tar.gz
R CMD INSTALL -l EXEDIR EXEDIR/broadgap.canary_<version>.tar.gz
6. Install Matlab compiled runtime:
   • For 64-bit:
     1. Download MCRInstaller.75.glnxa64.bin.gz and decompress into temporary directory.
     2. Run the following command:
        MCRInstaller.75.glnxa64.bin -console
     3. When prompted for a directory into which to install MCR, reply with EXEDIR/MCR75_glnxa64
     4. MCRInstaller.75.glnxa64.bin may be deleted after installation is complete.
   • For 32-bit:
     1. Download MCRInstaller.75.glnx86.bin.gz and decompress into temporary directory.
     2. Run the following command:
        MCRInstaller.75.glnx86.bin -console
     3. When prompted for a directory into which to install MCR, reply with EXEDIR/MCR75_glnx86
     4. MCRInstaller.75.glnx86.bin may be deleted after installation is complete.
   Note that you may do both of the above to run on both 32- and 64-bit systems.
7. Create a directory where all the Birdsuite metadata will go. We'll call this METADATADIR.
8. Unpack birdsuite_metadata_<version>.tgz into METADATADIR.
9. Copy the following files from the Affymetrix library bundle (genomewidesnp6_libraryfile.zip) into METADATADIR: GenomeWideSNP_6.Full.cdf, GenomeWideSNP_6.specialSNPs. For 5.0 chip, copy the analogous files.
10. Edit EXEDIR/birdsuite.sh, and add the following arguments to the java command line: --exeDir=EXEDIR --metadataDir=METADATADIR

If you wish to use the birdsuite_qc.py script, which makes running mini-DM and gender calling easier, follow these additional installation steps.

1. Copy or symlink GenomeWideSNP_6.cdf into METADATADIR. For 5.0 chip, copy the analogous files.
2. Copy or symlink apt-geno-qc (from APT) into EXEDIR.

Testing

1. Create a directory (TESTDIR) and unpack birdsuite_inputs_<version>.tgz there.
2. cd to TESTDIR.
3. Run the following command:
   
   EXEDIR/birdsuite.sh --basename=test --chipType=GenomeWideSNP_6 --outputDir=output --genderFile=test.gender --celFiles=test.cels --noLsf --apt_probeset_summarize.force
4. Create a directory (EXPECTEDDIR) and unpack birdsuite_outputs_<version>.tgz there.
5. Compare the files test.birdseed_calls, test.birdseed_confs, test.canary_calls, test.canary_confs, test.larry_bird_calls, test.larry_bird_confs in EXPECTEDDIR and TESTDIR/output. They should be identical.