jcvi-cloud/vicvb

Genome browsing with JBrowse for JCVI VIGOR output inside Galaxy

Venter Institute Cloud Viral Browser (VICVB)

This package takes the output of VIGOR (JCVI viral annotation pipeline) and generates the input for JBrowse genome browser. The primary goal is to use it as a visualization tool inside [Galaxy] (http://galaxyproject.org/) bioinformatics workbench. The package has an installer script that installs JBrowse into Galaxy, and the dataset convertion procedure generates files necessary for launching JBrowse from Galaxy.

Installation

This package uses Python Distribute module. You can use any of the build/install modes that Distribute supports (e.g. easy_install or python setup.py install). After the package is installed, you should run vicvb_postinstall script to install JBrowse and create corresponding config file.

For a one-stop typical installation into Galaxy, a script is provided, which can be executed like this:

git clone git://github.com/JCVI-Cloud/VICVB.git
cd VICVB
lib/VICVB/data/install/install_to_dir_full.sh <PACKAGE_INSTALL_DIR> <GALAXY_HOME> <GALAXY_ROOT_URL>

Replace the references to each bracketed name above with your actual values. Typically, GALAXY_ROOT_URL is just /. The command above will install the package using Python's --install-dir scheme - everything will be placed as eggs into PACKAGE_INSTALL_DIR. VICVB.rc file will be generated in PACKAGE_INSTALL_DIR. You should source it before using the package. The JBrowse will be installed under GALAXY_HOME/static.

The script above will copy all Python dependencies into PACKAGE_INSTALL_DIR. You can edit the install script to change that and other subtleties.

The package assumes that NCBI tbl2asn executable is available in system PATH.

Testing

Conversion to JBrowse and viewing outside of Galaxy

To test the conversion procedure, from the package source dir do:

python setup.py test
#replace run.test.OAtT6g.tmp below with the unique directory name that was created by the test run
cd test_run/run.test.OAtT6g.tmp
python -m SimpleHTTPServer 8090
#point your Web browser to localhost:8090 and 
#click on any of the directory names that look like *.jbrowse.
#You should see the genome(s)

Conversion and viewing inside Galaxy

To use this package in Galaxy, exposing it as a separate Galaxy tool is not needed. It can be just executed as part of any other tool command line, creating an extra output dataset, which can be then viewed in JBrowse. For testing purposes, we have provided a couple of demonstration tools which take a VIGOR output as a tarball (it should also contain nucleotide FASTA input to VIGOR) and generates JBrowse input (as Galaxy compound dataset). See bits/galaxy-tool/Readme.txt in the source tree (especially the part about switching off HTML sanitization). After you have installed the sample Galaxy tool, upload lib/VICVB/data/test_data/VIGOR/NCBI.one_arch/Rhinovirus.tgz into your Galaxy history and run the tool (under section VICVB) on this dataset. The code of these sample tools can be reviewed as examples of calling the package from Galaxy.

To view the output in JBrowse, click on the Eye icon of the resulting history item. You should see a genome in the central Galaxy panel.

Running

See the command lines of the sample Galaxy tools referenced in the previous section. To see other options, run the main script "vicvb_galaxy_tool --help". Note that the input can be supplied either as a tarball or as a directory produced by VIGOR.