For quick installation instructions for OS X and Ubuntu, see the README. This file has more detailed instructions for installation on other platforms or for developers.
After installation, proceed to the configuration instructions at the end of this document.
If you would like to install to a location other than /usr/local (if you don’t have permission to write to /usr/local, for example), see the “Installing to an alternative location” section.
This section lists required and optional prerequisites, with notes on specific versions we have tested and found to work.
To compile and install BioLite, you must at a minimum have:
A C/C++ compiler that supports OpenMP and the TR1 standard. Tested:
- gcc 4.4.6 (CentOS 6.3)
- gcc 4.6.3 (Ubuntu 12.04)
- XCode gcc 4.2.1 (OS X 10.8)
Python (2.7.2, 2.7.3) with packages:
- biopython (1.60, 1.61)
- dendropy (3.12.0)
- docutils (0.9.1)
- matplotlib (1.1.0, 1.1.1rc, 1.1.1)
- networkx (1.6, 1.7)
- numpy (1.6.1, 1.6.2)
BioLite provides a large collection of wrappers for the following 3rd party bioinformatics tools. While you do not have to install these to be able to load the BioLite python library or to use the BioLite command-line tools, a BioLite script that calls a wrapper must be able to find the corresponding program in your PATH. BioLite comes with a shell script to automate downloading and building many of these 3rd party programs. See the “Installing 3rd Party Software” section below for more details.
FastQC 0.10.0 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/)
Blast+ 2.2.27 (http://blast.ncbi.nlm.nih.gov/)
Bowtie 0.12.8 (http://bowtie-bio.sourceforge.net/)
Bowtie2 2.0.6 (http://bowtie-bio.sourceforge.net/bowtie2/)
samtools 0.1.18 (http://samtools.sourceforge.net/)
Velvet 1.2.08 (http://www.ebi.ac.uk/~zerbino/velvet/)
Note: with LONGSEQUENCES=1 and MAXKMERLENGTH >= 61, recommended 63
Oases 0.2.08 (http://www.ebi.ac.uk/~zerbino/oases/)
Note: with LONGSEQUENCES=1 and MAXKMERLENGTH >= 61, recommended 63
Trinity r2012-10-05 (http://trinityrnaseq.sourceforge.net/)
MACSE 0.9b1 (http://mbb.univ-montp2.fr/macse/)
Note: if you install manually, make sure the MACSE jar file is in your PATH.
RAxML 7.2.8-ALPHA (http://sco.h-its.org/exelixis/software.html)
Gblocks 0.91b (http://molevol.cmima.csic.es/castresana/Gblocks.html)
mcl 12-135 (http://micans.org/mcl)
Unpack the tarball (from https://bitbucket.org/caseywdunn/biolite/downloads):
tar xf biolite-X.X.X.tar.gz
cd biolite-X.X.X
Build and install 3rd party tools:
sudo ./build_3rd_party.sh /usr/local
Build and install BioLite:
./confgure make sudo make install
Proceed to Configuration at the end of this document.
(Skip this unless you are building a development version that you cloned from Bitbucket.)
First, create a fork of the repository so that you can later send a pull request so that we can incorporate your improvements. Then clone your forked repository.
Build and install 3rd party tools:
sudo ./build_3rd_party.sh /usr/local
See the Prerequisites section above for other software you may need to install.
You will need to have the automake, autoconf and libtool packages installed. To produce the configure scripts, use the command:
./autogen.sh
Then you can use the typical build sequence:
./configure
make
sudo make install
Proceed to Configuration at the end of this document.
The instructions above assume that you are installing BioLite to /usr/local, which requires root access. If you are installing to another location, modify the installation instructions as follows. These are not full instructions, they just explain how to modify the instructions above.
In the instructions below, we use [installation_path] as a placeholder for the path that you would like to install to, eg /home/lucy/local. Note that [installation_path] needs to be an absolute path.
Modify the command to build and install 3rd party tools to specify the alternative path:
./build_3rd_party.sh [installation_path]
Modify the configure command to both change the install location and tell it where third party packages were installed:
./configure --prefix=[installation_path]
make
make install
If you are installing somewhere that you have write access to, you don’t need to use sudo for build_3rd_party.sh or make install.
Proceed to the next section for instructions on setting paths.
If you install to a canonical location on your system, like ‘/usr/local’, the scripts and programs will already be in your PATH and the python module will be ready to import.
Otherwise, if you want to be able to call the programs without specifying their full path, you need to add the new ‘bin’ directory to your PATH. In bash (you can add add this to ~/.bashrc):
export PATH=[installation_path]/bin:$PATH
or in csh:
setenv PATH [installation_path]/bin:$PATH
To be able to import BioLite’s python modules in python, you will also need to add the full path to you PYTHONPATH, replacing the python version below with your version of python (most likely “2.7”). In bash:
export PYTHONPATH=[installation_path]/lib/python2.7/site-packages:$PYTHONPATH
or in csh:
setenv PYTHONPATH [installation_path]/lib/python2.7/site-packages:$PYTHONPATH
The BioLite source comes with a shell script that will download the required 3rd party software from a mirror at Brown’s Center for Computation and Visualization. Blast+ and FastQC are installed as binaries and the other packages are compiled from source. The usage for the script is:
./build_3rd_party.sh -h
usage: build_3rd_party.sh [PREFIX] [CC] [CXX] [OPT]
NOTE We use this script internally to install and test BioLite, and we have only tested it on our own systems. It is likely that you will need to manually install additional dependencies on OS X (or use a system like macports, homebrew, or Fink), or install additional packages on Linux through your distro’s package manager (especially some development packages that end in -dev or -devel).
To install to ‘/usr/local/’, you can call the script with no arguments. To use a different install path, specify a PREFIX, for instance your home directory:
./build_3rd_party.sh $HOME
If you want to specify a different compiler, use the CC and CXX options. For instance, your linux distro may have a gcc 4.6 package that installs ‘gcc46’, so you would use:
./build_3rd_party.sh /usr/local gcc46 g++46
Finally, if you want to specify more aggressive compiler optimizations, use the OPT option. If you have a newer CPU that supports SSE4.2 instructions (e.g. Intel Nehalem), you could use:
./build_3rd_party.sh /usr/local gcc g++ -msse4.2
The included release.sh script will update the git version, rebuild the documentation and run the necessary configure/make commands to create a tar ball.
To build the documentation, you must install the ‘sphinx’ Python package, for instance with:
sudo pip install sphinx
or:
sudo easy_install sphinx
After successfully installing BioLite with make install, you should see a message like:
|------------------------------------------------------------------------------
| BioLite has been installed to /usr/local
|
| Your default configuration file is located at:
|
| /usr/local/share/biolite/biolite.cfg
|
|------------------------------------------------------------------------------
pointing to the location of your default BioLite configuration file. This file serves as the default configuration for any user on the system. To override it on a per-user basis, simply copy the file to $HOME/.biolite/biolite.cfg and make any required changes.
You can also override the location of the configuration file with an environment variabl. In bash:
export BIOLITE_CONFIG=/your/path/to/biolite.cfg
or in csh:
setenv BIOLITE_CONFIG /your/path/to/biolite.cfg
Finally, the BIOLITE_RESOURCE environment variable allows you to temporarily override specific values in the resources section of the configuration. For instance, if your configuration file is set to 2 threads, but want to test out a run with 8 threads instead, you could use (in bash):
export BIOLITE_RESOURCES="threads=8"
The value of this variable can be a comma-separated list of key=value pairs.