Anaconda Installation¶

This section describes installation via Anaconda Python’s package manager, conda. If you want to do a direct install, you want the previous chapter.

If you don’t have Anaconda Python installed, get it by visiting the URL https://www.continuum.io/downloads. lorax was tested under version 3.6 of Anaconda Python. As of this writing, there is no compelling reason to use earlier or later versions. You must use a 64-bit version. The command-line installer version is fine, and 25% smaller. When you have downloaded the installer, issue the following command from a terminal window in the Downloads directory:

bash Anaconda3-<version>-<platform>.sh

where <version> is the current version number and <platform is your OS. As part of installation, please ensure that the Anaconda Python bin/ directory is prepended to your PATH (in ~/.bash_profile on the Mac) or else you will have to prepend it manually before continuing the installation and every time you wish to use lorax.

Configure channels¶

lorax requires some packages that are in non-default channels, which may be obtained by the following:

conda config --add channels conda-forge
conda config --add channels defaults
conda config --add channels bioconda

After the channels are configured, update the installation by issuing:

conda update --all

This command usually results in several packages being updated.

Create a virtualenv¶

It is not advisable to install lorax in the root environment because lorax’s dependencies could break other packages. As of the current conda version (4.3.25), there are four packages in the channels that are too old for use with lorax:

Package	Anaconda Version	Required Version
setuptools	27.2.0	>30.3.0
markupsafe	0.23	1.0.0
gunicorn	19.1.0	19.7.1
supervisor	3.3.2	4.0 (from git)

Replacing setuptools, in particular, could cause breakage of your root Anaconda Python distribution. For this reason, it is required to create a python virtual environment for lorax and its dependencies. I prefer to name this environment “loraxenv” to prevent confusion between package and environment names. You should create this environment as the user under which you intend the web server to run (using sudo -u USER if necessary), using the following commands:

conda create -y --name loraxenv biopython click hmmer raxml redis nginx
source activate loraxenv

Install gcc if using MacOSX¶

If you are installing under MacOS (and only under MacOS), gcc and libgcc MUST be installed in the current virtual environment:

conda install -y gcc libgcc

Failure to do this step will result in a message to the effect “Unable to resolve paths in MacOS executable” in the lorax installation step later.

Upgrade setuptools¶

Next continue the installation by updating setuptools and installing a special version of supervisord from a git repository. The first will end with an error; this is expected, so don’t be worried by it:

pip install -U setuptools
pip install -e 'git+https://github.com/LegumeFederation/supervisor.git@4.0.0#egg=supervisor==4.0.0'

Choose root location¶

When lorax installs itself, it will place some files in the bin/ and etc/ subdirectories of LORAX_ROOT. Normally LORAX_ROOT is set to the same directory as parent of the python installation in use (that is, in sys.prefix). This is usually fine for most installations, but if you wish to override it, you must define LORAX_ROOT before installation:

export LORAX_ROOT=some_directory_you_choose

Install lorax¶

Next, install the lorax dependencies and binaries:

pip install -U lorax

Put lorax_env script in alias or on path¶

The lorax_env script is the only executable that you will need to control lorax and associated process. You should usually not put the entire lorax bin/ directory into your path, as this may cause binaries from the virtual environment such as python to be used in contexts where they were not intended to be used. The easiest way to do this is with a symlink to a directory on your path, such as the Anaconda Python directory:

ln -s /path/to/anaconda/envs/loraxenv/bin/lorax_env /path/to/anaconda/bin

Deactivate virtual environment¶

You no longer need to work inside the virtual environment you created. Leave the virtual environment with the following command:

source deactivate

Next, follow the configuration instructions.