Skip to content

Latest commit

 

History

History
103 lines (73 loc) · 3.02 KB

environment.md

File metadata and controls

103 lines (73 loc) · 3.02 KB

RASE computational environment

The easiest way how to setup a computational environment for RASE is using Bioconda. This approach has been tested on multiple Unix and OS X machines, including clusters and virtual machines.

Dependencies

Setting up an environment

Using a separate BioConda environment

We recommend to create a separate software environment (here called rase):

conda create -n rase \
	prophyle ete3 pysam snakemake-minimal samtools parallel r-optparse

The environment can then be activated by

source activate rase

Using the default BioConda environment

Alternatively, the packages can also be installed directly into the default BioConda environment. Nevertheless, this is not always reliable since some of the RASE dependencies might collide with packages that were installed previously.

conda install prophyle ete3 pysam snakemake samtools parallel r-optparse

Alternative ways of installation

All the dependencies can also be installed without BioConda.

Many of these packages are distributed using standard package systems such as APT.

apt-get install build-essential python3 zlib1g-dev r-base r-cran-optparse ghostscript

All the Python packages (ProPhyle, PySAM, ETE 3, and Snakemake) can be installed using PIP:

pip3 install prophyle pysam ete3 snakemake

Known issues

libR.dylib Reason: image not found

At some systems, the R package distributed by Bioconda might not be properly built and would display messages such as

dyld: Library not loaded: @rpath/libintl.9.dylib
  Referenced from: /Users/user/miniconda/envs/rase/lib/R/lib/libR.dylib
  Reason: image not found
Abort trap: 6

The solution is then to create the rase environment without r-optparse, and to install R and the OptParse package manually.

ETE: cannot connect to X server

ETE 3 library, which is used for tree plotting, internally depends on QT and requires using an X-Server. This becomes problematic especially on virtual machines. For instance, on Ubuntu-based machines this can be solved by installing several additional packages:

apt-get install xvfb libqt4-dev libgl1-mesa-dev libglu1-mesa-dev xauth xfonts-base

and then prepending the following string to commands for building the database.

xvfb-run --server-args="-screen 0 1024x768x24 -noreset" \