Permalink
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
145 lines (114 sloc) 12.9 KB

BioBuilds 2015.11 was released on 23 November 2015.

The most significant change for this release has been the switch from a tarball-based installation process to using Miniconda from Continuum Analytics. We believe that this switch lets us keep the major benefits of the tarball (e.g., minimal installation complexity and self-contained environments), while adding functionality such as the ability to create custom BioBuilds "installs" (in case you don't need all the tools that come with BioBuilds) and access to a broader ecosystem of other packages available through anaconda.org.

In addition, we've opened up the process by which applications are built and added to BioBuilds. Conda recipes for every package included in BioBuilds are now available through the BioBuilds GitHub repository, and we will gladly consider/accept any pull requests should you want to add your favorite application.

Please send bugs reports to bugs@biobuilds.org or create a issue in the GitHub repository.

Minimum Requirements

BioBuilds 2015.11 runs on OS X on x86_64, Linux on x86_64, and Linux for POWER8. Approximately 2.5-GB of disk space is needed for a full installation of BioBuilds.

Please note that BioBuilds packages have been built with more aggressive compiler optimizations, and target more recent processor architectures and operating system releases, than most other Anaconda packages. If you are attempting to run BioBuilds on architectures and/or operating systems older than what is listed below, some or all of the included binaries might not function correctly:

  • OS X on x86_64:
    • Nehalem or later microarchitecture required. If your CPU was built before ~2009 and does not support SSE4.x instructions, some BioBuilds applications may not start, and others may generate "illegal instruction" errors when running.
    • OS X 10.9 (Mavericks) or later required. OS X 10.8 (Mountain Lion) and below have effectively reached end-of-life as of October 2015, and are not officially supported as BioBuilds host operating systems.
  • Linux on x86_64:
    • Nehalem or later microarchitecture required. If your CPU was built before ~2009 and does not support SSE4.x instructions, some BioBuilds applications may not start, and others may generate "illegal instruction" errors when running.
    • glibc 2.12 or later required. Most recent Linux distributions for x86_64 (including Debian 7 and later, Ubuntu 12.04 and later, and CentOS/RHEL 6 and later) meet this minimum and are supported. Note, however, that CentOS/RHEL 5 and distributions based on them run glibc 2.5 and therefore are not supported.
  • Linux on POWER:
    • POWER8 or later, running in little-endian mode required. Support for big-endian and/or POWER7 systems has been dropped as of this release; if you need to run one or more BioBuilds applications on such systems, please contact Lab7 Systems about a paid support contract.
    • glibc 2.17 or later required. Currently, both Ubuntu 14.04.2 and RedHat Enterprise Linux (RHEL) 7 are supported.

Several applications within the BioBuilds package have their own specific requirements. Please refer to the "Other Notes" section below for details.

Install

To install BioBuilds 2015.11, please first download the appropriate Miniconda installer from Continuum Analytics:

Then install Miniconda by running bash Miniconda-latest-<os>-<arch>.sh, where <os> is either "Linux" or "MacOSX", and <arch> is either "x86_64" or "ppc64le"; for further details, refer to the Conda Quick Install Guide. NOTE: If you are already using Continuum's Miniconda or Anaconda system, you can skip this part of the BioBuilds installation procedure.

Linux on POWER8 (ppc64le) users: Before running the rest of the installation, you must first upgrade conda itself by running conda update -c biobuilds -y conda; failing to do so will cause the install commands below to fail with an AssertionError: emboss-6.5.7-0.tar.bz2 MatchSpec(u'libgd') error.

Once miniconda has been installed, you can then install BioBuilds 2015.11 by running:

conda create -c biobuilds -p /path/to/biobuilds-2015.11 biobuilds=2015.11

You can then run BioBuilds applications by supplying their full path (i.e., /path/to/biobuilds-2015.11/bin/<app>), or simply by name after adding the bin directory to your $PATH environment variable (i.e., adding export PATH="/path/to/biobuilds-2015.11/bin:$PATH" to your ~/.bash_profile, then running <app> in your BASH shell).


Note: the above installation procedure does not create a standard Conda environment. If you would like to install and use BioBuilds 2015.11 like any other Conda environment, you should instead run:

conda create -c biobuilds -n biobuilds-2015.11 biobuilds=2015.11

This will create a Conda environment called biobuilds-2015.11, which you can then use standard Conda commands to manipulate; e.g., source activate biobuilds-2015.11 and source deactivate to enable or disable BioBuilds applications in your $PATH, respectively. Refer to the Conda "Managing Environments" documentation for more details.

(This procedure actually installs the BioBuilds environment into /path/to/miniconda/envs/biobuilds-2015.11, so you should be able to add /path/to/miniconda/envs/biobuilds-2015.11/bin to your $PATH and run BioBuilds applications without having to use the normal source activate/source deactivate procedure.)


Since BioBuilds packages are now available through an Anaconda.org channel, you can also use conda to create and use custom installations (environments) that contain only a subset of the BioBuilds applications. For example, if you need just bowtie2 and samtools, you could create a custom environment using:

conda create -c biobuilds -n my-bifx bowtie2 samtools

This will install only bowtie2, samtools, and any needed dependencies into /path/to/miniconda/envs/my-bifx/bin, without any of the other BioBuilds applications that you do not intend to use (e.g., ALLPATHS-LG or EMBOSS). You can then run these applications either by using their full path, by running source activate my-bifx to dynamically add them to your current shell $PATH, or by "permanently" adding /path/to/miniconda/envs/my-bifx/bin to your $PATH in your ~/.bash_profile file.

The -c biobuilds option in the above commands instruct conda to search the BioBuilds Anaconda channel for packages. You can avoid having to provide that option with every conda command by adding the BioBuilds channel to your ~/.condarc file, like so:

channels:
  - biobuilds
  - defaults

Doing so makes it much easier to create a collection of Conda environments, each tailored to the bioinformatics you intend to do; e.g.,

conda create -n basic-bifx blast emboss
conda create -n rnaseq-bifx trinity cufflinks tophat

Refer to Continuum's "Using conda" documentation for more details on using Conda to manage various run-time environments.

Other notes

While every effort has been made to make BioBuilds as self-contained as possible, various applications still require external libraries and interpreters be installed on the system. If these dependencies are currently not available on your system, you will need either sudo access or system administrator help to install them before you can run these applications.

  • Various applications include various Perl support scripts. Currently, an external Perl 5 interpreter must be available in your $PATH; on OS X and most Linux distributions, this should not be a major impediment as many system utilities already require Perl. NOTE: We intend to remove this dependency on the system Perl in the next release (BioBuilds 2016.04).
  • Linux on POWER8 (ppc64le) users will need libffi and the GCC runtime libraries on their systems. These can be installed by running sudo apt-get install libffi6 gcc g++ gfortran on Ubuntu 14.04 or sudo yum install libffi gcc gcc-g++ gcc-gfortran on RedHat 7.
  • Linux users on both x86_64 and ppc64le will need libX11, libXrender, and libXext to run many of the applications that produce graphical or image output (e.g., R or matplotlib). These can be installed by running sudo apt-get install libx11-6 libxrender1 libxext6 on Debian/Ubuntu-based systems or sudo yum install libX11 libXrender libXext on CentOS/RedHat-based systems.
  • A few applications, including IGV, Trinity, and Picard, require a Java Runtime Environment (JRE) be installed and accessible through your $PATH. The JRE can either be the "official" Oracle JRE or the OpenJDK JRE provided by many distribution vendors. In either case, the JRE must support Java 7 or above.
  • The SOAP3-DP and Barracuda applications are GPU-accelerated applications, and therefore, require that the appropriate nVidia hardware and drivers (including libcuda.so) be installed on your system. Additionally, SOAP3-DP also requires version 7.0 of libcudart.so be installed and accessible to the dynamic linker; this can be accomplished by (1) downloading and installing the CUDA 7.0 toolkit from the nVidia website, and (2) including /usr/local/cuda-7.0/lib64 in your LD_LIBRARY_PATH environment variable.

Applications in BioBuilds 2015.11

The following table lists applications included in the BioBuilds 2015.11 release; note that for brevity, certain support libraries and dependent packages (e.g., individual R packages) are not listed.

Application Version New/Upgrade status
ALLPATHS-LG 52488 Upgrade (Linux only)
BAMtools 2.4.0 New
Barracuda 0.7.107b New (Linux only)
bedtools 2.25.0 Upgrade
Bfast 0.7.0a No change
Bioconductor 3.2 Upgrade
BioPython 1.66 New
Boost 1.54.0 No change
Bowtie 1.1.2 Upgrade
Bowtie2 2.2.6 Upgrade
BWA 0.7.12 Upgrade
ClustalW 2.1 New
Cufflinks 2.2.1 No change
EBSEQ (R) 1.10.0 New
EMBOSS 6.5.7 New
FASTA 36.3.8b Upgrade
FastQC 0.11.4 Upgrade
gnuplot 5.0.0 New
Graphviz 2.38.0 No change
HMMER 3.1b2 Upgrade
HTSeq 0.6.1 No change
htslib 1.2.1 No change
IGV 2.3.65 Upgrade
iSAAC 15.04.01 Upgrade (Linux only)
Matplotlib 1.5.0 New
Mothur 1.36.1 Upgrade
NCBI BLAST+ 2.2.31 Upgrade (New on OS X)
Numpy 1.9.2 Upgrade
Oases 0.2.8 New
Picard 1.140 Upgrade
PLINK 1.07 No change
Pysam 0.8.3 New
Python 2.7.10 Upgrade
R 3.2.2 Upgrade
RSEM 1.2.25 New
Samtools 1.2.0 Upgrade
SHRiMP 2.2.3 No change (New on OS X)
SOAP3-DP r177 No change (Linux only)
SOAPaligner 2.20 No change
SOAPbuilder 2.20 No change
SOAPdenovo2 2.4.240 No change (New on OS X)
SQLite 3.8.4.1 No change
STAR 2.4.2a Upgrade (New on OS X)
tabix 1.2.1 Upgrade
TMAP 3.4.0 No change
TopHat 2.1.0 Upgrade
Trinity 2.1.1 Upgrade (New on OS X)
variant_tools 2.6.3 New
Velvet 1.2.10 No change