Skip to content

Latest commit

 

History

History
135 lines (102 loc) · 5.61 KB

cgb_install.md

File metadata and controls

135 lines (102 loc) · 5.61 KB
Raw

Installing CGB

CGB is distributed as a conda package. The following instructions refer to the installation of CGB using conda.

Installing conda

To install miniconda go to its website and download the appropriate Python 2.7 miniconda installer version [Linux will be assumed in the following].

For Linux, open up a terminal and execute the bash installer: bash Miniconda2-latest-Linux-x86_64 or bash Miniconda2-latest-Linux-x86.sh

(allow it to prepend to PATH the Miniconda2 install location; close terminal and reopen for changes to go into effect)

Creating a conda environment

Create the environment conda create -n cgb python=2.7

Activate the environment source activate cgb

Any changes made from now on (while environment is active) will be made only to this environment.

Setting up the conda environment

Install packages:

  • conda install biopython
  • conda install cached-property==1.3.0
  • conda install cycler==0.10.0
  • conda install scipy==0.17.1
  • conda install matplotlib==1.5.1
  • conda install pyparsing==2.1.4
  • conda install python-dateutil==2.5.3
  • conda install six==1.10.0
  • conda install pytz==2016.4
  • conda install -c conda-forge tqdm
  • conda install --channel https://conda.anaconda.org/etetoolkit ete3==3.0.0b35
  • conda install -c bioconda weblogo
  • conda install networkx

Installing git

If you have not installed git in your system, install git sudo apt-get install git

Link to your github account:

  • git config --global user.name "yourusername"
  • git config --global user.email "youremailaddress"

Creating keys

  • mkdir .ssh
  • ssh-keygen -t rsa -C "youremailaddress"
  • eval "$(ssh-agent -s)"
  • ssh-add ~/.ssh/id_rsa

Open ~/.ssh/id_rsa.pub file with a text editor and copy the contents of the file manually into GitHub --> Account settings --> SSH Keys --> New SSH Key

Test the key: ssh -T git@github.com

Make folder for repo and move there. Then, initialize git and link to and clone GitHub repo:

  • git init
  • git remote add origin https://github.com/ErillLab/cgb.git
  • git clone https://github.com/ErillLab/cgb.git

Dependencies

CGB requires three external programs to be installed: BLAST, CLUSTALO and BayesTraits. Optionally, HMMER can also be installed to enable querying of the COG, eggNOG or PFAM databases. The following instructions are for installing these programs on Linux.

BLAST

On a terminal: sudo apt-get install ncbi-blast+

Check that BLAST is properly installed by asking: which blastp which should return: /usr/bin/blastp

Create a folder where to store BLAST databases (e.g. under your home) mkdir BLASTdbs and export the environment variable export BLASTDB=/home/yourusername/BLASTdbs

CLUSTALO

Install simply through: sudo apt-get install clustalo

Check that CLUSTALO is properly installed by asking: which clustalo which should return: /usr/bin/clustalo

BayesTrait

BayesTrait should be already present in the cgb/bin folder after cloning. If using a 32-bit machine, download the BayesTrait 32-bit executable and replace the file BayesTraitsV2_linux in cgb/bin with the donwloaded executable.

HMMER

Install simply through: sudo apt-get install hmmer

Check that HMMER is properly installed by asking: which hmmsearch which should return: /usr/bin/hmmsearch

COG

In order to query COG with HMMER to assing COG IDs to orthologous groups, you'll need to first download a HMM file for the COG database. A HMM version of the COG database has been assembled by Dibrova et al. (2017) Biol Direct., and can be downloaded from the authors website.

Once donwloaded, simply gunzip and create database with hmmpress: hmmpress COG_database.hmm (you can delete the COG_database.hmm file after this)

eggNOG

In order to query eggNOG with HMMER to assign NOGs to orthologous groups, you'll need to first download the current version of the eggNOG database. You can download any NOG compilation that you deem applicable to your analyses, but for generality the BactNog compilation is recommended.

Donwload and uncompress the ~3 Gb BactNog tarball containing the bacterial NOG Hiden Markov Models (.hmm). tar xvzf bactNOG.hmm.tar.gz (this can take a while) (you can remove the bactNOG.hmm.tar.gz file after this)

Then concatenate the extracted files into a single file with: for f in *.hmm; do cat "$f" >> bact.hmmer; done (this can take quite a while) (you can delete all the .hmm files after this)

And use this file to create a HMMER database with: hmmpress bact.hmmer (this can take a while) (you can delete the bact.hmmer file after this)

PFAM

In order to query PFAM with HMMER to assign PFAM IDs to orthologous groups, you'll need to first download the current version of the PFAM database. See ftp://ftp.ebi.ac.uk/pub/databases/Pfam/releases/. You can download any PFAM compilation that you deem applicable to your analyses, but the latest PFAM full database is recommended.

Donwload and uncompress the PFAM zip file containing the bacterial NOG Hiden Markov Models (.hmm). gzip -d Pfam-A.hmm.gz

And use this file to create a HMMER database with: hmmpress Pfam-A.hmm