ClusterScan Installation
ClusterScan requires Python (v2.7.x). Bedtools (v2.25.0+) and R (v3.0.0+) are needed to be in the user path. In order to draw high quality clusters distributions for features in the top 10 clusters (by number of features) found, it is also required to install the R library ggplot2 (v2.0.0+).
- Installing Bedtools
- Installing R and the ggplot2 library
- Installing the required Python libraries
- Installing ClusterScan
To install bedtools you need to download the package via GitHub and compile from source:
wget https://github.com/arq5x/bedtools2/releases/download/v2.26.0/bedtools-2.26.0.tar.gz
tar -zxvf bedtools-2.26.0.tar.gz
cd bedtools2
make
Once you have finished you can add bedtools to your path opening a new terminal and typing:
sudo nano ~/.bashrc
scroll until you reach the bottom of the page and add the following string:
export PATH=$PATH:/<path-to-bedtools>/bedtools2/bin
Save and exit. To test your bedtools installation you must open a new terminal and type:
bedtools -h
More informations about the bedtools installation can be found here.
To install R simply type:
sudo apt install r-base r-base-dev
and enter the program with the command:
R
now you can install ggplotw with:
install.packages("ggplot2")
Finally you can test the ggplot installation with:
library(ggplot2)
ClusterScan needs few other Python libraries which can be easily installed via pip from PyPI. As first step install python setuptools and pip on your distribution:
sudo apt install python-pip
sudo apt install python-setuptools
Then install the required dependencies via pip typing:
pip install pandas==0.19.1
pip install rpy2==2.8.3
pip install pybedtools==0.7.8
You can find more informations at:
Now, you can download and extract ClusterScan in your preferred directory:
wget https://github.com/pyrevo/ClusterScan/archive/master.zip
unzip master.zip
Finally, you can add ClusterScan in your path by editing your bashrc file:
sudo nano ~/.bashrc
export PATH=$PATH:/<path-to-clusterscan>/ClusterScan-master
Save and exit. To test the clusterscan installation you must open a new terminal and type:
clusterscan.py -h
If you visualize the ClusterScan help page, you can proceed to run your first analysis by following the ClusterScan Tutorial.
We have tested ClusterScan on Ubuntu (v14.04LTS+). ClusterScan is distributed under the GNU General Public License (GPL) Version 3.