IMP is a pipeline that is composed of multiples major steps:
To get more insight about how the pipeline is working, please read the paper at bioRxiv.
IMP is designed to allow the use from two different setups:
The easiest way that is mostly automatic is to use the command-line utility
impy that will
take care of installing and running the workflow for you that uses docker containers in the backend.
You also may not be able to install docker in your infrastructure and you will have to use the base workflow to run IMP. It is powered by
We decided to use docker to package every dependencies for IMP. As quoted from their website:
“Docker containers wrap up a piece of software in a complete filesystem that contains everything it needs to run: code, runtime, system tools, system libraries – anything you can install on a server. This guarantees that it will always run the same, regardless of the environment it is running in.”
Docker website provides you the way to install the docker engine on multiple distributions.
pip install impy
If the default version of python in your environment is
python2 please use:
pip3 install impy
Now follow instructions on how to use impy.
To get the latests version of IMP:
git clone https://git-r3lab.uni.lu/IMP/IMP.git
If you don’t have
gitinstalled you could also download and unzip a packaged version available here.
IMP uses more than 40 tools packaged together, it is very difficult to provide a script that will install everything smoothly on every operating systems. We don’t provide such a script, but you can find some hints recipes under the
docker/ source code directory. Those recipes are for
ubuntu:trusty but should be similar for other *nix.
Dockerfile-dependencies list all low-levels dependencies that you must install on you base system.
Dockerfile-tools list all the tools installed and their version used int IMP.
Dockerfile list all
R dependencies installed inside the container, and contains some modifiers for environment variables.
Now follow instructions on how to use snakemake.
Facing some problems installing/running IMP? You may get help by opening an issue on the Gitlab issues page where you can directly get in touch with the developers.