Developer Installation

If you would like to help with Pathomx development you will need to install a source version of the code. Note: This is not necessary if you just want to contribute plugins, as these can be developed against the binary installation.

Getting Started

The development code is hosted on Github. To contribute to development you should first create an account on Github (if you don’t have one already), then fork the pathomx/pathomx repo so you have a personal copy of the code. If you’re not familiar with Github, there is a useful guide available here.

On your version of the repo (should be <username>/pathomx) you will see an url to clone the repo to your desktop. Take this and then from the command line (in a folder where you want the code to live) enter:

git clone <repository-url>

After a while you will get a folder named pathomx containing the code.

The following sections list platform-specific setup instructions required to make Pathomx run. Follow the instructions from the section and then you should be ready to run from the command line using:

python -m pathomx.Pathomx


Install Qt5 (Qt5.2) for Windows. Make the decision at this point whether you’re installing 64bit or 32bit and stick to it.

Install Python 2.7.6 Windows installer from the Python download site.

You can get Windows binaries for all required Python libraries from the Pythonlibs library. At a minimum you will need to install NumPy, SciPy, Scikit-Learn, Matplotlib. Make sure that the installed binaries match the architecture (32bit/64bit) of the installed Python.

For NMR data processing, you will need to install NMRGlue binaries.

For the dynamic pathway drawing plugin MetaViz you will also need to install Graphviz.

To run Pathomx from the command line, change to the cloned git folder and then enter:

python -m pathomx.Pathomx


The simplest approach to setting up a development environment is through the MacOS X package manager Homebrew. It should be feasible to build all these tools from source, but I’d strongly suggest you save yourself the bother.

Install Homebrew as follows:

ruby -e "$(curl -fsSL"

Once that is in place use brew install to install python, PyQt5 (which will automatically install Qt5) and graphviz. Install pip for Python and add the packages numpy, scipy, pydot, nmrglue, gpml2svg, poster, wheezy, sklearn, icoshift, matplotlib. This can be done in a one liner with pip:

pip install numpy scipy pydot nmrglue gpml2svg poster wheezy sklearn icoshift matplotlib

That should be enough to get Pathomx up and running from the command line. For development a useful tool to install is Total Terminal, which gets you access to the command line via a hotkey.

To run Pathomx from the command line, change to the cloned git folder and then enter:

python -m pathomx.Pathomx


The development version (available via git) supports Python 3 and so can now be run on Linux (tested on Ubuntu Saucy Salamander). Note: Python 3 PyQt5 is only available from 13.10. To install on earlier releases of Ubuntu you will need to install from source.

There are a number of packages that need to be installed first:

sudo apt-get install python3-pyqt5 python3-matplotlib python3-requests python3-numpy python3-scipy python3-yapsy

pip3 install scikit-learn

Note that aside from python3-pyqt5 you can also install the other packages using pip3 (the names on PyPi are the same as for the packages minus the python3- prefix). Once installation of the above has completed you’re ready to go.

To run Pathomx from the command line, change to the cloned git folder and then enter:

python -m pathomx.Pathomx