Salad is an Open Source Python library to segment tissues into different biologically relevant regions based on Hidden Markov Random Fields. The algorithm is loosely inspired from Zhang et al..

Starting from a networkX graph where each cell possesses a categorial feature cell_type, Salad creates a network with an identical structure with homogeneous regions corresponding to specific mixes of cells. This allows for unsupervised segmentation of the tissue.


If it’s not already done, install Anaconda. Then you should install Salad in a local environment. For the moment, only installation via the repository is possible. In the command prompt enter:

git clone

This should download the project. Then install the required packages. In the terminal, navigate to the package directory and type:

pip install .

Congratulations! You’re ready to play. The library is now installed on your computer and is ready to be used from jupyter.


A working example of the library can be accessed here


View Github