Our main focus is the developement of a living tissues simulation tool suite. At its core is an open source library:

The tyssue simulation library

A nice banner

The tyssue library seeks to provide a unified interface to implement bio-mechanical models of living tissues. It’s main focus is on vertex based epithelium models.

You can browse its documentation on readthedocs.


What kind of Models does it implement?

The first model implemented is the one described in Monier et al. [monier2015apico]. It is an example of a vertex model, where the interactions are only evaluated on the apical surface sheet of the epithelium. The second class of models is still at an stage. They implement a description of the tissue’s rheology, within a dissipation function formalism.

The two models considered

General Structure of the modeling API

Design principles

APIs not apps

Each biological question, be it in morphogenesis or cancer studies is unique, and requires tweeking of the models developed by the physicists. Most of the modelling softwares follow an architecture based on a core C++ engine with a combinaison of markup or scripting capacities to run specific simulation.

In tyssue, we rather try to expose an API that simplifies the building of tissue models and running simulations, while keeping the possibilities as open as possible.

Separate structure, geometry and models

We seek to have a design as modular as possible, to allow the same epithlium mesh to be fed to different physical models.

Accessible, easy to use data structures

The core of the tyssue library rests on two structures: a set of pandas DataFrame holding the tissue geometry and associated data, and nested dictionnaries holding the model parameters, variables and default values.

Tyssue data structure

The API thus defines an Epithelium class. An instance of this class is a container for the datasets and the specifications, and implements methods to manipulate indexing of the dataframes to ease calculations.

The mesh structure is heavily inspired by CGAL Linear Cell Complexes, most importantly, in the case of a 2D vertex sheet for example, each junction edge between the cells is “splitted” between two oriented half edges.

Creating an Epithelium

## Core object
from tyssue.core.sheet import Sheet
## Simple 2D geometry
from tyssue.geometry.planar_geometry import PlanarGeometry
## Visualisation (matplotlib based)
from tyssue.draw.plt_draw import sheet_view

sheet = Sheet.planar_sheet_2d('basic2D', nx=6, ny=7,
                              distx=1, disty=1)


  • Easy data manipulation.
  • Multiple geometries (Sheets in 2D and 3D, monolayers, bulk, cell centered models…).
  • Easy to extend.
  • 2D (matplotlib) and 3D (threeJS) customisable visualisation.

Mailing list: -

Subscribe ➙ Unsubscribe ➙


  • Bertrand Caré - @bcare
  • Cyprien Gay - @cypriengay
  • Guillaume Gay (maintainer) - @glyg
  • Hadrien Mary (build wizard) - @hadim
  • François Molino
  • Magali Suzanne


This project is distributed under the terms of the Modzilla Public Licence.


[monier2015apico]: Monier, B. et al. Apico-basal forces exerted by apoptotic cells drive epithelium folding. Nature 518, 245–248 (2015).

[Tamulonis2013]: Tamulonis, C. Cell-based models. (Universiteit ven Amsterdam, 2013). doi:10.1177/1745691612459060.

[Tlili2013]: Tlili,S. et al. Mechanical formalism for tissue dynamics. 6, 23 (2013).

[1]: The fact that the LCC model uses the term cell as it’s core concept is unfortunate. This will be hidden in the python API of the project.