tyssue : An epithelium simulation library

[ANN] Working on a 1.0 release

Dear tyssue user or prospective user, I come with good bearings.

This commit to the **new default branch main is the official start of the 1.0 release. It is long overdue and there is a lot of house keeping.

Short term

• fix CI w/ github actions
• Readthedocs with mkdoc
• Tests
• Notebook tests
• Merge PRs #263 #274
• Tests
• Check Polarization notebooks

---

Name	Downloads	Version	Platforms

tyssue is now published in the Journal of Open Source Software!

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.

Overview

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.

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.

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 import Sheet
## Simple 2D geometry
from tyssue import PlanarGeometry
## Visualisation (matplotlib based)
from tyssue.draw import sheet_view

sheet = Sheet.planar_sheet_2d('basic2D', nx=6, ny=7,
                              distx=1, disty=1)
PlanarGeometry.update_all(sheet)
sheet.sanitize()
fig, ax = sheet_view(sheet)

Features

• Easy data manipulation.
• Multiple geometries (Sheets in 2D and 3D, monolayers, bulk).
• Easy to extend.
• 2D (matplotlib) and 3D (ipyvolume) customisable visualisation.
• Easy quasistatic model definition.
• Self collision detection. new in 0.3

Documentation

• The documentation is now browsable on tyssue.io
• The old documentation is still browsable online here
• Introduction notebooks are available here.

Mailing list:

[email protected] - https://framalistes.org/sympa/info/tyssue

Subscribe ➙ https://framalistes.org/sympa/subscribe/tyssue Unsubscribe ➙ https://framalistes.org/sympa/sigrequest/tyssue

Authors

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

Dependencies

As all the dependencies are already completely supported in python 3.x, we won't be maintaining a python 2.x version, because it's time to move on...

Core

• CGAL > 4.7
• Python >= 3.6
• numpy
• scipy
• matplotlib
• pandas
• pytables
• jupyter
• notebook
• quantities
• ipywidgets
• pythreejs
• ipyvolume
• vispy

Tests

• pytest
• coverage
• pytest-cov

Install

You can install the library with the conda package manager

conda install -c conda-forge tyssue

Through PyPi

You can also install tyssue from PyPi, this is a CGAL-less version (pure python), lacking some features:

python -m pip install --user --upgrade tyssue

From source

See INSTALL.md for a step by step install, including the necessary python environment.

Licence

Since version 0.3, this project is distributed under the terms of the General Public Licence.

Versions 2.4 and earlier were distributed under the Mozilla Public Licence.

If GPL licencing is too restrictive for your intended usage, please contact the maintainer.

Bibliography

• There is a Bibtex file here with collected relevant publications.

The tyssue library stemed from a refactoring of the leg-joint code used in [monier2015apico].

[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).

Research notice

Please note that this repository is participating in a study into sustainability of open source projects. Data will be gathered about this repository for approximately the next 12 months, starting from June 2021.

Data collected will include number of contributors, number of PRs, time taken to close/merge these PRs, and issues closed.

For more information, please visit our informational page or download our participant information sheet.