This code is associated with the paper from Verger et al., "A tension-adhesion feedback loop in plant epidermis". eLife, 2018. http://dx.doi.org/10.7554/eLife.34460

Cell Separation Image Ananlysis Pipeline

Copyright 2018 INRA - CNRS

File author(s): Stéphane Verger stephane.verger@ens-lyon.fr
File contributor(s): Guillaume Cerutti guillaume.cerutti@inria.fr

Distributed under the Cecill-C License.
See accompanying file LICENSE.txt or copy at
http://www.cecill.info/licences/Licence_CeCILL-C_V1-en.html

Please cite the original publication: Verger et al. (2018). The tension-adhesion feedback loop in plant epidermis.

Links: DOI:...
Github: https://github.com/sverger/Cell_separation_analysis

Description:

This python script allows the semi-automatic analysis of cell separations in 2D images, and can further perform comparaisons of mean cell separation area between two conditions or genotypes, as well as the analysis of cell separation orientation and anisotropy of multiple conditions or genotypes. This is a "semi-automatic" pipeline, because the analysis requires a preliminary manual step that needs to be be performed for each image before running the script: an appropriate threshold properly separating the cell signal from the cell separation background has to be defined manually. For more details see Verger et al. (2018). The tension-adhesion feedback loop in plant epidermis. This script was developed on Linux (Ubuntu 14.04) in the tissuelab environment of the OpenAleaLab platform (github.com/VirtualPlants/tissuelab, Cerutti G et al., (2017). FrontPlantSci 8:353. doi:10.3389/fpls.2017.00353). It is designed to run with python 2.7x and has been tested on Linux (Ubuntu 14.04), Mac (OSX...?) and windows (...?), using the "recommended" install with miniconda (see "install" below).

Prerequist: (e.g. using ImageJ)

For each image:

• From a raw confocal Z-stack, make a maximal intensity Z-projection to obtain a 2D image.
• If necessary enhance the contrasts.
• Smooth the image with a median filter to remove noise (Radius ~ 2 pixels).
• With the threshold tool in ImageJ, dertermine the appropriate threshold that separates the best the cell separations from the cells. Change the lut to "Grays" for easier visualization.
• Depending on the quality of the images it may be difficult or impossible to segment the cell separtations based on a threshold. If some of your images are in this situation, you may exclude them from your analysis. If most of your images are in this situation, this image analysis pipeline is not appropriate for your case.
• Then save the images in 8 bit .tif or .jpg and add "_xxxthld" at the end of the name (e.g. "sample_1_055thld.tif). The value before thld is the threshold value that will be used for the image segmentation (It has to be 3 digits).
• Then the file arborescence has to be organised as such: A "main" directory (updir), containing subdirectories for each condition/mutant, each containing all the images corresponding to the given condition/mutant.

Install:

You can either:

• Directly run the script if your python environment already has all the required dependencies (list below).

• Install the missing dependencies and run the script.

• (Recommended) Install miniconda and create a conda environment to run the script. For this, follow the instruction in the INSTALL.txt file.

Dependencies:

Designed to run on linux based operating systems.

• Python 2.7x https://www.python.org/
• Python modules:
  • matplotlib https://matplotlib.org/
  • nose http://nose.readthedocs.io/
  • numpy www.numpy.org/
  • pandas https://pandas.pydata.org/
  • pillow https://pillow.readthedocs.io/
  • pycircstat https://github.com/circstat/pycircstat
  • scipy https://www.scipy.org/

Settings:

• Download/open the script "Cell_separation_analysis.py"
• Before running the script, define the paramenters in the section of the script called "Parameters".
• Define a directory containing all the data to analyse and compare (updir).
• Define the pixel size of your image in micrometer (pixel_size).
• Define min and max area of crack to eliminate areas that are too small (min), and/or the background (max).
• Define the threshold type "min" or "max". "max" will detect and segment zones with lowest intensity of signal (black gaps between separated cells).
• Output global cracks size analysis? True or False. This is if you want to run the analysis of gap size comparing two mutants or conditions in the updir folder.
• Output global crack orientation analysis? True or False. This is if you want to run the analysis of gap orientation.
• Save the file to save the new parameters.

running the script:

In a terminal run

python 

or

ipython

depending on your installation.

In the python console, type

%run Cell_separation_analysis.py

This should run the script, write some output in the python console and create the output images and files.

Output:

• For each image, a .csv file is created containing for each segmented cell separation, the label number, center position, area in pixels and um square, the principal angle (orientation of the separation), the anisotropy, the eigen values and vectors and the standard deviations.
• Also for each image, an inverted version of the image, overlaid with the segmented areas as well as a representation of the anisotropy and principal angle for each area, is saved as a vectorial .pdf.
• Finally, for each image, a polar histogram representing the distribution of cell separation orientation in the image is created and saved as a vectorial .pdf.
• If running the "Global_Output_Size", statistical tests will be run on the compared samples, and the summary of these tests will be saved in a .txt file in the "updir", as well as a box plot as a vectorial .pdf.
• If running the "Global_Polarhist_Output", a global polar histogarm will be created for each condition/mutant and saved as a vectorial .pdf and the circular mean angle, the resultant vector length and the mean anisotropy will be computed and saved in a summary .txt file.