/chemokine_buffering_paper

Code for python-based quantitative image and data analysis in "Dynamic buffering of extracellular chemokine enables robust adaptation during directed tissue migration" by Wong and colleagues.

Primary LanguageJupyter NotebookMIT LicenseMIT

chemokine_buffering_paper

This repo hosts the code for the python-based quantitative image and data analysis used in the paper entitled "Dynamic buffering of extracellular chemokine enables robust adaptation during directed tissue migration" by Wong and colleagues, as described in the METHOD DETAILS chapter in the sections High resolution imaging and processing and 3D image analysis and quantification. All code was written by Jonas Hartmann.

Analyses Included

  • Intensity analysis (as shown in figures 3J and 3L)
  • Point cloud analysis (as shown in figures 2L, 4B, 6D)
  • Colocalization analysis (as shown in figures 5J and 6N)

Code Structure

  • The code in this repo is structured in three 'layers':
    • Modules (quant, coloc, util) contain refactored functions, mainly for the RUN notebooks
    • RUN notebooks are pipelines built on the modules; they take in images and output quantitative measures
    • ANALYSIS notebooks take in quantitative measures and produce various plots and statistics

Data Availability

This repository only hosts a couple of example images (in data_ex) to help follow through the RUN notebooks. These examples have been zipped to comply with GitHub's 100MB maximum file size policy; they need to be unzipped before use.

The ANALYSIS notebooks are set up to reproduce the analyses presented in the paper and expect the full data to be present in a directory named data_full. This large dataset is available only on request via the Lead Contact, Darren Gilmour (Email DG).

Workflow

  1. Convert raw images to 8bit .tif images using macro_8bit.ijm in Fiji with a suitable intensity range (logged in metadata.xlsx).
    • The example data provided in data_ex has already been converted.
    • These example images must be unzipped before use and the resulting .tif files must be located directly in the data_ex directory.
  2. Manually identify the coordinates of the tissue's apical focus point, note them in metadata.xlsx and save the file as a tab-separated text file called metadata.txt.
    • For the example data provided and the study's full dataset, coordinates and metadata.txt are already provided.
  3. For both point cloud and intensity analyses, run RUN_preprocessing.ipynb to mask the overall tissue.
  4. For intensity analysis, run RUN_intensity_quantification.ipynb.
  5. For point cloud analysis, run RUN_landmark_extraction.ipynb.
  6. For colocalization analysis, run RUN_colocalization.ipynb.
  7. Use the the various ANALYSIS notebooks to visualize and analzye the respective results.

Dependencies

  • Python 2.7.13 (we recommend the Anaconda distribution)
  • Scientific python stack including numpy, scipy, scikit-image, matplotlib
    • The versions used by us were numpy 1.11.3, scipy 2.0, scikit-image 0.13.0, and matplotlib 1.5.1
  • The tifffile module (can be installed from Anaconda cloud)

Contact and Support

  • The study's corresponding authors are Mie Wong (Email MW) and Darren Gilmour (Email DG).
  • For questions regarding the code in this repository, contact Jonas Hartmann (Email JH) or open an issue on GitHub. Note that we cannot promise support for any use cases other than direct reproduction of the study's results.
  • To request the full data or any other materials, resources or reagents, please get in touch with the study's Lead Contact, Darren Gilmour (Email DG).