Running liv_zones

Before running the liv_zones processing it is important to make sure that your images are in the correct format and that your folders are organized correctly

  • Images are expected to be saved as .tif files, although we are working on also supporting .zarr

  • Images channels are expected to be organized as follows:

    1. Actin

    2. Mitochondria

    3. Lipid Droplets

    • additional channels are allowed, but they will not be processed

  • There must be a seperate folder to store the results of each input image

    • outputs are saved with pre-defined names and will override previous outputs if that are any already in the designated folder

    • I personally recommend creating a seperate folder for each image, then storing the original image as well as any outputs within

  • The liv_zones analysis also requires 2 other inputs, saved in the output folder:

    • cv_distnace.tif: a binary image segmentation of the central vein, where the central vein is marked by 0s

    • pv_distance.tif: a binary image segmentation of the protal vein, where the portal vein marked by 0s

    • these are used to caclulate the realtive position of each organelle and cell along the ascinus

    • lastly, it is important that these are saved with these exact names, otherwise the program will not recognize them

In order to run the liv_zones processing, we use the run.py script

  1. Open run.py in any text editor

from liv_zones import preprocess as pre
from liv_zones import organelle as org
from liv_zones import cell as c
from liv_zones.ascini import plot_properties, plot_cell, plot_ascinus_annotated

  1. Next we specify the scale and the needed file paths

    • scale: the number of pixels per micron in the image, easiest to find using FIJI

    • image_paths: a list of file paths to specific images for processing

    • save_paths: a list of fie paths to the folders where results for each image are saved

    • the image_paths and save_paths lists are paired, ie. the results of the first image will be saved to the first save_path and so on…

    • double check to make sure that both lists are the same length

scale =  14.4024 # pixels per micron
image_file_paths = [ 'path/to/first/image.tif',
                     'path/to/second/image.tif'
                     ]

save_file_paths = [ 'path/to/first/folder',
                    'path/to/second/folder',
                    ]

  1. Next we specify which analyses would like to run

  1. run_preprocessing: True or False, If True, will check to make sure that all the files needed for the analysis are present in the specific save_path folder, and will create any that are missing. The necessary files are:

  • cell_mask.npy: an instance segmentation mask of all cells in the image

  • mito_mask.npy: an instance segmentation mask of all mitochondria in the image

  • lipid_droplet_mask.npy: an instance segmentation mask of all lipid droplets in the image

  • cv_distance.npy: a distance transform from the central vein.

  • pv_distance.npy: a distance transform from the portal vein

  • boundry_distance.npy: a distance transform from the boundry of each cell

  • It is important that each of these files are named exactly as stated above, if not the program will not recognize them

# Do you want to run the preprocessing?
run_preprocessing = False

# comment out any features that you don't want re-calculated
feature_list = [
  'cell_mask',
  'mito_mask',
  'lipid_droplet_mask',
  'cv_distance',
  'pv_distance',
  'boundry_distance'
  ]

  1. organelle_features: True or False, If True, will extract features of the organelles provided in organelle_list and save them in a csv file

  • organelle_list: possible options are mitochondria or lipid_droplets, comment out any that you do not wish to calculate

#  Do you want to extract individual organelle features?
organelle_features = False

# comment out any organelles you dont want re-calculated
organelle_list = [
   'mitochondria',
   'lipid_droplets',
   ]

  1. Visualization options:

  • plot_labled_ascinus: True or False: If True save an image of the ascinus with each cell labeled with its ID #

  • plot_props: True or False: If True generate plots of average properties per cell as a function of ascinus position. These plots will be saved in a sub-folder ascini_trends

  • show_individual_cell: True or False:

# Do you want an image of the ascinus with each cell labeled?
plot_labeled_ascinus = False

# Do you want to plot per cell properties over the length of the ascinus
plot_props = False

# Do you want to visualize a specific cell?
show_individual_cell = False

# the cell number corresponding to the labeled ascinus
cell_number = 10

  1. Run the analysis by calling saving your changes and running run.py from the terminal

python run.py