Appendix A — Testing stress granules segmentation

This notebook tests the stress granules segmentation algorithm that uses difference of gaussians (DoG) implemented in scikit-image filters.

A.1 Granule properties

	area
count	472
mean	1.89066
std	1.19623
min	0.486
25%	0.9639
50%	1.6848
75%	2.5353
max	11.502

A.2 Granule images

Images of the cytoplasm, filtered cytoplasm and granules identified by DoG.

# Testing stress granules segmentation This notebook tests the stress granules segmentation algorithm that uses difference of gaussians (DoG) implemented in scikit-image filters. ```{python} from glob import glob import imageio.v3 as iio import skimage as ski import numpy as np import matplotlib.pyplot as plt import re from os.path import basename import pandas as pd from IPython.display import display, HTML, Markdown ``` ```{python} PIXEL_SIZE = 0.18 ``` ```{python} ims_cyto = [] ims_cyto_filtered = [] ims_granules = [] granule_data = [] for im_fn in glob('data/ome-tiff/*/*tif')[30:40]: im = iio.imread(im_fn) # maximum intensity projection im_nuclei = im[0].max(axis=0) im_cytoplasm = im[1].max(axis=0) # Process cytoplasm and find granules im_cytoplasm_filtered = ski.filters.difference_of_gaussians(im_cytoplasm, 3) threshold = 0.0009 # we use a fixed global threshould because all images were collected at the same intensity im_granules = im_cytoplasm_filtered > threshold im_granules = ski.morphology.remove_small_objects(im_granules, 15) im_granules_labeled = ski.measure.label(im_granules) granule_props = ski.measure.regionprops_table(im_granules_labeled, properties=['label', 'area', 'centroid', 'eccentricity', 'perimeter']) match = re.search(r'imagem(\d+)', im_fn) image_number = match.group(1) if match else 'unknown' granule_props['area'] = granule_props['area'] * PIXEL_SIZE * PIXEL_SIZE granule_props['perimeter'] = granule_props['perimeter'] * PIXEL_SIZE granule_props['image_number'] = image_number granule_data.append(pd.DataFrame(granule_props)) ims_cyto.append(im_cytoplasm) ims_cyto_filtered.append(im_cytoplasm_filtered) ims_granules.append(im_granules) granules_df = pd.concat(granule_data, ignore_index=True) ``` ## Granule properties ```{python} Markdown(granules_df['area'].describe().to_markdown()) ``` ## Granule images ```{python} #|fig-cap: "Images of the cytoplasm, filtered cytoplasm and granules identified by DoG." fig, axs = plt.subplots(len(ims_cyto), 3, constrained_layout=True, figsize=(6, 2*len(ims_cyto))) for idx in range(len(ims_cyto)): axs[idx,0].imshow(ims_cyto[idx], cmap='inferno', vmax=10) axs[idx,1].imshow(ims_cyto_filtered[idx], vmin=0, vmax=0.003, cmap='magma') axs[idx,2].imshow(ims_cyto[idx], cmap='inferno', vmax=10) axs[idx,2].contour(ims_granules[idx], colors='white', linewidths=1.0) axs[idx,0].axis('off') axs[idx,1].axis('off') axs[idx,2].axis('off') ```