Home

Macrophage Analysis Wiki

This is a short overview on the scripts used for. The analysis tools are written in Python 3.6, the following python modules need to be installed:

• numpy
• seaborn
• matplotlib
• pandas as pd
• scipy
• skimage

Usage

The workflow is shown in the diagram above. As input the following files need to be provided

Metafile

mouse_name	type	analysis_file_vessels	analysis_file_macrophages	x_resolution [um]	y_resolution [um]	vessel_channel	macrophage_channel
mouse_1	4 weeks	vessel_file1.tif	macrophages_1.csv	400	400	0	1
mouse_1	4 weeks	vessel_file2.tif	macrophages_2.csv	400	400	0	1
mouse_2	2 weeks	vessel_file3.tif	macrophages_3.csv	300	300	1	0

Series of vessel images

Binary images containing segemented vessels

X-Y positions of macrophages

	X	Y
1	225	37
1	186	220
1	160	279

Parameter file

The parameter file has the following content

    "data_directory": "/directory/to/your/data/",
    "output_directory": "/directory/to/your/output/",
    "meta_data_file" : "meta_file",
    "subfolder_vessel_images" : "subfolder/",
    "type_column" : "tumor_type",
    "macrophage_channel"  : ["macrophage_channel1", "macrophage_channel2"],
    "macrophage_position_files"   : ["macrophage_channel1_files","macrophage_channel2_files"],  
    "selected_feature" : ["distance_vessels"],
    "occupancy_radius" : 10.0,
    "occupancy_sampling" : 10000

Here you can adjust folder names and channel names etc.

Features

The following features can be calculated

1. nearest distance of macrophages to the vessels
2. nearest distance of macrophages to thick vessels (here thick is defined by a threshold on the vessel area)
3. nearest distance of macrophages to thin vessels (here thin is defined by a threshold on the vessel area)
4. normalized distance of macrophages to the vessels (here the distances are normalized to the distances that would stem from a completely random distribution)
5. occupancy of vessels
6. n-nearest neighbor distances among macrophages

Acknowledgements