Home
This is a short overview on the scripts used for analyse spatial patterns of macrophage distributions. 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
The workflow is shown in the diagram above. As input the following files need to be provided
| 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 |
Binary images containing segemented vessels
| X | Y | |
|---|---|---|
| 1 | 225 | 37 |
| 1 | 186 | 220 |
| 1 | 160 | 279 |
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.
The following features can be calculated
- nearest distance of macrophages to the vessels
- nearest distance of macrophages to thick vessels (here thick is defined by a threshold on the vessel area)
- nearest distance of macrophages to thin vessels (here thin is defined by a threshold on the vessel area)
- normalized distance of macrophages to the vessels (here the distances are normalized to the distances that would stem from a completely random distribution)
- occupancy of vessels
- n-nearest neighbor distances among macrophages
