This application was created to bulk process and analyze batches of monochromatic images providing localization (masks), intensities, and similar quantifying values (markers), including co-localizations of multichannel monochromatic immunofluorescent (or IHC, etc.) images. It creates both image and value (excel) outputs.

It was originally developed for use in:

J. Franco-Barraza et.al 2017. eLife https://dx.doi.org/10.7554/eLife.20600 Cukierman Lab Fox Chase Cancer Center, Philadelphia, PA 19111.

You can install a Windows executable for the latest version of SMIA at https://github.com/cukie/SMIA/releases/tag/2.1.0 If you want to build from source, or are on a platform other than windows, download or clone the source code and run...

python smia/setup.py
python smia/smia_ui.py

Note that to build from source, you'll need Python 2.7 and a C++ compiler. The C++ compiler for Windows can be found at https://www.microsoft.com/en-us/download/details.aspx?id=44266

To run the SMIA user interface, run the executable or from source:

python smia.run_smia_cukie config_file.txt

A script and ui to aid in making batches can also be downloaded and run as a Windows executable. This was developed specifically for the Cukierman lab, and will probably be of no use to others. To run from source...

python batch_maker_ui.py

If downloading the source, unit tests can be found in the tests package. Some sample batches and sample images can be found in tests/test_fixtures/test_images You can run them by editing tests/test_fixtures/test_images_config.config to point to the base_dir and output_to the correct paths on your system. From the command line...

python smia/run_smia_cukie tests/test_fixtures/test_images_config.config

The base directory for images will be a folder that holds all other folders. Each folder is a "batch" and a typical file structure for analysis must be in the following format:

Base_Directory
	-> Batch1_Directory
		-> Image1
		-> Image2
		-> Image3
		-> Image4
		-> Image5

	-> Batch2_Directory
        -> Image1
        -> Image2
        -> Image3
        -> Image4
        -> Image5

	-> More_Batches as needed...
        -> etc...

1. mask and marker names unique
2. mask and marker names must not be a subset of each other
3. each mask or marker name must correspond uniquely to an image name in each batch
4. Note that the names of images will be identified by the image file name and only characters after the final ] character will be included. ] is not a necessary component of the image name. This is to keep compatibility with the format of the imaging mechanism first used for the original study.

For every run of the application, you must first create a configuration file. Configuration files are just plain JSON. A sample config file can be seen in sample_config.txt

base_directory: 
	A string representing the absolute path of the top level 
	directory containing all the batches of images.
	NOTE in Windows you will need to use double backslashes as escapes.
	E.g. Windows - "C:\\path\\to\\dir\\here"
	*nix - "/path/to/dir/here"

marker_names:
    A list of all markers in the following format: [file prefix, desired name, threshold]
    file prefix - a unique identifier for the name of this specific image frame. See restriction on names above.
    desired name - the name you would like to use to identify the file. This is the name that will represent the marker in your results. 
    threshold - a threshold value for your marker where a valid pixel is defined as having a value greater than or equal to the threshold. 

mask_names:
	A list of mask names defined the same way as marker names

num_markers: 
	integer value of number of markers (this will be removed soon)

num_masks:
	integer value of number of masks (this will be removed soon)

num_pictures:
	integer value of total number of images per batch (this will be removed soon)

overlay_white_list:
	a list of strings corresponding to the operations you would like the application to perform. They must all be of the form "singlemarker under however, many, masks" In order to denote the inverse of a mask or marker, simply prepend the word "NOT" to the front of the name. 
	E.g. if we have a marker called nuclei, then its inverse would be NOTnuclei. (The name of this parameter is weird and will change in the future)

output_to:
	A string representing the absolute path of where you wish to output all results (excel file, copy of configuration file used, thumbnails, etc.)

For general inquiries about the papaer please contact Dr. Edna Cukierman. https://www.foxchase.org/edna-cukierman Email: [email protected]

If you find issues with the software, please file a github issue.