chemotaxis-cli

A command-line interface to measure chemotaxis index (CI) from chemotaxis assay images

Install

1. Install anaconda if its not already installed.

2. Create a new conda environment with python 2.7.12 and activate it. The env name used below chemotaxis_py2.7.12 is optional, name it as you like.

   # create a new conda environemnt
   conda create --name chemotaxis_py2.7.12 python=2.7.12 
   
   # activate it
   conda activate chemotaxis_py2.7.12

3. Clone the chemotaxis-cli repo to the directory of your choice.

   # navigate to the directory you want to clone the repo to
   cd <path_to_your_dir>
   
   # clone the chemotaxis-cli repository there
   git clone https://github.com/AndersenLab/chemotaxis-cli.git

4. Install the requirements and ct into your environment.

   # navigate into the repository directory
   cd <path_to_your_chemotaxis-cli_directory>
   
   # install the requirements with pip
   pip install -r requirements.txt
   
   # install ct
   python setup.py install

5. Run ct from within the chemotaxis_py2.7.12 env following example usage below. If you don’t plan to run ct its a good idea to deactivate your environment. See details of conda environment management here.

---

Example Usage

1. Activate the conda environment you installed ct into.

   # activate your chemotaxis env
   conda activate chemotaxis_py2.7.12

2. Navigate to the directory with your chemotaxis assay plate images, then run the ct command on images to calculate the chemotaxis index (CI). The command below can be edited to run on example images in the chemotaxis-cli repo you cloned in the install.

   # navigagte to your directory with images. You can use our examples if desired.
   cd <your_path_to_chemotaxis_cli>/examples/example1
   
   # Run ct on all .jpg files within a directory.
   ct *.jpg  --radius 920 --fp_sigma 1 --crop 20 --center 5 --small 100 --large 1200 --debug --header > results1.txt
   
   # Run on the second example with different parameters
   cd <your_path_to_chemotaxis_cli>/examples/example2
   ct *.jpg  --radius 781 --fp_sigma 2 --crop 20 --center 5 --small 100 --large 1200 --debug --header > results2.txt

3. Process the .txt file as you please. The output variables are defined below:

fname	q1	q2	q3	q4	n	total_q	total	ci
the file name without file extension	pixel count in top left quadrant	pixel count in top right quadrant	pixel count in bottom left quadrant	pixel count in bottom right quadrant	pixel count in origin (center)	sum of q1-q4	sum of all	chemotaxis index (CI)

ct arguments

• --radius sets the radius of n pixels used to find plates in the image. It can be helpful to measure this pixel value in imageJ for one image and provide it to ct so the program has an optimal starting radius for finding plates.
• --fp_sigma float, sets the standard deviation of the Gaussian filter used in canny edge detection when finding the plate center and radius. Higher values will find only the most obvious plate features. The default is 1.
• --crop will crop the plate radius by n pixels when creating the plate mask. This is useful if ct identifies the correct plate radius but the plate edges are identified as objects.
• --center sets the denominator value for the equation (plate radius / x), which is used to specify the radius of the origin (center) mask. The origin (center) mask is used to avoid counting nematode objects that have not migrated away from the origin (center) of the plate. The default --center value is 5.
• --small sets the pixel count used to filter out small objects.
• --large sets the pixel count used to filter out large objects.
• --debug will create many additional debug files that will help the user assess what objects are being counted and filtered.
• --header will attach a column header to output text file.
• --rev will reverse the sign of the chemotaxis index. This is helpful if the test and control compounds are switched from the default locations. Normally test compounds are in the top left and bottom right quadrants and controls are in the top right and bottom left quadrants.

Experimental workflow

1. Bleach synchronize strains to be assayed.
2. Use the COPAS BIOSORT with our custom chemotaxis plate holder to sort 50 synchronized L4 animals to the origin of chemotaxis assay plates.
   1. The CAD files and drawings required to fabricate our 3 part plate holder are here.
   2. The parts are assembled from bottom to top in the following order: base, support, then plates.
3. Spot control and test compounds onto the plates using a multi-channel pipet. The spot locations are etched into the chemotaxis plate holder for convenience.
4. If necessary, wick the M9 fluid used to dispense the nematodes to the center of the chemotaxis plates away using the corner of a kimwipe.
5. Once the dispensing fluid is removed allow the nematodes to respond to the compounds for 1 hour at 20°C.
6. After 1 hour, transfer the chemotaxis plates to 4°C prior to imaging.
7. Image the plates within 5 days, when imaging ensure that the edges of the plate are visible so that ct can setup the plate regions correctly.
8. Prior to analyzing images with ct, open one image in imageJ and manually measure the diameter of a plate in pixels.
   1. Select the straight line tool
   2. Draw a line across the diameter of the plate then click command+m to record a Length measurement. This is the diameter of the plate in pixels, divide by 2 to get the plate radius.
9. Run ct on your images and provide your plate radius --radius <your radius> as an argument.
10. Process the results