Automated Neuronal Differentiation Analyzer (ANDA) is an image analysis tool used for analysis of microscopy images from 2d neuronal cell cultures. The pipeline is a series of Python scripts executed in succession by the workflow manager Snakemake. Important metrics in neurodevelopment, namely cell bodies, neurites and neurite attachment points are retrieved from images and analysed using Fiji.
All necessary Python packages can be installed when installing Anaconda.
Download ANDA:
git clone https://github.com/EskelandLab/ANDA.gitBefore using ANDA, be sure to have back up of all images. Also ensure that you have enough space to run the analysis as the analysis creates output files that require some space. At least twice the space used by the images should be available.
Place all images in a folder separate from the ANDA directory.
From the terminal and while in the ANDA directory, start the Bash script by typing "bash ANDA.sh" or "./ ANDA.sh" and hit Enter.
In the next window set the analysis parameters by filling out the required information (figure 1).
- Choose the directory with the images as image directory.
- Select the full path to the ImageJ program. Usually this is /home/.../Fiji.app/
- Set neurite aspect ratio threshold for exclusion of false positive neurites. Set this value to zero if you want to include every identified object.
- Select analysis metrics and cell line.
- Select if you want to save motif outlines or not.
- Start the analysis by pressing "Save and quit".
Figure 1: TkInter graphical user interface for selecting parameters for image analysis.
The tool will now run through the images and apply the analysis metric(s) you chose. Depending on the system you are running on, and the number of images you are analysing this will take a while. To abort press "Ctrl+C".
When the analysis has finished, the results are stored in csv files named "results_{metric}.csv" in the same directory as the image subfolders. Note that there are also csv files stored in the directories with names ending with "results". These csv files contain unsummarized data on every identified particle in every image.
-
When running or restarting the analysis by running ANDA.sh, an error message regarding already existing folders might occur, causing the pipeline to halt. To overcome this, either delete the folders in the error message or comment lines 32-39 in the ANDA.sh file.
-
If Snakemake skips a rule and outputs an error about missing files in rule 'data_sort', re-do the analysis by running this:
snakemake --snakefile data_snakefile. Snakemake will not restart the analysis all over, but continue from where it stopped previously. Another option is to run the analysis using fewer cores by using the --cores N or -c N parameter.