TABLE 1.
Code used in the step-by-step protocol for establishing a container using a definition file. Please note that copying some code directly from the table might induce errors due to formatting. Hence, we suggest to type the commands newly in case of errors, and refer to the files on github: https://github.com/DaetwylerStephan/Containerize_ImageAnalysis or in the Supplementary Material.
| Code | Project title |
|---|---|
| Code 2 | > mkdir container_example/ |
| > cd container_example/ | |
| Code 3 | > vagrant init sylabs/singularity-3.7-ubuntu-bionic64 |
| Code 4 | > vim Vagrantfile |
| Code 5a | config.vm.synced_folder "C:/your/local/cellimages/filepath", "/home/vagrant/cellpose_testdata" |
| Code 5b | config.vm.provider "virtualbox" do |vb| #Customize the amount of memory on the VM: vb.memory = "9000" end |
| Code 6 | > vagrant up |
| Code 7 | > vagrant ssh-config |
| Code 8 | > scp -P 2222 python_main.py vagrant@127.0.0.1:. |
| Code 9 | > vagrant ssh |
| Code 10 | vagrant@vagrant> sudo apt-get update |
| vagrant@vagrant> sudo apt-get install vimvagrant@vagrant> vim cellpose_container.recipe | |
| Code 11 | Bootstrap: docker From: ubuntu:18.04 |
| Code 12 | %labels AUTHOR: Shilpita Mitra-Behura, Reto Fiolka, Stephan Daetwyler |
| Code 13 | %files /home/vagrant/python_main.py /mnt |
| Code 14 | %post #Downloads the latest package lists. apt-get update -y |
| #Install python and other tools. #Non-interactive is used to ensure prompts are not needed. DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \ python3 python3-pip\ python3-setuptools | |
| #Update pip python3 -m pip install --upgrade pip==21.2.4 | |
| #Install python libraries needed for Cellpose to run python3 -m pip install wheel==0.37 python3 -m pip install --no-cache-dir torch==1.8.1 python3 -m pip install opencv-python-headless==4.5.3.56 python3 -m pip install cellpose==0.6.5 | |
| Code 15 | %runscript |
| exec /usr/bin/python3 /mnt/python_main.py “$@” | |
| Code 16 | %help This is a Singularity container to segment images using Cellpose. To run this container, use, for example, 'singularity run cellpose_container.sif --filedir /folder/where/files/to/segment/are --savedir /folder/where/segmented/files/should/be/saved --pretrained_model cyto --chan 2 --save_tif' |
| The user must put the folder where the files to be segmented are after the flag “--filedir.” The directory where the user wants to save their files is optional, with the tag “--savedir.” The flag “--pretrained_model” is required and must have either the input “cyto” or “nuclei.” The flags “--chan” and “--chan2” are optional and allude to the channels of the image that will be segmented. The flag “--diameter” is optional and can be added for the user to specify the diameter of the nuclei in image. The flags “--flow_threshold”, “--cellprob_threshold” are further optional flags to specify cellpose parameters. The flags “--save_png” and “--save_tif” are used to denote what file type the segmented files are saved as. The default is to save as a Tiff. | |
| Code 17 | vagrant@vagrant> sudo singularity build cellpose_container.sif cellpose_container.recipe |
| Code 18 | vagrant@vagrant> singularity run cellpose_container.sif --filedir /home/vagrant/cellpose_testdata --savedir /home/vagrant/cellpose_testdata/save_segmentedimages --pretrained_model nuclei --chan 1 --flow_threshold 0 --cellprob_threshold −1 --diameter 19 --save_tif |
| Code 19 | > scp -P 2222 vagrant@127.0.0.1:/home/vagrant/cellpose_container.sif . |
| Code 20 | > sbatch run_on_cluster.sh |