The input and output files and directories of all steinbock CLI commands can be controlled via command-line options. However, without explicitly specifying input/output arguments, the following typical directory structure is assumed when working with the steinbock CLI:
steinbock data/working directory | ├── raw (user-provided, when starting from raw data) | ├── img (user-provided, when not starting from raw data) ├── panel.csv (user-provided, when not starting from raw data) ├── images.csv | ├── ilastik_img ├── ilastik_crops ├── pixel_classifier.ilp ├── ilastik_probabilities | ├── cell_segmentation.cppipe ├── masks | ├── intensities ├── regionprops └── neighbors
Depending on the choice of preprocessing approaches, either the
raw directory containing the raw data, or the
img directory containing the images and a
panel.csv file must be provided by the user. All other files and directories are generated by the steinbock Docker container when following a supported workflow.