Skip to main content

Outputs

The configuration generator produces configuration files and also prints an experiment list for downstream batch execution.

Configuration File Structure​

The generated configuration files are organized in the following structure:

configs/
└── {analysis_step}/          # Analysis step (preprocess/main/analysis)
    └── {experiment_set_name}/  # Experiment set name
        ├── {exp_id_1}/         # Experiment ID
        │   └── config.yaml
        ├── {exp_id_2}/
        │   └── config.yaml
        └── ...

Example Configuration Directory Structure​

For a main analysis experiment set named main_ft, the structure would look like:

configs/
└── main/
    └── main_ft/
        ├── D10_0/
        │   └── config.yaml
        ├── D10_1/
        │   └── config.yaml
        ├── D10_2/
        │   └── config.yaml
        └── D10_3/
        │   └── config.yaml

Experiment List for Bash Scripts​

The configuration generator outputs an experiment list for bash scripts. For example:

experiments=(
  "D10_0"
  "D10_1"
  "D10_2"
  "D10_3"
)

This text can be copied to the bash script to run the downstream experiments. Example bash scripts can be found in the launcher/ directory:

  • launcher/run_preprocess_ft.sh - For preprocessing experiments
  • launcher/run_main_ft.sh - For main analysis experiments
  • launcher/run_analysis_ft.sh - For downstream analysis experiments

Generated Configuration Files​

Each generated config.yaml file contains:

Complete Parameter Set​

  • All parameters from the corresponding template file
  • Experiment-specific values from the CSV design table
  • Proper type conversions (strings, integers, floats, booleans, lists)

Hierarchical Structure​

  • Nested YAML structure matching the template
  • Parameters organized by functional groups (data, channels, patching, etc.)
  • Consistent formatting and indentation

Example Generated Configuration​

exp_id: D10_0

data:
  file_name: /path/to/D10_tissue_0.ome.tiff
  antibodies_file: /path/to/antibodies.tsv
  image_mpp: 0.5
  generate_channel_stats: true

channels:
  nuclear_channel: DAPI
  wholecell_channel:
    - Pan-Cytokeratin
    - E-cadherin

patching:
  split_mode: full_image
  patch_height: -1
  patch_width: -1
  overlap: 0.1

visualization:
  visualize_whole_sample: false
  visualize_patches: true
  save_all_channel_patches: false

Output Validation​

The generator performs validation to ensure:

  • All required parameters are present
  • File paths reference existing files (when applicable)
  • Parameter types match expected formats
  • Experiment IDs are unique within the set
  • YAML syntax is valid and properly formatted