Ariadne

🌱 is a small software package for analyzing images of Arabidopsis thaliana roots.

📷 It features a GUI for semi-automated image segmentation

⏰ with support for time-series GIFs

☠️ that creates dynamic 2D skeleton graphs of the root system architecture (RSA).

🔍 It's designed specifically to handle complex, messy, and highly-branched root systems well — the same situations in which current methods fail.

📊 It also includes some (very cool) algorithms for analyzing those skeletons, which were mostly developed by other (very cool) people^1,2. The focus is on measuring cost-performance trade-offs and Pareto optimality in RSA networks.

⚠️ This is very much a work-in-progress! These are custom scripts written for an ongoing research project — so all code is provided as-is.

🔨 That said, if you're interested in tinkering with the code, enjoy! PRs are always welcome. And please reach out with any comments, ideas, suggestions, or feedback.

Installation

Ariadne is installed as a Python package called ariadne-roots. We recommend using a package manager and creating an isolated environment for ariadne-roots and its dependencies. You can find the latest version of ariadne-roots on the Releases page.

Installation (Users)

We recommend installing Ariadne in an isolated environment using uv. You can install it with uv to keep your environment clean and reproducible. There are two main ways to install and run Ariadne:

---

Option 1. Local Environment

This creates a local .venv folder to hold Ariadne and its dependencies.

# Create a local environment with pip + setuptools pre-seeded
uv venv --seed .venv

# Activate it (Linux/macOS)
source .venv/bin/activate

# Activate it (Windows PowerShell)
.venv\Scripts\activate

# Install Ariadne
uv pip install ariadne-roots

Then run the GUI:

ariadne-trace

---

Option 2. One-liner Install & Run (no manual environment necessary)

You can also run Ariadne directly with uvx, which installs it into an isolated cache and exposes the CLI:

uvx ariadne-trace

This will launch the GUI without needing to set up or activate a venv manually.

---

Usage

If installed in a local environment

Activate the environment and run:

source .venv/bin/activate    # or .venv\Scripts\activate on Windows
ariadne-trace

If using the one-liner

Simply run:

uvx ariadne-trace

conda environment installation

Follow the instructions to install Miniforge3.

Step-by-Step Installation

1. Create an isolated environment:

   mamba create --name ariadne python=3.11

2. Activate your environment:

   mamba activate ariadne

3. Install ariadne-roots using pip:

   pip install --pre ariadne-roots  # Use --pre to include pre-release versions

   • Omit the --pre flag if you only want to install stable releases.

Usage

1. Activate your environment:

   mamba activate ariadne

2. Open the GUI:

   ariadne-trace

Trace with Ariadne

1. Click on “Trace” to trace roots.

2. The following window should open:

   

3. Click on “Import image file” and select the image to trace the roots.

4. Select the zoom factor for your image

   • A window should popup asking for the zoom factor
   • Use the "Zoom in" and Zoom out" button to adjust the zoom needed to trace the root system with high precision
   • Click on "OK"
   • After this, any new image imported will be opened with the identical zoom factor
   • Remark: after closing Ariadne, the zoom factor will be canceled. Therefore, take note of the zoom factor used and reapply the same everytime when restarting Ariadne.

5. Trace the first root:

   • Start tracing the entire primary root first (it should appear green).
   • To save time, place a dot on each region where a lateral root is emitted.

6. Save the traced root:

   • When the first root is fully traced, click on the “Save” button on the left-hand menu of Ariadne or press “g” on your keyboard.
   • A new window will pop up asking for the plant ID. Tap any letter “A” or number "1".
     • Each time you click on “Save”, a .json file will be saved in the folder at the location of Location_1 (see above).

7. Trace additional roots:

   • When you are done tracing the first root, click on the “Change root” button on the left-hand menu of Ariadne.
   • Select a new plant ID, like “B”, to trace the second root.
   • Continue tracing each root on your image following these steps.

8. Finish tracing:

   • When you have traced all roots on your image, click on “Change root” and repeat from “Step 3” above for any new images.

Analyze with Ariadne

1. Organize your files:

   • Gather all the .json files stored at the location where Ariadne has been installed into a new folder named “OUTPUT_JSON” (referred to as “location_1” later on).
   • Create a folder named “RESULTS” (referred to as “location_2”).
   • Create a new folder named “Output”.

2. Prepare for analysis:

   • Close Ariadne but keep the terminal open.
   • Follow the instructions in step 2 above to set up the terminal.

3. Run the analysis:

   • Click on “Analyze” in Ariadne.
   
   • Select the .json files to analyze from “location_1”.
   • Then select “location_2” for the output.
   • The software will analyze all the selected .json files.

Results

• In the “location_3” folder, you will find:
  • A graph for each root showing the Pareto optimality.
  • A .csv file storing all the RSA traits for each root.

The RSA traits included in the CSV are

• Total root length: Total root length
• Travel distance: Sum of the length from the hypocotyl to each root tip (Pareto related trait)
• Alpha: Trade-off value between growth and transport efficiency (Pareto related trait)
• Scaling distance to front: Pareto optimality value (Pareto related trait)
• Total root length (random): Random total root length
• Travel distance (random): Random sum of the length from the hypocotyl to each root tip (Pareto related trait)
• Alpha (random): Random trade-off value between growth and transport efficiency (Pareto related trait)
• Scaling distance to front (random): Random Pareto optimality value (Pareto related trait)
• PR length: Length of the primary root
• PR minimal length: Euclidean distance from the hypocotyl to the primary root tip
• Basal zone length: length from the hypocotyl to the insertion of the first lateral root along the primary root
• Branched zone length: length from the insertion of the first lateral root to the insertion of the last lateral root along the primary root
• Apical zone length: length from the last lateral root to the root tip along the primary root
• Mean LR lengths: Average length of all lateral roots
• Mean LR minimal distances: Average Euclidean distance between each lateral root tip and its insertion on the primary root for all lateral roots
• Median LR lengths: Median length of all lateral roots
• Median LR minimal distances: Median Euclidean distance between each lateral root tip and its insertion on the primary root for all lateral roots
• Sum LR minimal distances: Sum of the Euclidean distances between each lateral root tip and its insertion on the primary root for all lateral roots
• Mean LR angles: Average lateral root set point angles
• Median LR angles: Median lateral root set point angles
• LR count: Number of lateral roots
• LR density: Number of lateral roots divided by primary root length
• Branched zone density: Number of lateral roots divided by Branched zone length
• LR lengths: Length of each individual lateral root
• LR angles: Lateral root set point angle of each individual lateral root
• LR minimal distance: Euclidean distance between each lateral root tip and its insertion on the primary root for each lateral root
• Barycentre x displacement: Vertical distance between the hypocotyl base to the barycenter of the convex hull
• Barycentre y displacement: Horizontal distance between the hypocotyl base to the barycenter of the convex hull
• Total minimal distance: Sum of LR minimal distances plus PR minimal length
• Tortuosity (Material/Total Distance Ratio): Total root length divided by total minimal distance

Keybinds

• Left-click: place/select node.
• Ctrl: Hold Ctrl to scroll through the image with the mouth
• t: toggle skeleton visibility (default: on)
• e: next frame (GIFs only)
• q: previous frame (GIFs only)
• r: toggle proximity override. By default, clicking on or near an existing node will select it. When this override is on, a new node will be placed instead. Useful for finer control in crowded areas (default: off)
• i: toggle insertion mode. By default, new nodes extend a branch (i.e., have a degree of 1). Alternatively, use insertion mode to intercalate a new node between 2 existing ones. Useful for handling emering lateral roots in regions you have already segmented (default: off)
• g: Save output file
• d: Delete currently selected node(s)
• c: Erase the current tree and ask for a new plant ID
• +: Zoom in
• -: Zoom out
• Ctrl-Z: Undo last action

Contributing

Follow these steps to set up your development environment and start making contributions to the project.

1. Navigate to the desired directory

   Change directories to where you would like the repository to be downloaded

   cd /path/on/computer/for/repos

2. Clone the repository

   git clone https//github.com/Salk-Harnessing-Plants-Initiative/Ariadne.git

3. Navigate to the root of the cloned repository

   cd Ariadne

🛠️ For Developers

Requirements

• uv for dependency management
• Python 3.11+

Setting Up a Development Environment

Clone the repository:

git clone https://github.com/Salk-Harnessing-Plants-Initiative/Ariadne.git
cd Ariadne

🛠️ Development with uv

We use uv for dependency management and tooling.

• This workflow is tested in github actions using .github\workflows\test-dev.yml.

There are two main commands you’ll use all the time:

1. uv sync

This sets up (or updates) your project environment.

uv sync

Alternatively, you can use uv pip install if you are working in an isolated environment and want to pip install in editable mode with dev dependencies.

uv pip install -e ".[dev]"

• Creates .venv (if it doesn’t exist).
• Installs your runtime dependencies plus the dev group (tests, linters, etc.) from pyproject.toml.
• You usually only run this when first cloning the repo, or after editing pyproject.toml.

---

2. uv run

This runs commands inside the project environment without needing to manually activate .venv.

Examples:

• Run the test suite with coverage:

  uv run pytest --cov=ariadne_roots --cov-report=term-missing

• Check code formatting:

  uv run black --check .

• Run the CLI:

  uv run ariadne-trace

---

Do I need both?

• Yes, but at different times:
  • Use uv sync when you need to install/update dependencies.
  • Use uv run whenever you want to execute something inside that environment.

You could also source .venv/bin/activate (or .\.venv\Scripts\activate on Windows) and then run pytest, black, etc. directly. But uv run is cross-platform and doesn’t require activation, which makes it ideal for CI and scripts.

---

3. Building artifacts

To build source and wheel distributions:

uv build

Artifacts will be created in the dist/ directory.

---

Alternative: install dev extras with pip

If you’re not using uv, you can still install everything with pip:

pip install -e ".[dev]"

This installs runtime + dev dependencies into your current environment.

---

Instructions for conda environment

1. Create a development environment

   This will install the necessary dependencies and the ariadne-roots package in editable mode

   mamba create --name ariadne_dev python=3.11 # python 3.11, 3.12, 3.13 are tested in the CI

2. Activate the development environment

   mamba activate ariadne_dev

3. Install dev dependencies and source code in editable mode

   pip install -e ".[dev]"

Development Rules

1. Create a branch for your changes

   Before making any changes, create a new branch

   git checkout -b your-branch-name

2. Code

   Make your changes. Please make sure your code is readable and documented.

   • The Google style is preferred.
   • Use docstrings with args and returns defined for each function.
   • Typing annotations are preferred.
   • Use comments to explain steps of calculations and algorithms.
   • Use consistent variable names.
     • Please use full words and not letters as variable names so that variables are readable.

3. Commit often

   Commit your changes frequently with short, descriptive messages. This helps track progress and makes it easier to identify issues.

   git add <changed_files>
   git commit -m "Short, descriptive commit message"

4. Open a pull request

   Before you make any changes, you can write a descriptive plan of what you intend to do and why. Once your changes are ready, push your branch to the remote repository. Provide a clear explanation of what you changed and why.

   git push origin your-branch-name

   • Go to the repository on GitHub.
   • Click on Compare & pull request.
   • Fill in the title and description of your pull request.
   • Click Create pull request.

5. Test your changes

   Ensure your changes pass all tests and do not break existing functionality.

6. Request a review

   In the pull request, request a review from the appropriate team members. Notify them via GitHub.

7. Merge your changes to main

   After your code passes review, merge your changes to the main branch.

   • Click Merge pull request on GitHub.
   • Confirm the merge.

8. Delete your remote branch

   Once your changes are merged, delete your remote branch to keep the repository clean.

Releasing ariadne-roots

The GitHub Action workflow .github/workflows/python-publish.yml results in the package, ariadne-roots, being released at PyPI.

To release a new package, follow these instructions:

Follow contributing instructions above

1. Make a new branch to record your changes

   git checkout -b <your_name>/bump_version_to_<version>

2. Modify version

   The pyproject.toml file contains the information for the pip package. Incrementally increase the "version" with each release.

   Semantic Versioning

   Semantic versioning (SemVer) is a versioning system that uses the format: MAJOR.MINOR.PATCH

   • MAJOR: Increase when you make incompatible API changes.
   • MINOR: Increase when you add functionality in a backward-compatible manner.
   • PATCH: Increase when you make backward-compatible bug fixes.

   For example:

   • If the current version is 1.2.3:
   • A breaking change would result in 2.0.0.
   • Adding a new feature would result in 1.3.0.
   • Fixing a bug would result in 1.2.4.

   Learn more about the rules of semantic versioning here.

3. Commit changes

   After making the required modifications, commit your changes:

   git add pyproject.toml
   git commit -m "Bump version to <version>"
   git push origin <your_name>/bump_version_to_<version>

4. Open a pull request

   1. Go to the repository on GitHub.
   2. You should see a banner prompting you to compare & create a pull request for your branch. Click it.
   3. Fill in the pull request title and description. For example:
      • Title: Bump version to <version>
      • Description: "This PR updates the version to <version> for release."
   4. Click Create pull request.

5. Request a review

   After creating the pull request, in the right-hand sidebar, click on Reviewers and select the appropriate reviewer(s). Notify the reviewer(s) via GitHub.

6. Merge your changes to main after review

   Once the reviewer approves your pull request, merge it into the main branch.

7. Release to trigger the workflow

   1. Go to the release page.
   2. Draft a new release:
      • Create a new tag with the version number you used in the repository.
      • Have GitHub draft the release notes to include all the changes since the last release.
      • Modify the release name to include ariadne-roots, so that it says ariadne-roots v<version> like the rest.
   3. Please ask for your release to be reviewed before releasing.

8. Verify the release

   Check PyPI and the GitHub Actions of our repository to make sure the pip package was created and published successfully.

   • You should see the latest release with the correct version number at pypi.org.
   • The Github Actions should have green checkmarks and not red X's associated with your release.

Contributors

• Kian Faizi
• Matthieu Platre
• Elizabeth Berrigan

Contact

For any questions or further information, please contact:

• Matthieu Platre: matthieu.platre@inrae.fr

References

1. Chandrasekhar, Arjun, and Navlakha, Saket. "Neural arbors are Pareto optimal." Proceedings of the Royal Society B 286.1902 (2019): 20182727. https://doi.org/10.1098/rspb.2018.2727 ↩

2. Conn, Adam, et al. "High-resolution laser scanning reveals plant architectures that reflect universal network design principles." Cell Systems 5.1 (2017): 53-62. https://doi.org/10.1016/j.cels.2017.06.017 ↩