In life sciences, tracking objects from movies is pivotal for quantifying behaviors of particles, organelles, bacteria, cells, and whole animals. CellTracksColab bridges the gap between tracking and analysis. CellTracksColab simplifies the journey from data compilation to analysis.

• 📘 Holistic View: Comprehensive analysis across fields of view, biological repeats, and conditions.
• 🖥️ User-Centric: Intuitive GUI designed for all users.
• 🔍 Visualization: Track visualization and filtering.
• 📊 Analysis: Deep-dive into track metrics and statistics.
• 🧪 Reliability: Check experimental variability using hierarchical clustering.
• 🔧 Advanced Tools: Harness the power of UMAP, t-SNE, and HDBSCAN.
• 💼 Flexibility: Tailor and adapt to your needs.

TrackMate	CellProfiler	Icy	ilastik	Fiji Manual Tracker

May also be compatible with other tracking software exporting tracking results that meet our minimal requirements. More info here.

Tutorial 1: Getting Started with CellTracksColab using Google Colab

Tutorial 2: Using CellTracksColab locally using Jupyter

Tutorial 3: Using CellTracksColab locally using Google Colab

I2K 2024: CellTracksColab tutorial

ℹ️ Tutorials 2 and 3 show the previous way of running CellTracksColab locally (manual Anaconda/Jupyter setup). The recommended local option is now the CellTracksColab desktop app; see the Quick Start below.

CellTracksColab notebooks can run in two ways. The notebooks and analyses are the same in both; only where they run changes.

1. Pick a notebook from the tables below and click its badge.
2. (Recommended) Save your own copy with File > Save a copy in Drive.
3. Run the Load key dependencies cell (section 1.1). In Colab it downloads CellTracksColab into the session and asks for permission to connect your Google Drive.
4. Point the notebook at your data on Drive (paths start with /content/gdrive/MyDrive/), or use the test dataset where the notebook offers one.

More details: Running CellTracksColab using Google Colab.

The desktop app is built with LabConstrictor. It bundles Python, JupyterLab and every dependency, so you don't need to set up Conda or Python yourself.

1. Install: follow the installation guide for your operating system. Installers are also on the Releases page.
2. Launch: open CellTracksColab from the Start Menu (Windows), the Applications folder (macOS) or your applications menu (Linux). A terminal window opens (keep it open while you work) and JupyterLab starts in your browser with the Welcome notebook.
3. Open a notebook: in the Welcome dashboard, click Open the Notebook next to the analysis you want. The Welcome notebook can also check for notebook updates.
4. Run it: your data stays on your computer. Paste the path of a local folder into the text box (on Windows and Linux you can also pick it with the folder selector). See how to run notebooks in the desktop app to learn how to run cells when the code is hidden and how to show it with Show/Hide Code.

More details: Using the notebooks after installation · Troubleshooting the desktop app.

git clone https://github.com/CellMigrationLab/CellTracksColab.git
cd CellTracksColab
conda env create -f environment.yaml   # Python 3.12 + JupyterLab, environment "celltrackscolab"
conda activate celltrackscolab
pip install -r requirements.txt        # NVIDIA GPU users can use requirements_gpu.txt instead
pip install -e .                       # makes the `celltracks` package (in src/) importable
jupyter lab

We provide three notebooks for loading and analyzing your data depending on its format. The Link column opens each notebook in Google Colab. In the desktop app, all notebooks are listed in the Welcome dashboard.

Notebook	Purpose	Required File Format	Link
CellTracksColab - TrackMate	Load and analyze TrackMate data. More info here.	CSV or XML files
CellTracksColab - Custom	Analyze data from CellProfiler, ICY, ilastik, or Fiji Manual Tracker. More info here.	CSV files
CellTracksColab - Viewer	Load and share data in the CellTracksColab format.	CellTracksColab format

These notebooks require your dataset to be in the CellTracksColab format.

Notebook	Purpose	Required File Format	Link
CellTracksColab - Dimensionality Reduction	Utilize advanced dimensionality reduction techniques.	CellTracksColab format
CellTracksColab - Track Spatial Clustering Analysis	Dive deeper into your dataset with track clustering analysis.	CellTracksColab format
CellTracksColab - Distance to ROI	Analyze movement tracks in relation to designated ROIs.	CellTracksColab format

More to come

• Handle TrackMate CSV files structured in a plate format, such as file names commonly produced by incubator microscopes like Incucytes.
• 

CellTracksColab is inspired by several key projects in cell tracking and analysis. We acknowledge the influential contributions of Traject3d, CellPhe, CelltrackR, the MotilityLab website, and Cellplato on Zenodo. The innovative use of UMAP and HDBSCAN for analyzing tracking data, as featured in CellTracksColab, was first introduced in CellPlato.

• Test Dataset: Start exploring with our test datasets in CellTracksColab CSV format, or TrackMate CSV format.
• Data Structure: Organize with our two-tiered folder hierarchy. Details here.
• Data Requirements: Note that CellTracksColab does not yet support track merging or splitting.

• 📖 CellTracksColab analysis documentation: Visit the comprehensive wiki.
• ☁️ Running in Google Colab: Colab guide.
• 🖥️ Installing the desktop app: Installation guide (Windows, macOS and Linux, including how to uninstall).
• 📓 Using notebooks in the desktop app: Notebook usage guide · Running cells with hidden code.
• 🛠️ Problems with the desktop app: Troubleshooting guide.

• Estibaliz Gómez-de-Mariscal
• Hanna Grobe
• Joanna W. Pylvänäinen
• Laura Xénard
• Ricardo Henriques
• Jean-Yves Tinevez
• Guillaume Jacquemet

We welcome your insights and improvements! There are several ways you can contribute to the CellTracksColab project:

If you encounter any bugs, have suggestions for improvements, or want to discuss new features, please raise an issue on our GitHub Issues page.

We are excited to see new analysis notebooks built on the CellTracksColab platform. If you have developed a new notebook, please submit it via a pull request. All submitted notebooks should include a test dataset to showcase their functionality. Each notebook will be tested by a member of the team before being released.

We expect all contributors to adhere to our simple code of conduct:

• Be respectful and considerate of others.
• Provide constructive feedback.
• Collaborate openly and honestly.

By participating in this project, you agree to abide by these guidelines.

Thank you for contributing to CellTracksColab! Your support and contributions help us improve and expand the platform for everyone in the community.

Licensed under the MIT License. Details here.

If you use CellTracksColab in your research, please cite the following paper:

Gómez-de-Mariscal, E., Grobe, H., Pylvänäinen, J. W., Xénard, L., Henriques, R., Tinevez, J.-Y., & Jacquemet, G. (2024). CellTracksColab is a platform that enables compilation, analysis, and exploration of cell tracking data. PLOS Biology. Published August 8, 2024. https://doi.org/10.1371/journal.pbio.3002740

Figure 1: Compile your data	Figure 2: Visualise your tracks
Figure 3: Compute additional metrics	Figure 4: Plot track parameters
Figure 5: Compute Similarity Metrics between Field of Views and between Conditions and Repeats	Figure 6: Perform UMAP
Figure 7: Identify clusters using HDBSCAN	Figure 8: Understand your clusters using a heatmap