Single Molecule Tracking and Diffusion

Daniel Sage — Ecole Polytechnique Fédérale de Lausanne, Switerzland

This module complements the youngSMLMS 2025 workshop.

This hands-on module guides you to the quantitative single-particle diffusion analysis. Across four practices, you will:

• Analyze SMLM data in Fiji
• Track particles
• Understand diffusion Model in Python
• Integrate a SPT pipeline

By the end, you’ll understand the full SMLM/SPT pipeline—localization, tracking, diffusion modeling, and quantitative readouts through simple examples.

Setting up Fiji

Prerequisite: A laptop with installation permissions.

1. Download Fiji from the official website: https://fiji.sc/
2. Update Fiji to the latest version.
3. Install ThunderSTORM: https://zitmen.github.io/thunderstorm/

Setting up the Jupyter environment

Prerequisite: A Google account (preferably private).

1. Open Google Colab
2. Upload the notebooks from this GitHub repository.
3. Go to File > Save a copy in Drive to keep your own editable version.

Prerequisite: A working Conda installation.

conda create --name spt-diffusion python=3.11
conda activate spt-diffusion

pip install notebook
pip install scikit-image scikit-learn
pip install andi-datasets deeptrack trackpy

jupyter notebook

Practice A - SMLM Data Analysis

On Fiji

• Handling sequences of Frames — Fiji commances
• 2D SMLMS — ThunderSTORM
• Demo 3D — QuickPALM

---

Practice B - Particle Tracking

On Fiji

• Tracking of molecules — TrackMate

---

Practice C - Modeling Molecular Diffusion

Python Notebook

simulation_brownian_motion.ipynb

• Isotropic Brownian Motion
• Anisotropic Brownian Motion
• Non-homogeneous Brownian Motion

simulation_anomalous_diffusion.ipynb

• Test the 5 models provided by the ANDI Challenge
• Reference: https://github.com/AnDiChallenge/andi_datasets

Practice D - Estimation of Molecular Diffusion

classification_trajectory.ipynb

• Compute simple features of trajectory
• Classify the trajectory using a Random Forest classifier

simple_single_particle_tracking.ipynb

• Simulation 2 groups of blinking particles (slow and fast diffusion) and generating corresponding noisy frames
• Detection and tracking of particles
• Recover diffusion coefficient of the detected particles

Exercise

• Accurate localization of particles and track them
• Recover diffusion coefficient of the localized particles