Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

41 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PySMACKS

Overview

PySMACKS is a Python framework for analysis of single-molecule FRET (smFRET) data, from the raw microscopy movies to the corrected FRET histogram and kinetic analysis.

The initial motivation for PySMACKS was to port the SMACKS framework, used within our group, into a modern Python framework. Additional features have then been integrated in order to provide a complete analytical pipeline within a comprehensive framework. The choice of Python was motivated by the vast community of Python developers in smFRET and biophysics fields, and the availability of state-of-the-art computational packages, especially in machine-learning and deep-learning methods.

PySMACKS includes features that are necessary for our ongoing research projects, and for shich we could not find existing solutions in the smFRET fields. For imstance, the ability to correct for the microscope stage drift is critical for long smFRET experiments, hence the development and integration of this feature into PySMACKS.

PySMACKS is modular in its architecture and is meant to follow the FAIR data principles. Especially, the choice of the OpenFRET format for the storage of the traces datasets is motivated by an interoperability purpose, with the idea that this open format makes it easy to translate to another open format (for instance netCDF) without loss of metadata. Open science is a core value of the development team of PySMACKS and we hope that the smFRET community will benefit from this framework and build upon it.

PySMACKS is provided as a Python package and is freely available from https://pypi.org/project/PySMACKS/.

Table of Contents

Key Features

In the current version (1.0.0):

  • Channels registration
    • including Optical Flow or Affine transformation correction methods
    • QC plot of the registration results
  • Traces extraction
    • including ALEX- and non-ALEX- smFRET data
    • Microscope stage drift correction
    • local background correction over time using the median of surrounding pixels, or advanced strategy (Total Variation, Minimum of Total Variation)
    • Storage of traces within OpenFRET JSON files
  • Traces viewer and filtering
    • interactive plot
    • manual labelling tool for downstream analysis of highlighted sections and/or ML-DL curation of training datasets
  • Stoichiometry-Efficiency FRET histogram
    • automatic calculation of FRET correction factors
  • Kinetic analysis
    • Trace-by-Trace HMM
    • Ensemble HMM
  • Metadata viewer

Installation

PySMACKS is available as a Python package from https://pypi.org/project/PySMACKS/

The suggested installation method is to use pip within a virtual environment:

  1. Set a virtual environment
```
python -m venv NAME_OF_YOUR_VENV
```

And activate it

```
NAME_OF_YOUR_VENV\Scripts\activate
```
  1. Install the PySMACKS package using pip
```
pip install PySMACKS
```
  1. Launch the PySMACKS GUI using the command
```
PySMACKS_GUI
```

Dependencies

Libraries Licence Note
numpy BSD-3-Clause
matplotlib Python Software Foundation License
pandas BSD 3-Clause
PyQt5 GPL v3
scipy BSD License
scikit-image BSD License
big-fish BSD 3-Clause paper
hmmlearn BSD License
opencv-python Apache 2.0
openfret MIT GitHub repo
pillow MIT-CMU
ruptures BSD-2-Clause paper * not used in the current version of PySMACKS*
seaborn BSD License
trackpy GitHub repo
tvdcondat2013 CeCILL FREE SOFTWARE GitHub repo
uuid7 MIT

Workflow

The analysis pipeline can be divided into two separate sections:

PySMACKS is meant to be modular and interoperable: raw movies can be processed by PySMACKS to extract the single-molecule traces, and analyzed within PySMACKS; or these traces can be exported to JSON files and processed using another framework.

Basically, the data management within PySMACKS follows the logic: TIFF --> JSON (OpenFRET format).

Traces extraction

Input data

In the current version (1.0.0), PySMACKS only handles TIFF stacks as data input for the traces extraction. Especially, the two channels (donor and acceptor) should be in separate TIFF files.

Additionally, for memory efficiency, the full movies should be split into multiple submovies, (usually a submovie < 500 MB), to avoid saturation of the RAM.

ALEX smFRET

PySMACKS supports both ALEX and non-ALEX smFRET data; you can specify it during the traces extraction process.

If your data are ALEX, it is expected that the acceptor submovies frames alternate between acceptor-acceptor frame (ALEX) and donor-acceptor frame (FRET), with the first frame of the movie being ALEX.

Channels registration

You can correct the chromatic aberrations and channels misalignments using optical beads TIFF movies (donor and acceptor separate files) and the registration module of PySMACKS.

Currently, two methods can be used for the registration:

  • Optical flow
  • Affine transformation

The results of the registration step are stored within numpy files (*.npy)

Extraction step

The individual traces are detected using a LoG filter and a tracking algorithm.

The drift of the microscope stage can be corrected at this stage.

The individual traces are stored within a JSON file (compressed as .json.zip) following the OpenFRET format (source repo: https://github.com/simol-lab/OpenFRET)

All metadata are stored within the OpenFRET structure, including the local background intensity over time for each individual traces.

Traces analysis

Input data

For the trace analysis (filtering, FRET histogram, kinetics), the input data is the JSON file (compressed as .json.zip) following the OpenFRET format.

The outputs of the analysis pipeline are saved within the same .json file through its metadata.

Trace filtering

PySMACKS contains a trace viewer, which you can use to visualize traces one-by-one and manually filter them. You can also highlight relevant sections of these traces using an interactive tool.

Especially, this labelling interactive tool can be used to curate datasets for ML/DL training purposes.

NB: the development of a fully integrated deep learning approach for traces filtering in under development and will be implemented in future releases of PySMACKS.

Stoichiometry-Efficiency FRET histogram

If your data are ALEX-smFRET, after filtering and labelling of FRET, Donor-Only and Acceptor-Only sections of individual traces, you can plot the stoichiometry-efficiency (SE) FRET histogram using this module.

Especially, the calculation of the FRET corrections factors (alpha, delta, beta and gamma) is fully automated with intermediate quality check by the user.

A corrected FRET efficiency histogram can be computed.

Kinetic analysis

The last module of the pipeline is dedicated to the extraction of the kinetic rates governing states transitions, using a Hidden-Markov Model (HMM) framework.

In details, this HMM approach relies on the SMACKS framework (original study: 10.1016/j.bpj.2016.08.023), initially developed in IGOR language, and ported to Python within PySMACKS.

Users can conduct HMM inferences on a Trace-by-Trace basis (one individual HMM model optimized for each individual traces), or on an Ensemble way.

The outputs are rates matrix (TbT or Ensemble), related to the discrete states transition rates of the single-molecules.

Tutorial dataset

To illustrate the functionalities of PySMACKS and help new users to learn how to navigate through this framework, we provide in a Zenodo repository a complete dataset that you can use as inputs for each modules of PySMACKS.

This dataset contains experimental data from a Holliday junction sample (two states model) acquired with our custom TIRF setup (Schmid lab, University of Basel, 2026).

Citation: Molcrette, B.& Schmid, S. (2026). Dataset for the PySMACKS tutorial (Version 1.0.0) [Dataset]. Zenodo. https://doi.org/10.5281/zenodo.21793247

Documentation

The full documentation associated with PySMACKS is available at the following URL (English):

https://molcretb.github.io/PySMACKS

Contributors

Bastien Molcrette [0000-0002-5995-5376] (University of Basel): development, documentation, maintenance, primary contact

Sonja Schmid [0000-0002-3710-5602] (University of Basel): support, funding

Contributing

Contributions to PySMACKS are very welcome! If you would like to join the development team, please follow these steps:

  1. Fork the PySMACKS repository
  2. Create a new branch
  3. Add your new features
  4. Commit the changes
  5. Push to the newly created branch
  6. Open a Pull Request

Thanks for your help and support!

Acknowledgements

  • Thanks to all early testers of PySMACKS!
  • Special thanks to all current and past members of the Schmid lab, including all the master students that have joined us for their internship.
  • The past crew of GigaScience and the biocurators community, for their dedication to open science and FAIR principle.
  • The Stack Overflow and Image.sc communities, for providing great insights in software development and image analysis.

Funding

This work is supported by:

  • the University of Basel
  • the Swiss Nanoscience Institute (SNI)
  • the Swiss National Science Foundation (SNSF)
  • the National Centres of Competence in Research (NCCR)
  • the Innovation Office of the University of Basel

Citation

Molcrette, B., & Schmid, S. (2026). PySMACKS (Version 1.0.0b2) [Computer software]. https://github.com/molcretb/PySMACKS

Persistent identifiers

License

Distributed under the MIT License. See LICENSE for more information.

Contact

Please use the GitHub issues to report any problem with PySMACKS, thanks.

Other requests: send email to bastien.molcrette@unibas.ch

About

Modular Python framework for single-molecule FRET data analysis

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages