Note: If you are looking for MeXpose v1 you can find it under the mexpose_v1 branch.

Welcome

MeXpose 2.0 is a major update to our interactive workflow for comprehensive analysis of multiplexed imaging data, optimised for spatial metallomics. The new release streamlines the existing pipeline by replacing third-party software with custom Jupyter notebooks for image reconstruction, preprocessing, and data extraction. Additional new features such as limit-of-detection calculation, cytometry-style gating, and integrated spatial statistics enhance both exploratory and quantitative analyses within a single, intuitive environment.

MeXpose 2.0 covers the following aspects of multiplexed image analysis by providing custom notebooks for each step:

• Image reconstruction (TOFWerk and Nu Instruments specific)
• Single-cell data extraction
• Single-cell analysis
  • size normalisation & outlier filtering
  • clustering, dimensionality reduction and spatial analysis
  • histograms, heatmaps and spatial single-cell overlays
• Manual population gating
• Quantification
• Limit-of-detection calculations

Installation

Environment Setup

All dependencies can be installed by creating a new mamba/conda environment using the provided mexpose_v2.yaml file:

# Create the environment from the yaml file
mamba env create -f mexpose_v2.yaml

# Activate the environment
mamba activate mexpose_v2

Note: Replace mamba with conda when using Anaconda.

JupyterLab Configuration

To handle large file uploads (>10 MB) in the notebooks, you'll need to create a custom JupyterLab configuration:

Linux/macOS:

1. Create a JupyterLab config file:

   jupyter lab --generate-config

2. Edit the configuration file (located at ~/.jupyter/jupyter_lab_config.py) and add:

   c.ServerApp.tornado_settings = {"websocket_max_message_size": 1024 * 1024 * 1024}

Windows:

1. Create a JupyterLab config file:

   jupyter lab --generate-config

2. Edit the configuration file (located at %USERPROFILE%\.jupyter\jupyter_lab_config.py) and add:

   c.ServerApp.tornado_settings = {"websocket_max_message_size": 1024 * 1024 * 1024}

Alternative (All platforms):

You can use the provided jupyter_lab_config.py file:

jupyter lab --config=jupyter_lab_config.py

This increases the upload limit to 1 GB, allowing you to work with larger imaging datasets directly within the notebooks.

Supported Setups and File Types

MeXpose 2.0 was tested on the following setups:

TOFWerk

Software	Version	Manufacturer
TOFPilot	2.15	TOFWERK, Thun, Switzerland
Chromium	3.2	Teledyne Photon Machines, Bozeman, USA
ActiveView2	1.5.1.30	Elemental Scientific Lasers (ESL), Bozeman, USA

Using raw combined laser- and TOF data in HDF5 format

Nu Instruments

Software	Version	Manufacturer
NuQuant	1.2.8739.1	Nu Instruments Ltd., Wrexham, UK
Chromium	3.2	Teledyne Photon Machines, Bozeman, USA
ActiveView2	1.5.1.30	Elemental Scientific Lasers (ESL), Bozeman, USA

Using preprocessed (NuQuant) combined laser- and TOF data in CSV format

Supported file types

• HDF5
• CSV
• TIFF, OME.TIFF, BIGTIFF, PNG, JPEG

Support for additional file formats is provided by the use of the imageio Python library.

Segmentation

MeXpose is focused around single-cell analysis. In order to extract single-cell data as well as for several visualisations a greyscale segmentation mask is required.
We recommend using (a recent version of) Cellpose as we find it performs very well and works reliably while additionally allowing annotation and training in a user-friendly GUI.

Visit the Cellpose github for installation instructions and their documentation for additional information.

You can also try Cellpose-SAM here:

• Try cellpose-sam

Pachitariu, M., Rariden, M., & Stringer, C. (2025). Cellpose-SAM: superhuman generalization for cellular segmentation. bioRxiv.

Citing MeXpose

Please cite the following papers when using either the original MeXpose or MeXpose 2.0 in your work:

MeXpose 2.0

To be updated

MeXpose 1.0

MeXpose - A modular imaging pipeline for the quantitative assessment of cellular metal bioaccumulation. Gabriel Braun and Martin Schaier, Paulina Werner, Sarah Theiner, Juergen Zanghellini, Lukas Wisgrill, Nanna Fyhrquist, Gunda Koellensperger. JACS Au 2024 4 (6), 2197-2210; DOI: 10.1021/jacsau.4c00154

@article{Braun_Schaier2024,
author = {Braun, Gabriel and Schaier, Martin and Werner, Paulina and Theiner, Sarah and Zanghellini, Jürgen and Wisgrill, Lukas and Fyhrquist, Nanna and Koellensperger, Gunda},
title = {MeXpose─A Modular Imaging Pipeline for the Quantitative Assessment of Cellular Metal Bioaccumulation},
journal = {JACS Au},
volume = {4},
number = {6},
pages = {2197-2210},
year = {2024},
doi = {10.1021/jacsau.4c00154},
URL = {https://doi.org/10.1021/jacsau.4c00154},
eprint = {https://doi.org/10.1021/jacsau.4c00154}
}

---

Licensing

The code included with MeXpose 2.0 is distributed under an MIT license.
The code included with MeXpose 1.0, namely all python scripts, jupyter notebooks and fiji macros, is distributed under an MIT license. The MeXpose container hosted on DockerHub is distributed under a GPLv3 license.