Example datasets

The spectral-unmixing tutorials are backed by a small public example-dataset collection that is available as a Zenodo dataset release:

Musacchio, F. (2026). Example datasets for the spectral-unmixing pipeline [Data set]. Zenodo. https://doi.org/10.5281/zenodo.20984021

The goal of this collection is to let users replay the tutorials with the same inputs that are used throughout the documentation and the interactive scripts in user_scripts/.

What is included

The example data are organized into three main folders:

  • example_data/PICASSO_examples/ public two-channel, three-channel, and five-channel PICASSO-family example images used by the blind-unmixing tutorials

  • example_data/synthetic_data/ synthetic TZCYX data with controlled two-channel bleed-through, used for the full-stack unmixing tutorial

  • example_data/Gockel_Nieves_Rivera_2026/ a cropped real microscopy stack used for the helper tutorial on registration, filtering, histogram matching, and projection

Each folder contains a dedicated README file with more details about the dataset, its provenance, and the license under which it is redistributed. The entire collection is licensed under CC BY 4.0.

Dataset-to-tutorial mapping

The main public tutorial mappings are:

Dataset descriptions and provenance

PICASSO_examples

This folder contains a subset of the public example images shared with the PICASSO publication.

It includes public two-channel, three-channel, and five-channel example stacks that are suitable for:

  • standard two-channel unmixing examples,

  • bidirectional unmixing examples,

  • and PICASSO-family blind-unmixing tutorials.

Source:

Chang, Jae-Byum; Seo, Junyoung; Sim, Yeonbo; Kim, Jeewon; Kim, Hyunwoo; Cho, In; et al. (2022). PICASSO allows ultra-multiplexed fluorescence imaging of spatially overlapping proteins without reference spectra measurements. figshare. Figure. https://doi.org/10.6084/m9.figshare.19596682.v1

Associated paper:

Seo, J., Sim, Y., Kim, J. et al. PICASSO allows ultra-multiplexed fluorescence imaging of spatially overlapping proteins without reference spectra measurements. Nature Communications 13, 2475 (2022). https://doi.org/10.1038/s41467-022-30168-z

License:

synthetic_data

This folder contains synthetic tutorial data created specifically for spectral-unmixing.

The main stack is:

  • synthetic_bleedthrough_T9_Z20_C2.tif

It is a synthetic TZCYX stack with:

  • T=9

  • Z=20

  • C=2

The stack contains two time-varying 3D Gaussian structures and controlled bleed-through from channel 0 into channel 1. It is especially useful for the full-stack tutorial because the construction is known and the behavior of the unmixing methods can be tested in a controlled setting.

Source:

  • generated within this repository using additional_scripts/generate_synthetic_bleedthrough_stack.py

License:

Gockel_Nieves_Rivera_2026

This folder contains a cropped real microscopy stack used in the helper tutorial for:

  • intra-stack z-drift correction,

  • across-time registration,

  • histogram matching across time,

  • filtering,

  • and max-z projection.

Source:

Gockel, N., Nieves-Rivera, N., Musacchio, F., Druart, M., Jaako, K., Fuhrmann, F., Rozkalne, R., Poll, S., Baiba, J., Fuhrmann, M., & Le Magueresse, C. (2025). Example Datasets for Microglial Motility Analysis Using the MotilA Pipeline [Data set]. Zenodo. https://doi.org/10.5281/zenodo.15061566

The file used here is a cropped derivative of that public source dataset.

License:

Because the stack in this folder is a cropped derivative, it is redistributed under the same CC BY-SA 4.0 license.

Citation

If you are using the example datasets in your own work, please cite the Zenodo dataset release:

Musacchio, F. (2026). Example datasets for the spectral-unmixing pipeline [Data set]. Zenodo. https://doi.org/10.5281/zenodo.20984021