Example datasets for the spectral-unmixing pipeline

This archive provides the public example datasets used to reproduce the tutorials of the Python package spectral-unmixing. The collection is intended to let users run the tutorial scripts, inspect the example stacks, and compare their own results against the documented workflows for spectral unmixing, blind unmixing, filtering, registration, and projection. The example datasets are grouped into three folders: Gockel & Nieves-Rivera et al. (2026) The folder Gockel_Nieves_Rivera_2026 contains a cropped microscopy stack used for the helper tutorials on registration, histogram matching, filtering, and max-z projection. What it shows A cropped 5D microscopy stack used as a realistic example for downstream post-unmixing processing, especially temporal registration and intra-stack z-drift correction. Source Derived as a cropped cut-out from: 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 License Creative Commons Attribution Share Alike 4.0 International (CC BY-SA 4.0) https://creativecommons.org/licenses/by-sa/4.0/ Because the files in this folder are derived from that source dataset, they are redistributed under the same CC BY-SA 4.0 license. PICASSO example data sets The folder PICASSO_examples contains a subset of the public example images shared with the PICASSO publication. These files are used in the tutorials for PICASSO-family blind unmixing, multi-channel unmixing, and selected two-channel and bidirectional examples. What they show Public 2-channel, 3-channel, and 5-channel example images and simulations that are suitable for demonstrating blind unmixing and source-sink unmixing workflows. Source Subset of the example data shared at: 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 publication: 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 Creative Commons Attribution 4.0 International (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/ Synthetic data set The folder synthetic_data contains synthetic microscopy data created specifically for the `spectral-unmixing` tutorials. What it shows A synthetic `TZCYX` stack with `T=9`, `Z=20`, and `C=2`, containing two time-varying 3D Gaussian structures and controlled bleed-through from channel 0 into channel 1. The dataset is intended for tutorial reproduction and for testing alpha-estimation methods on data with known construction. Source Generated within the `spectral-unmixing` repository using additional_scripts/generate_synthetic_bleedthrough_stack.py. License Creative Commons Attribution 4.0 International (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/