spora#

The unified, multimodal, AI-ready platform for spatial proteomics data.

Access the data

Artificial Intelligence in Molecular Medicine EPFL

The spora Platform#

spora [data]

Standardized Datasets

Standardized multiplex images from 35 cohorts, with paired H&E, tissue and cell masks, cell types, and patient metadata in one format. Hosted on the BioImage Archive.

Browse the collection
spora [io]

Data Loading

A Python library to load tissues and tiles, combine modalities, and sample across cohorts, with scripts to convert new datasets into the spora format.

Documentation
spora [bench]

Benchmarks

Standardized tasks and protocols to evaluate spatial proteomics foundation models at the subcellular, cell, and tissue level.

Documentation

The spora ecosystem is one pipeline covering AI development in spatial proteomics, from the initial data composition to the final evaluation.

  1. FormatImages and annotations in one format
  2. LoadLoaded as tissues, tiles, and cohorts
  3. EvaluateEvaluated on standardized tasks

The spora [data] Collection#

13,546 standardized multiplex images from 5,477 patients, spanning 35 cohorts, 14 tissue origins, and 5 imaging technologies. Each sample comes with its multiplex image and, where available, a paired H&E image, tissue and cell masks, cell types, patient data, and metadata.

Tissue Origin944,680 multiplex tiles, log scale
  1. Pancreas418,132
  2. Colon219,132
  3. Breast144,960
  4. Lung62,110
  5. Brain28,890
  6. Liver28,781
  7. Lymph node18,859
  8. Spleen11,426
  9. Head and neck6,864
  10. Skin2,063
  11. Other1,913
  12. Soft tissue1,065
  13. Bladder242
  14. Tonsil242
TechnologyShare of multiplex tiles
  • IF 43.2%
  • Orion 21.0%
  • IMC 20.2%
  • CODEX 10.9%
  • MIBI 4.7%
Pairing with H&EShare of multiplex tiles
Multimodal 30.7%
  • Unimodal 69.3%
  • Multimodal 30.7%
  • Registered 26.1%
  • Matched 4.7%

Percentages refer to the total number of multiplex tiles. Download the datasets from the BioImage Archive; the format reference describes the files.

Citation#

If you use spora in your research, please cite:

@inproceedings{querfurth2026spora,
  title = {{spora: A Unified Multimodal Dataset and Platform for Spatial Proteomics}},
  author = {von Querfurth, Benedikt and Wenckstern, Johann and Jain, Eeshaan and Klein, Lukas and Cheng, Phil F. and Cheng, Yexiang and Vincent-Cuaz, Cédric and Frossard, Pascal and Haberecker, Martina and Wicki, Andreas and Michielin, Olivier and Bunne, Charlotte},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS) Evaluations & Datasets (E&D) Track},
  year = {2026}
}

Please also cite the original papers of the cohorts you use within spora [data].