fragmentum
CHAPTER THREE
LITERATURE
VOL. I  ·  ISSUE 01
BERLIN  ·  2026
§3.0   THE PUBLISHED RECORD

Already in Nature.
Already on arXiv.
Already cited.

Album was not built in a closet. Four publications across Nature Communications, Nature Protocols, and arXiv, plus the Chan Zuckerberg Initiative's CZII CryoET competition with 1,200+ teams, already use or build on the framework. The literature is the proof.

Selected publications.

1.
arXiv 2026 v2 · latest
Album: executable building blocks for scientific imaging routines, from sharing to LLM-assisted orchestration.
J. P. Albrecht, D. Schmidt, K. Harrington.
The latest version of the album framework paper. v2 updates the original 2021 preprint with the LLM-assisted orchestration angle: how solutions are drafted, revised, and chained by language models through album-mcp. The technical foundation everything else rests on, now extended.
arXiv:2110.00601 (v2, 2026-04-21) · read
2.
Nature Communications 2025 Original Research
Polarity-JaM: an image analysis toolbox for cell polarity, junction and morphology quantification.
W. Giese*, J. P. Albrecht*, O. Oppenheim, E. B. Akmeric, J. Kraxner, D. Schmidt, K. Harrington, H. Gerhard.
A modular image-analysis pipeline for endothelial cell polarity, packaged as album solutions and shipped to the public catalog. The first end-to-end demonstration of album as the distribution mechanism for a Nature-level computational toolbox.
Nat. Commun. 6(1):1474, 2025-02-08.
3.
Nature Protocols 2024 Protocol
Modular segmentation, spatial analysis and visualization of volume electron microscopy datasets.
A. Müller*, D. Schmidt*, J. P. Albrecht, L. Rieckert, M. Otto, L. Garcia, G. Fabig, M. Solimena, M. Weigert.
A protocol paper that explicitly distributes its critical computational steps as Album solutions for deep-learning segmentation, spatial analysis, and 3-D rendering. The paper bridges compatibility issues between open-source tools that previously could not be combined.
DOI: 10.1038/s41596-024-00957-5
4.
arXiv 2025 Preprint
ARI3D: A software for interactive quantification of regions in X-Ray CT 3D images.
J. P. Albrecht*, J. Godinho*, C. Hübers, D. Schmidt.
An interactive 3-D region-quantification tool for X-Ray CT, distributed as a versioned album solution. Demonstrates album's multi-language reach: the GUI is interactive, the compute is heavy, the distribution is one command.
arXiv:2508.xxxxx · 2025-08-13

The CZII CryoET competition.

In November 2024 the Chan Zuckerberg Initiative launched the CZII CryoET Object Identification challenge on Kaggle: identify six classes of protein complex inside three-dimensional electron tomograms. The data layer was copick, a catalog hosted on the album stack. Over twelve hundred teams competed for seventy-five thousand dollars in prizes, and the ML community got a stress-test of the album distribution model at competition scale.

1,200+ Teams competing on copick-derived data
$75K Prize pool funded by Chan Zuckerberg Initiative
6 Protein classes apo-ferritin, β-amylase, β-galactosidase, &c.
The fastest path from a research idea to a citation, today, is a Git-hosted catalog of versioned tools. The slowest is a Docker container nobody can find. Fragmentum, internal

Next: §4 · Investigators.

Who is building this. Who advises.

Continue → §4 Back to §2 ↩