SCUDDO: an unsupervised clustering algorithm for single-cell Hi-C maps using diagonal diffusion operators
Luka Maisuradze, Mark D Shattuck, Corey S O’Hern, Peter Robinson · 2026
Read SCUDDO: an unsupervised clustering algorithm for single-cell Hi-C maps using diagonal diffusion operators by Luka Maisuradze, Mark D Shattuck, Corey S O’Hern, Peter Robinson (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13202327.txt.
Description
Luka Maisuradze and co-authors present an algorithm that sorts single cells by type from sparse chromosome contact maps, published in 2026. Single-cell Hi-C maps record which DNA regions touch inside one nucleus, but the data are so sparse that clustering them into cell types is notoriously hard. Top methods barely beat an accuracy index of 0.4 on tough sets. Their method, SCUDDO, embeds the maps using diagonal diffusion operators before clustering.
On four difficult datasets it beat rival algorithms by a clear margin on the standard agreement index. It kept its lead even when restricted to fewer chromosome maps per cell or a small slice of each map's contacts. The authors release the code and data openly. They argue the diffusion view extracts latent structure that sparsity hides from other methods.
Catalog record
| CREATOR | Luka Maisuradze, Mark D Shattuck, Corey S O’Hern, Peter Robinson |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |