Overcoming Artificial Structures in Resolution‐Enhanced Hi‐C Data by Signal Decomposition and Multi‐Scale Attention
Qinyao Li, Kelly Yichen Li, Chiara Nicoletti, Stephen Kwok‐Wing Tsui, Pier Lorenzo Puri, Qin Cao, Kevin Y Yip · 2026
Read Overcoming Artificial Structures in Resolution‐Enhanced Hi‐C Data by Signal Decomposition and Multi‐Scale Attention by Qinyao Li, Kelly Yichen Li, Chiara Nicoletti, Stephen Kwok‐Wing Tsui, Pier Lorenzo Puri, Qin Cao, Kevin Y Yip (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-13554405.txt.
Description
Hi-C maps show how a genome folds by recording which DNA regions touch, but the maps come out blurry. Software sharpeners based on deep learning fix the blur yet leave fingerprints: fake structures where the map was cut into patches for processing. Qinyao Li and co-authors present a cleaner sharpener, SHARP, in a 2026 paper.
Their method first splits the signal into three kinds: effects of plain distance along the DNA, blocky domain patterns, and fine detail. Only the fine detail goes through the neural network, so patch edges cannot corrupt the rest. Attention at several scales lets the network read both neighborhood and wide context. On tests including new samples and another species, SHARP beat leading methods in accuracy and in spotting real interactions.
Catalog record
| CREATOR | Qinyao Li, Kelly Yichen Li, Chiara Nicoletti, Stephen Kwok‐Wing Tsui, Pier Lorenzo Puri, Qin Cao, Kevin Y Yip |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 92 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |