Learning to See Peaks: Attention-Based Feature Extraction for Automated Chromatographic Peak Detection
Daniel Walter, Mathias Helbig, Birgit Weydanz, Dominik Voltmer, Juan Jose Bonfiglio, Carsten Marr, Tobias Großkopf · 2026
Read Learning to See Peaks: Attention-Based Feature Extraction for Automated Chromatographic Peak Detection by Daniel Walter, Mathias Helbig, Birgit Weydanz, Dominik Voltmer, Juan Jose Bonfiglio, Carsten Marr, Tobias Großkopf (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-13261407.txt.
Description
Daniel Walter and colleagues built in 2026 a machine learning tool that finds peaks in chromatography traces for antibody medicines. Size-exclusion chromatography checks antibody quality, but overlapping peaks and drifting baselines make human reading slow and variable. Their system, PFE-1, uses an attention-based transformer trained on millions of simulated traces.
The simulator was tuned to match real routine data for antibodies and related large molecules. The model flags likely peak regions and events, and transparent rules then draw the final peak boxes. On both simulated and real test sets it beat convolutional networks and classic derivative methods. The margin grew largest where peaks overlapped or took odd shapes. It also needs no hints about where peaks should appear. The authors offer the framework as a reproducible base for routine analysis.
Catalog record
| CREATOR | Daniel Walter, Mathias Helbig, Birgit Weydanz, Dominik Voltmer, Juan Jose Bonfiglio, Carsten Marr, Tobias Großkopf |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |