Year: 2026
Type: preprint
Abstract: Discovery of peptides that undergo liquid–liquid phase separation (LLPS) is important for establishing sequence–property relationships but systematic approaches have remained elusive so far. Here, we report a tandem mass spectrometry platform to identify peptides prone to LLPS directly from combinatorial peptide libraries encompassing over 44,000 sequences and achieving an 86% true-positive rate across 38 validated candidates. This mixture-based strategy reveals that LLPS forming peptides stabilize aggregation-prone sequences, revealing emergent behavior undetectable in conventional single-peptide analysis. The resulting dataset establishes a foundation for machine learning applications to enable high-throughput prediction of peptide aggregation states.
Source: ChemRxiv
Authors: Kübra Kaygisiz, Bradley L. Pentelute
Language: English
Subfield: Spectroscopy
Field: Chemistry
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7164585116