High-accuracy glycan de novo prediction for N- and O-linked glycopeptides across multiple fragmentation techniques

peer-reviewed · Nature Communications · 2026

peer-reviewed · Nature Communications · 2026. Qianqiu Zhang et al. N- and O-glycosylation are structurally diverse post-translational modifications that affect a range of…
Date 2026-09-10
Type peer-reviewed
Venue Nature Communications
Publisher Springer Science and Business Media LLC
Contribution adjacent
DOI 10.1038/s41467-026-77012-2
Citations (OpenAlex) 0
Venue 2-year citedness 17.60

Abstract

N- and O-glycosylation are structurally diverse post-translational modifications that affect a range of biological functions. Glycoproteomics faces substantial challenges, particularly in the analysis of O-glycans due to their diversity compared to that of N-glycans. In addition, tandem mass spectrometry data patterns exhibit variability between different fragmentation methods. Existing de novo algorithms often lack sensitivity and are limited to sceHCD fragmentation, restricting their practical application. To address these limitations, we introduce DeepGlycan, a deep-learning-based method for de novo glycopeptide sequencing that captures relationships between glycopeptide spectra and fragment ions from both N- and O-glycans. DeepGlycan achieves over 92% glycan recall and around 95% glycan precision on N-glycopeptide spectra generated using both sceHCD and EThcD. In addition, it enables O-glycan de novo sequencing without additional training. Beyond benchmarking, DeepGlycan identifies an O-glycopeptide in mouse heart tissue whose assignment is supported by exoglycosidase treatment and comparison with a synthetic standard.

Authors

  1. Qianqiu Zhang · University of Waterloo
  2. Zeping Mao · Bioinformatics Solutions Inc., University of Waterloo
  3. Yuling Chen · Tsinghua University
  4. Baozhen Shan · Bioinformatics Solutions Inc.
  5. Haiteng Deng · Tsinghua University, Yangtze Delta Region Institute of Tsinghua University
  6. Ming Li · Bioinformatics Solutions Inc., Central China Institute of Artificial Intelligence, Peng Cheng Laboratory, University of Waterloo, University of Western Ontario

Methods and tools

  • DeepGlycan: Deep-learning de novo glycopeptide sequencing that learns the relationship between glycopeptide spectra and fragment ions for both N- and O-glycans, rather than being tied to one fragmentation method. Reports over 92% glycan recall and about 95% precision on N-glycopeptide spectra under both sceHCD and EThcD, and sequences O-glycans with no additional training. Identifies an O-glycopeptide in mouse heart tissue confirmed by exoglycosidase treatment and a synthetic standard.

Seen in the charts

Back to the full map

Back to top