Deep learning-driven fragment ion series classification enables highly precise and sensitive de novo peptide sequencing

preprint · bioRxiv · 2023

preprint · bioRxiv · 2023. Daniela Klaproth-Andrade et al. Unlike for DNA and RNA, accurate and high-throughput sequencing methods for proteins are lacking, hindering…
Date 2023-01-05
Type preprint
Venue bioRxiv
Publisher Cold Spring Harbor Laboratory
Contribution post-processor
DOI 10.1101/2023.01.05.522752
Citations (OpenAlex) 8

Peer-reviewed version: Deep learning-driven fragment ion series classification enables highly precise and sensitive de novo peptide sequencing (2024-01-02, Nature Communications)

Abstract

Unlike for DNA and RNA, accurate and high-throughput sequencing methods for proteins are lacking, hindering the utility of proteomics in applications where the sequences are unknown including variant calling, neoepitope identification, and metaproteomics. We introduce Spectralis, a new de novo peptide sequencing method for tandem mass spectrometry. Spectralis leverages several innovations including a new convolutional neural network layer connecting peaks in spectra spaced by amino acid masses, proposing fragment ion series classification as a pivotal task for de novo peptide sequencing, and a new peptide-spectrum confidence score. On spectra for which database search provided a ground truth, Spectralis surpassed 40% sensitivity at 90% precision, nearly doubling state-of-the-art sensitivity. Application to unidentified spectra confirmed its superiority and showcased its applicability to variant calling. Altogether, these algorithmic innovations and the substantial sensitivity increase in the high-precision range constitute an important step toward broadly applicable peptide sequencing.

Authors

  1. Daniela Klaproth-Andrade · Technical University of Munich
  2. Johannes Hingerl · Technical University of Munich
  3. Nicholas H. Smith · Technical University of Munich
  4. Jakob Träuble · Technical University of Munich
  5. Mathias Wilhelm · Technical University of Munich
  6. Julien Gagneur · Helmholtz Center Munich, Technical University of Munich

Methods and tools

Trained on

Evaluated on

Cites (20)

Cited by (3)

Seen in the charts

Back to the full map

Back to top