Pretrained spectrum representations for diverse tandem mass spectrometry proteomics tasks

peer-reviewed · Molecular & Cellular Proteomics · 2026

peer-reviewed · Molecular & Cellular Proteomics · 2026. Justin Sanders et al. Mass spectrometry is the dominant technology in the field of proteomics, enabling high-throughput analysis of…
Date 2026-10-01
Type peer-reviewed
Venue Molecular & Cellular Proteomics
Publisher Elsevier BV
Contribution algorithm
DOI 10.1016/j.mcpro.2026.101673
Citations (OpenAlex) 0
Venue 2-year citedness 4.69

Preprint version: Foundation model for mass spectrometry proteomics (2025-05-16, arXiv)

Abstract

Mass spectrometry is the dominant technology in the field of proteomics, enabling high-throughput analysis of the protein content of complex biological samples. Due to the complexity of the instrumentation and resulting data, sophisticated computational methods are required for the processing and interpretation of acquired mass spectra. Machine learning has shown great promise to improve the analysis of mass spectrometry data, with numerous purpose-built methods for improving specific steps in the data acquisition and analysis pipeline reaching widespread adoption. Each of these tasks rely on a shared fundamental understanding of the information present in a mass spectrum. Here, we propose a transfer learning framework, CasanovoTL, which utilizes pretrained spectrum representations, derived from a de novo sequencing model trained on a massive proteomics dataset, rather than training a spectrum encoder from scratch for each task. We show that using these pre-trained spectrum representations significantly improves our performance on the four downstream tasks of spectrum quality prediction, chimericity prediction, phosphorylation prediction, and glycosylation status prediction. Finally, we perform end-to-end finetuning, as well as multi-task training, and find that finetuned spectrum representations further improve performance on each individual task. Overall, our work demonstrates that a pretrained spectrum encoder for tandem mass spectrometry proteomics trained on de novo sequencing learns generalizable representations of spectra, improves performance on downstream tasks where training data is limited, and can ultimately enhance data acquisition and analysis in proteomics experiments.

Authors

  1. Justin Sanders · University of Washington
  2. Melih Yilmaz · University of Washington
  3. Jacob H. Russell · University of Washington
  4. Wout Bittremieux · Biomedical Informatics Research Center Antwerp, Indiana University, University of Antwerp, University of California San Diego, University of Washington
  5. William E. Fondrie · Talus Bioscience
  6. Nicholas M. Riley · University of Washington
  7. Sewoong Oh · University of Washington
  8. William Stafford Noble · University of Washington

Methods and tools

  • Casanovo Foundation: Foundation model for tandem-MS proteomics: pre-trained the Casanovo spectrum encoder on 30M labelled spectra from MassIVE-KB, then reused the encoder off-the-shelf for downstream tasks (de novo sequencing, spectrum quality, chimericity, phosphorylation, glycosylation prediction). Successor in spirit to Casanovo v1/v2/v5. The peer-reviewed version (MCP 2026) names the framework CasanovoTL.

Trained on

Evaluated on

Data used

Data this paper ran on, where the catalog has not recorded whether for training or for evaluation.

Cites (6)

Seen in the charts

Back to the full map

Back to top