Real-time de novo sequencing of peptide antigens using Bruker ProteoScape™ for 'Run & Done' 4D-immunopeptidomics

Importance of the Topic

The study addresses the growing need for ultra-fast, accurate peptide identification in immunopeptidomics workflows. By integrating a real-time de novo sequencing algorithm directly into the acquisition pipeline of a timsTOF Pro 2 mass spectrometer, users can obtain peptide assignments as data are collected. Immediate feedback on instrument performance and sample quality streamlines decision-making in proteomics laboratories and supports high-throughput immunopeptide discovery.

Objectives and Study Overview

This work aimed to implement and benchmark a timsTOF-optimized de novo sequencing module (BPS Novor) within Bruker ProteoScape™. Specific goals included:

• Training the algorithm on timsTOF-generated spectra from multiple protease digests.
• Evaluating real-time processing speed to match or exceed data acquisition rates.
• Comparing identification precision and recall with existing software and standard de novo workflows.
• Demonstrating practical application in immunopeptidomics of a mouse colon cancer cell line.

Methodology and Used Instrumentation

Data were generated on a timsTOF Pro 2 instrument employing PASEF for high-speed MS/MS acquisition (>150 Hz). The BPS Novor module was trained using ProLuCID-GPU database search results filtered at 1% PSM FDR, covering tryptic and non-tryptic (GluC, Pepsin, Elastase, Chymotrypsin) digests of K562 lysate. The trained model was deployed in ProteoScape during acquisition, streaming MS/MS spectra to BPS Novor for on-the-fly de novo interpretation.

Key Results and Discussion

Processing performance:

• Average throughput of 1,338 ± 226 spectra/second, enabling completion of >130,000 spectra in under three minutes.
• Processing speeds consistently kept pace with acquisition at PASEF scan rates.

Identification accuracy:

• Trained BPS Novor achieved >70% amino acid precision at 75% recall, outperforming the non-optimized algorithm (∼45% recall).
• Peptide-level correct assignments exceeded 60% with de novo score thresholds above 70.

Immunopeptidomics application:

• Analysis of CT26 colon cancer HLA-bound peptides yielded similar length distributions and motif patterns to database search results.
• Precision-recall profiles confirmed competitive performance against two commercial de novo tools.
• Real-time feedback allowed immediate evaluation of sample preparation quality and instrument stability.

Benefits and Practical Applications

The integrated real-time de novo sequencing offers several advantages for proteomics and immunopeptidomics:

• Accelerated decision-making on sample handling and experimental parameters.
• Elimination of lengthy offline data processing steps.
• Enhanced sensitivity for discovery of non-canonical and post-translationally modified peptides.

Future Trends and Opportunities

Real-time sequencing modules are expected to expand into additional 4D-proteomics applications, including top-down workflows and targeted glycopeptide analysis. Integration with machine learning-driven confidence scoring and adaptive acquisition strategies may further boost identification rates and streamline clinical research pipelines.

Conclusion

BPS Novor, embedded within Bruker ProteoScape and coupled to timsTOF PASEF, achieves rapid, high-accuracy de novo peptide sequencing in real time. This “Run & Done” platform accelerates immunopeptidomics research by delivering immediate peptide assignments and supports timely, data-driven decisions in high-throughput proteomic workflows.