TIMSrescore: timsTOF-optimized PSM rescoring boosts identification rates for immunopeptidomics

Significance of the Topic

The immunoproteome drives antigen presentation and immune response across diseases. High quality peptide identification in immunopeptidomics is often limited by non enzymatic cleavage and expanded search spaces. Novel rescoring approaches can improve detection sensitivity and specificity.

Objectives and Study Overview

• Introduce TIMSrescore, an algorithm tailored for timsTOF PASEF data to enhance peptide spectrum match scoring.
• Evaluate its impact on identification rates in HeLa cell lysates digested with elastase and HLA enriched tumor samples.
• Demonstrate integration of peptide fragmentation, retention time, and ion mobility predictions for rescoring.

Methodology and Instrumentation

Experiments used reduced and alkylated HeLa protein lysate digested by elastase at 37 °C for 3 hours. Peptides were separated on a nanoElute 2 UHPLC with a 22 minute gradient and analyzed on a CSI Ultra coupled to a timsTOF Ultra mass spectrometer. Various TIMS ramp times (100, 150, 200 ms) and collision energies spanning 20–70 eV were tested. HLA class I and II enriched RCC tumor samples were processed by the same pipeline. Data were searched with the Sage engine and rescored using MS2Rescore branch (TIMSrescore) incorporating MS2PIP fragmentation, DeepLC retention time, and IM2Deep collision cross section predictions.

Key Results and Discussion

• HeLa elastase dataset showed on average 37% increase in peptide spectrum matches and 35% more unique peptides after rescoring.
• Immunopeptidomics dataset yielded a 21.8% gain in peptide identifications, highlighting benefits for large search spaces.
• MS2PIP optimization for timsTOF improved median Pearson correlation from 0.53 to 0.88 between predicted and observed spectra.
• Feature analysis revealed complementary contributions of fragmentation, retention time, and ion mobility scores to discriminating target and decoy matches.

Benefits and Practical Applications

TIMSrescore enhances peptide recovery in timsTOF immunopeptidomics workflows without instrument modifications. Increased sensitivity and specificity support more comprehensive antigen profiling in clinical and research settings. The open source integration with Sage and MS2Rescore facilitates adoption across laboratories.

Future Trends and Opportunities

• Expansion of predictive models to novel fragmentation methods and post translational modifications.
• Real time rescoring integration during data acquisition for adaptive acquisition strategies.
• Combined analysis with spectral libraries and machine learning based peak detection.

Conclusion

TIMSrescore provides a robust PSM rescoring workflow optimized for timsTOF data, delivering consistent 15–40% improvements in peptide identifications across diverse immuno and whole cell datasets. Its integration of multi dimensional predicted peptide properties positions it as a key tool for advancing immunopeptidomics.

Reference

1. Hoenisch Gravel N. et al. TOFIMS mass spectrometry based immunopeptidomics refines tumor antigen identification. Nat Commun. 14, 7472 (2023)
2. Lazear M.R. Sage: An open source tool for fast proteomics searching and quantification at scale. J Proteome Res. 22, 3652–3659 (2023)
3. Declercq A. et al. MS2Rescore: Data driven rescoring dramatically boosts immunopeptide identification rates. Mol Cell Proteomics. 21, (2022)