A large-scale unified deep learning model for peptide mass spectrum interpretation trained on multimodal data

Machine Learning


  • Aebersold, R. & Mann, M. Mass-spectrometric exploration of proteome structure and function. Nature 537, 347–355 (2016).

    Article 

    Google Scholar 

  • Eng, J. K., McCormack, A. L. & Yates, J. R. An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. JAm. Soc. Mass Spectrom. 5, 976–989 (1994).

    Article 

    Google Scholar 

  • Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized ppb-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 26, 1367–1372 (2008).

    Article 

    Google Scholar 

  • Chi, H. et al. Comprehensive identification of peptides in tandem mass spectra using an efficient open search engine. Nat. Biotechnol. 36, 1059–1061 (2018).

    Article 

    Google Scholar 

  • Strauss, M. T. et al. AlphaPept: a modern and open framework for MS-based proteomics. Nat. Commun. 15, 2168 (2024).

    Article 

    Google Scholar 

  • Kong, A. T., Leprevost, F. V., Avtonomov, D. M., Mellacheruvu, D. & Nesvizhskii, A. I. MSFragger: ultrafast and comprehensive peptide identification in mass spectrometry–based proteomics. Nat. Methods 14, 513–520 (2017).

    Article 

    Google Scholar 

  • Yang, H., Chi, H., Zeng, W.-F., Zhou, W.-J. & He, S.-M. pNovo 3: precise de novo peptide sequencing using a learning-to-rank framework. Bioinformatics 35, 183–190 (2019).

    Article 

    Google Scholar 

  • Ma, B. et al. PEAKS: powerful software for peptide de novo sequencing by tandem mass spectrometry. Rapid Commun. Mass Spectrom. 17, 2337–2342 (2003).

    Article 

    Google Scholar 

  • Frank, A. & Pevzner, P. PepNovo: de novo peptide sequencing via probabilistic network modeling. Anal. Chem. 77, 964–973 (2005).

    Article 

    Google Scholar 

  • Yilmaz, M., Fondrie, W., Bittremieux, W., Oh, S. & Noble, W. S. De novo mass spectrometry peptide sequencing with a transformer model. Proc. Mach. Learn. Res. 162, 25514–25522 (2022).

  • Schulte, D. & Snijder, J. A handle on mass coincidence errors in de novo sequencing of antibodies by bottom-up proteomics. J. Proteome Res. 23, 3552–3559 (2024).

    Article 

    Google Scholar 

  • Wen, B. & Zhang, B. PepQuery2 democratizes public MS proteomics data for rapid peptide searching. Nat. Commun. 14, 2213 (2023).

    Article 

    Google Scholar 

  • Mann, M., Kumar, C., Zeng, W.-F. & Strauss, M. T. Artificial intelligence for proteomics and biomarker discovery. Cell Syst. 12, 759–770 (2021).

    Article 

    Google Scholar 

  • Zhou, X.-X. et al. pDeep: predicting MS/MS spectra of peptides with deep learning. Anal. Chem. 89, 12690–12697 (2017).

    Article 

    Google Scholar 

  • Gessulat, S. et al. Prosit: proteome-wide prediction of peptide tandem mass spectra by deep learning. Nat. Methods 16, 509–518 (2019).

    Article 

    Google Scholar 

  • Ma, C. et al. Improved peptide retention time prediction in liquid chromatography through deep learning. Anal. Chem. 90, 10881–10888 (2018).

    Article 

    Google Scholar 

  • Zhou, W.-J., Wei, Z.-H., He, S.-M. & Chi, H. pValid 2: a deep learning based validation method for peptide identification in shotgun proteomics with increased discriminating power. J. Proteom. 251, 104414 (2022).

    Article 

    Google Scholar 

  • Zeng, W.-F. et al. AlphaPeptDeep: a modular deep learning framework to predict peptide properties for proteomics. Nat. Commun. 13, 7238 (2022).

    Article 

    Google Scholar 

  • Yang, K. L. et al. MSBooster: improving peptide identification rates using deep learning-based features. Nat. Commun. 14, 4539 (2023).

    Article 

    Google Scholar 

  • Yu, Y. & Li, M. Towards highly sensitive deep learning-based end-to-end database search for tandem mass spectrometry. Nat. Mach. Intell. 7, 85–95 (2025).

  • Tran, N. H., Zhang, X., Xin, L., Shan, B. & Li, M. De novo peptide sequencing by deep learning. Proc. Natl Acad. Sci. USA 114, 8247–8252 (2017).

    Article 

    Google Scholar 

  • Qiao, R. et al. Computationally instrument-resolution-independent de novo peptide sequencing for high-resolution devices. Nat. Mach. Intell. 3, 420–425 (2021).

    Article 

    Google Scholar 

  • Yilmaz, M. et al. Sequence-to-sequence translation from mass spectra to peptides with a transformer model. Nat. Commun. 15, 6427 (2024).

    Article 

    Google Scholar 

  • Mao, Z., Zhang, R., Xin, L. & Li, M. Mitigating the missing-fragmentation problem in de novo peptide sequencing with a two-stage graph-based deep learning model. Nat. Mach. Intell. 5, 1250–1260 (2023).

    Article 

    Google Scholar 

  • Zhang, X. et al. π-PrimeNovo: an accurate and efficient non-autoregressive deep learning model for de novo peptide sequencing. Nat. Commun. 16, 267 (2025).

    Article 

    Google Scholar 

  • Li, J., Li, D., Xiong, C. & Hoi, S. BLIP: bootstrapping language–image pre-training for unified vision–language understanding and generation. Proc. Mach. Learn. Res. 162, 12888–12900 (2022).

  • Zhao, J., Zhang, Z., Li, Y. & Cen, L. Multidimensional information assisted deep learning realizing flexible recognition of vortex beam modes. IEEE Photon. J. 13, 1–6 (2021).

    Google Scholar 

  • Hayes, T. et al. Simulating 500 million years of evolution with a language model. Science https://doi.org/10.1126/science.ads0018 (2024).

  • Zhao, J., Zhuang, W., Song, J., Li, Y. & Lu, S. Pre-training protein bi-level representation through span mask strategy on 3D protein chains. Proc. Mach. Learn. Res. 235, 1–15 (2024).

  • Zhou, G. et al. Uni-Mol: a universal 3D molecular representation learning framework. The Eleventh International Conference on Learning Representations (2023).

  • Tran, N. H. et al. NovoBoard: a comprehensive framework for evaluating the false discovery rate and accuracy of de novo peptide sequencing. Mol. Cell. Proteom. 23, 100849 (2024).

  • Burq, M. et al. Tesorai Search: large pretrained model boosts identifications in mass spectrometry proteomics without the need for Percolator. Preprint at bioRxiv https://doi.org/10.1101/2024.08.19.606805 (2024).

  • Jun, A. et al. DDA-BERT: leveraging transformer architecture pre-training for data-dependent acquisition mass spectrometry-based proteomics. Nat. Commun. https://doi.org/10.1038/s41467-026-72246-6 (2026).

  • Wang, K.-F., Wu, Y.-Z. & Chi, H. A universal database reduction method based on the sequence tag strategy to facilitate large-scale database search in proteomics. Int. J. Mass Spectrom. 483, 116966 (2023).

    Article 

    Google Scholar 

  • Yang, H. et al. Open-pNovo: de novo peptide sequencing with thousands of protein modifications. J. Proteome Res. 16, 645–654 (2017).

    Article 

    Google Scholar 

  • Gavin, P. G. et al. Intestinal metaproteomics reveals host–microbiota interactions in subjects at risk for type 1 diabetes. Diabetes Care 41, 2178–2186 (2018).

    Article 

    Google Scholar 

  • Bassani-Sternberg, M. et al. Direct identification of clinically relevant neoepitopes presented on native human melanoma tissue by mass spectrometry. Nat. Commun. 7, 13404 (2016).

    Article 

    Google Scholar 

  • Wang, P. et al. A large-scale proteogenomic atlas of pear. Mol. Plant 16, 599–615 (2023).

    Article 

    Google Scholar 

  • Cao, Z., Qin, T., Liu, T.-Y., Tsai, M.-F. & Li, H. Learning to rank: from pairwise approach to listwise approach. In Proc. 24th International Conference on Machine Learning 129–136 (Association for Computing Machinery, 2007).

  • Radford, A. et al. Learning transferable visual models from natural language supervision. Proc. Mach. Learn. Res. 139, 8748–8763 (2021).

  • Yifan, W., Doersch, C., Arandjelović, R., Carreira, J. & Zisserman, A. Input-level inductive biases for 3D reconstruction. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition 6176–6186 (IEEE, 2022).

  • Ridnik, T. et al. Asymmetric loss for multi-label classification. Proc. IEEE/CVF International Conference on Computer Vision 82–91 (IEEE, 2021).

  • Cheng, K. et al. Metalab 2.0 enables accurate post-translational modifications profiling in metaproteomics. J. Am. Soc. Mass Spectrom. 31, 1473–1482 (2020).

    Article 

    Google Scholar 

  • Wen, B. et al. Assessment of false discovery rate control in tandem mass spectrometry analysis using entrapment. Nat. Methods 22, 1454–1463 (2025).

  • Potgieter, M. G. et al. Metanovo: an open-source pipeline for probabilistic peptide discovery in complex metaproteomic datasets. PLOS Comput. Biol. 19, 1011163 (2023).

    Article 

    Google Scholar 

  • Wilburn, D. B. et al. Deep learning from harmonized peptide libraries enables retention time prediction of diverse post translational modifications. Preprint at bioRxiv https://doi.org/10.1101/2023.05.30.542978 (2023).

  • Jurtz, V. et al. NetMHCpan-4.0: improved peptide–MHC class I interaction predictions integrating eluted ligand and peptide binding affinity data. J. Immunol. 199, 3360–3368 (2017).

    Article 

    Google Scholar 

  • Gfeller, D. et al. The length distribution and multiple specificity of naturally presented HLA-I ligands. J. Immunol. 201, 3705–3716 (2018).

    Article 

    Google Scholar 

  • Sharma, K. et al. Cell type- and brain region-resolved mouse brain proteome. Nat. Neurosci. 18, 1819–1831 (2015).

    Article 

    Google Scholar 

  • Wilhelm, M. et al. Mass-spectrometry-based draft of the human proteome. Nature 509, 582–587 (2014).

    Article 

    Google Scholar 

  • Zolg, D. P. et al. Building ProteomeTools based on a complete synthetic human proteome. Nat. Methods 14, 259–262 (2017).

    Article 

    Google Scholar 

  • Chick, J. M. et al. A mass-tolerant database search identifies a large proportion of unassigned spectra in shotgun proteomics as modified peptides. Nat. Biotechnol. 33, 743–749 (2015).

    Article 

    Google Scholar 

  • Marcu, A. et al. HLA ligand atlas: a benign reference of HLA-presented peptides to improve T-cell-based cancer immunotherapy. J. Immunother. Cancer 9, 002071 (2021).

    Article 

    Google Scholar 

  • Liu, Z., Wang, F., Chen, J., Zhou, Y. & Zou, H. Modulating the selectivity of affinity absorbents to multi-phosphopeptides by a competitive substitution strategy. J. Chromatogr. A 1461, 35–41 (2016).

    Article 

    Google Scholar 

  • Piersma, S. R. et al. Feasibility of label-free phosphoproteomics and application to base-line signaling of colorectal cancer cell lines. J. Proteom. 127, 247–258 (2015).

    Article 

    Google Scholar 

  • Marx, H. et al. A large synthetic peptide and phosphopeptide reference library for mass spectrometry-based proteomics. Nat. Biotechnol. 31, 557–564 (2013).

    Article 

    Google Scholar 

  • Bekker-Jensen, D. B. et al. An optimized shotgun strategy for the rapid generation of comprehensive human proteomes. Cell Syst. 4, 587–599 (2017).

    Article 

    Google Scholar 

  • Müller, J. B. et al. The proteome landscape of the kingdoms of life. Nature 582, 592–596 (2020).

    Article 

    Google Scholar 

  • Kim, M.-S. et al. A draft map of the human proteome. Nature 509, 575–581 (2014).

    Article 

    Google Scholar 

  • Schuster, H. et al. The immunopeptidomic landscape of ovarian carcinomas. Proc. Natl Acad. Sci. USA 114, 9942–9951 (2017).

    Article 

    Google Scholar 

  • Racle, J. et al. Machine learning predictions of MHC-II specificities reveal alternative binding mode of class II epitopes. Immunity 56, 1359–1375 (2023).

    Article 

    Google Scholar 

  • Guo, X., Trudgian, D. C., Lemoff, A., Yadavalli, S. & Mirzaei, H. Confetti: a multiprotease map of the HeLa proteome for comprehensive proteomics. Mol. Cell. Proteom. 13, 1573–1584 (2014).

    Article 

    Google Scholar 

  • Wang, Y. et al. Multi-protease strategy identifies three PE2 missing proteins in human testis tissue. J. Proteome Res. 16, 4352–4363 (2017).

    Article 

    Google Scholar 

  • Sun, J. et al. Multiproteases combined with high-pH reverse-phase separation strategy verified fourteen missing proteins in human testis tissue. J. Proteome Res. 17, 4171–4177 (2018).

    Article 

    Google Scholar 

  • Gueto-Tettay, C. et al. Multienzyme deep learning models improve peptide de novo sequencing by mass spectrometry proteomics. PLOS Comput. Biol. 19, 1010457 (2023).

    Article 

    Google Scholar 

  • Marcu, A. et al. The HLA Ligand Atlas—a resource of natural HLA ligands presented on benign tissues. Preprint at bioRxiv https://doi.org/10.1101/778944 (2020).

  • Prianichnikov, N. et al. MaxQuant software for ion mobility enhanced shotgun proteomics. Mol. Cell. Proteom. 19, 1058–1069 (2020).

    Article 

    Google Scholar 

  • Meier, F. et al. Online parallel accumulation–serial fragmentation (pasef) with a novel trapped ion mobility mass spectrometer. Mol. Cell. Proteom. 17, 2534–2545 (2018).

    Article 

    Google Scholar 

  • Wigger, L. et al. Multi-omics profiling of living human pancreatic islet donors reveals heterogeneous beta cell trajectories towards type 2 diabetes. Nat. Metab. 3, 1017–1031 (2021).

    Article 

    Google Scholar 

  • Kim, M. L. et al. Hydroxychloroquine inhibits the mitochondrial antioxidant system in activated T cells. iScience 24, 103509 (2021).

  • Zhao, J. pUniFind training set. Zenodo https://doi.org/10.5281/zenodo.18976195 (2026).

  • Zhao, J. pUniFind. Zenodo https://doi.org/10.5281/zenodo.19052067 (2026).

  • Zhao, J. et al. pUniFind: a unified, large-scale pretrained model for pushing the limits of mass spectra interpretation. Code Ocean https://doi.org/10.24433/CO.8567692.v1 (2026).



  • Source link