AI proteomics: from protein identification to virtual cells.

Yingying Sun; Jun A; Zhiwei Liu; Rui Sun; Liujia Qian; Samuel H Payne; Wout Bittremieux; Markus Ralser; Chen Li; Yi Chen; Zhen Dong; Yasset Perez-Riverol; Asif Khan; Chris Sander; Ruedi Aebersold; Juan Antonio Vizca&#xed;no; Jonathan R Krieger; Jianhua Yao; Wen Han; Linfeng Zhang; Yunping Zhu; Yue Xuan; Benjamin Boyang Sun; Liang Qiao; Henning Hermjakob; Haixu Tang; Huanhuan Gao; Yamin Deng; Qing Zhong; Cheng Chang; Nuno Bandeira; Ming Li; Weinan E; Siqi Sun; Yuedong Yang; Gilbert S Omenn; Yue Zhang; Ping Xu; Yan Fu; Xiaowen Liu; Christopher M Overall; Yu Wang; Eric W Deutsch; Luonan Chen; J&#xfc;rgen Cox; Vadim Demichev; Fuchu He; Jiaxing Huang; Huilin Jin; Chao Liu; Nan Li; Zhongzhi Luan; Jiangning Song; Kaicheng Yu; Wanggen Wan; Tai Wang; Kang Zhang; Le Zhang; Peter A Bell; Matthias Mann; Bing Zhang; Tiannan Guo

doi:10.1038/s41592-026-03085-y

Home/Archive/PubMed

PubMed · 42521824

AI proteomics: from protein identification to virtual cells.

View original source ↗Export metadata ↓

Abstract

Artificial intelligence (AI) is transforming scientific research, including proteomics. In this Perspective, we highlight key mass spectrometry (MS)-based proteomics areas where AI is driving innovation, ranging from protein identification to building AI virtual cells. These include improving peptide and protein identification and quantification; characterizing protein-protein interactions and protein complexes; advancing spatial and perturbation proteomics; integrating multi-omics data; and, ultimately, enabling AI virtual cells. Finally, we call for global collaboration among data producers, data consumers and other stakeholders to establish an AI-friendly ecosystem for MS-based proteomics, laying the foundation for transformative advancements in proteomics driven by AI.

Explore related subjects

Proteomics Mass Spectrometry Artificial Intelligence Humans Proteins Multiomics Animals Journal Article Review

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Yingying Sun, Jun A, Zhiwei Liu, Rui Sun, Liujia Qian, Samuel H Payne, Wout Bittremieux, Markus Ralser, Chen Li, Yi Chen, Zhen Dong, Yasset Perez-Riverol, Asif Khan, Chris Sander, Ruedi Aebersold, Juan Antonio Vizcaíno, Jonathan R Krieger, Jianhua Yao, Wen Han, Linfeng Zhang, Yunping Zhu, Yue Xuan, Benjamin Boyang Sun, Liang Qiao, Henning Hermjakob, Haixu Tang, Huanhuan Gao, Yamin Deng, Qing Zhong, Cheng Chang, Nuno Bandeira, Ming Li, Weinan E, Siqi Sun, Yuedong Yang, Gilbert S Omenn, Yue Zhang, Ping Xu, Yan Fu, Xiaowen Liu, Christopher M Overall, Yu Wang, Eric W Deutsch, Luonan Chen, Jürgen Cox, Vadim Demichev, Fuchu He, Jiaxing Huang, Huilin Jin, Chao Liu, Nan Li, Zhongzhi Luan, Jiangning Song, Kaicheng Yu, Wanggen Wan, Tai Wang, Kang Zhang, Le Zhang, Peter A Bell, Matthias Mann, Bing Zhang, Tiannan Guo. 2026-07-28. AI proteomics: from protein identification to virtual cells.. https://doi.org/10.1038/s41592-026-03085-y

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2026

Multi-omic analyses of the same sample using metabolomics, lipidomics, proteomics, phosphoproteomics, and glycoproteomics.

Mass spectrometry (MS)-based multi-omics offers powerful tools to comprehensively characterize proteins, post-translational modifications, metabolites, and lipids. However, these measurements are typically performed using separate sample preparation workflows and modality-specific liquid chromatography mass spectrometry (LC-MS) platforms, limiting integration and constraining applications to small amounts of sample materials, especially scarce clinical specimens. Here, we describe a unified nano-LC-MS framework that enables metabolomic, lipidomic, proteomic, phosphoproteomic, and glycoproteomic analyses from the same starting material using a single nano-LC-MS platform, with only the chromatographic conditions, acquisition methods, and enrichment procedures tailored to each omics. This integrated strategy reduces workflow complexity and sample consumption while improves analytical continuity across molecular layers. By enabling deep multi-omics characterization from the same sample, this platform provides a practical foundation for comprehensive analysis of precious clinical samples.

Proteomics↗

PubMed2026

4D-DIA proteomics reveals distinct proteolytic landscapes induced by mechanical stress, Agrobacterium, and a viral capsid precursor.

Nicotiana benthamiana is a widely used platform for plant molecular farming, yet recombinant protein yields are frequently compromised by the host's innate defense mechanisms, particularly proteolytic degradation. While the general effects of Agroinfiltration are known, the distinct contributions of mechanical injury, bacterial perception, and product-specific stress remain poorly resolved. Here we utilized high-depth 4D-DIA proteomics to dissect the host response across three dimensions: physical stress (buffer infiltration), pathogen-associated stress (Agrobacterium), and product-associated stress (GFP vs. the FMDV capsid precursor P1_2A). We demonstrate that buffer infiltration is not a neutral event but an independent inducer of cell wall remodeling and oxidative stress. By filtering out these background effects, we defined a core Agrobacterium-responsive proteome characterized by a growth-defense trade-off. We also expanded the known protease repertoire of N. benthamiana to 1,505 enzymes through improved genomic annotation. We found that the expression of the FMDV capsid precursor P1_2A was associated with a distinct and more pronounced protease profile compared to soluble GFP, characterized by the upregulation of subtilases and cysteine proteases. These findings suggest that host proteolytic responses vary with the recombinant cargo, a factor worth considering when designing engineering strategies for the production of complex biopharmaceuticals in plants.

Proteomics↗

PubMed2026

An Instrumental Optimization of a Label-Free Proteomic Method for Trace Protein Input.

Liquid chromatography-mass spectrometry (LC-MS)-based proteomics of trace-level samples, such as tens of cells or spatially resolved tissue regions, offers unique biological insights but is often constrained by the requirement for specialized, costly instrumentation. In this study, we developed a scalable workflow for the deep proteomic analysis of low- to ultralow-input samples by systematically optimizing a widely adopted Orbitrap and UHPLC platform to maximize sensitivity, precision, and throughput. This optimized workflow identified over 5600 proteins from 5 ng of peptides and 3400 proteins from 20 sorted cells, achieving a throughput of 30 analyses per day while maintaining deep proteome coverage and high quantitative reproducibility. Furthermore, by applying this method to spatially resolved proteomics, we identified over 6100 proteins from microscale regions of interest (ROIs) within a formalin-fixed, paraffin-embedded (FFPE) tissue. A data-driven normalization strategy was employed to correct for variable cellularity across tissue regions, effectively revealing intratumor heterogeneity and distinct molecular and functional signatures, including pathway activations not apparent in parallel spatial transcriptomic analysis. Ultimately, this accessible, high-performance method substantially lowers the instrumentation barrier for the deep proteomic profiling of trace-level biological samples.

Proteomics↗