henhao
Sign in
BiochemistryMolecular & Cellular Proteomics
Peer-reviewed journal article

Higher-Throughput Proteome Profiling Enabled by Parallelized Pre-Accumulation and Optimized Ion Processing in the Orbitrap Astral Zoom Mass Spectrometer

Ulises H. GuzmániD, Martin RykæriD, Ivo Alexander HendriksiD, Hamish I. StewartiD, Eduard V. Denisov, Bernd Hagedorn, J. Petzoldt, Arne KreutzmanniD, Yannick Mueller, Tabiwang N. ArreyiD, Immo Colonius, Ole ØstergaardiD, Claire KoenigiD, Julia Kraegenbring, Kyle L. Fort, Erik CouzijniD, Jan-Peter Hauschild, Daniel J. Hermanson, Vlad ZabrouskoviD, Christian Hock, Eugen DamociD, Jesper.V. OlseniD

Abstract

High-throughput proteomics is critical for understanding biological processes, enabling large-scale studies such as biomarker discovery and systems biology. However, current mass spectrometry technologies face limitations in speed, sensitivity, and scalability for analyzing large sample cohorts. The Thermo Scientific Orbitrap Astral Zoom mass spectrometer (MS) was developed to address these limitations by improving acquisition speed, ion utilization, and spectral processing, which are all essential for advancing proteome depth in high-throughput proteomics. The Orbitrap Astral Zoom MS achieves ultra-fast MS/MS scan rates of up to 270 Hz with enhanced ion utilization through pre-accumulation, enabling the identification of [~]100,000 unique peptides and >8,400 proteins in a single 300 samples-per-day (SPD) analysis of human cell lysate. The optimized system reduces analysis time by 40%, achieves near-complete proteome coverage (>12,000 proteins) in 2.7 hours, and enables ultra-high-throughput workflows, identifying >7,000 proteins in a 500 SPD method with exceptional reproducibility (Pairwise Pearson correlations >0.99). These advancements establish the Orbitrap Astral Zoom MS as a new benchmark in proteomics, significantly enhancing speed, sensitivity, and scalability, paving the way for routine large-scale proteome studies with applications in clinical research and systems biology. TeaserHigh-Speed Human Proteome Analysis using the Orbitrap Astral Zoom Mass Spectrometer

PDFDOI
Full-text digest ↓

02 /Specification & Classification

Click any tag to browse matching papers
Citations
24
FWCI
31.93
Field percentile
top 1%
OpenAlex Topics

In-depth summary

✦ Multi-agent verified full-text synthesis

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (19 Jul 2025) · CC BY 4.0

Orbitrap Astral Zoom upgrades enable 270 Hz DIA scanning, identifying 7184 protein groups at 500 samples/day and >12,000 proteins from fractionated HEK293 in 2.7 h.

Proteomics for large cohorts, systems biology and drug screening needs deep, reproducible coverage across hundreds to thousands of samples, but shorter LC gradients demand faster MS/MS, and the Orbitrap Astral still loses duty cycle to almost 2 ms of scan-to-scan overhead. The Zoom recovers that overhead through parallel pre-accumulation and faster ion handling, gaining speed without shortening ion accumulation.

Key results

  1. ShownShorter scan overheads and parallel bent-trap pre-accumulation (Pre-Ac) raise narrow-window DIA to about 270 Hz and add peptides. Pre-Ac drives most of the gain in fast 1.5 ms methods; enhanced spectral processing (ESP) drives it in long-injection, wide-window methods.

    Median scan-to-scan time 4.32 ms (Astral) vs 3.49 ms (Zoom); ~270 Hz at injection time (IT) ≤1.5 ms vs ~220 Hz at ≤2.5 ms; ~80,000 vs ~67,000 unique peptides (200 ng HEK293, 300 SPD). At 50 ng (6 ms IT, 4 Th, 50 SPD) all features together gave +51% peptides vs Astral; Pre-Ac 31%, high acquisition rate (HAR) 38%, ESP 49% alone. · Optimized ion processing stage enables increased…; Benchmarking the Impact of Hardware and Software…; Figure legends

  2. ShownVery short gradients keep depth and reproducibility. Single HEK293 runs at 300 and 500 SPD yield thousands of protein groups (PGs) with replicate Pearson r of 0.99 or above, and a 32-cell-line panel gives nearly equal coverage at 300 and 180 SPD.

    200 ng HEK293: 300 SPD gave ~100,000 peptides and 8,400 PGs, r 0.99; 500 SPD (2.2 min gradient) gave 7184 PGs from ~67,000 peptides, >15,000 more peptides than Astral, r >0.99. 32 cell lines in quadruplicate: mean 9645 (180 SPD) vs 9549 (300 SPD) proteins, >11,500 overall. · Fast and deep proteome profiling with Orbitrap…; A Comprehensive Draft of the Human Proteome…; Figure legends

  3. ShownOffline high-pH fractionation into 34 fractions, run at 300 SPD, reaches near-complete HEK293 proteome coverage in 2.7 h. Sequence coverage is high enough to detect phosphosites and protein variants without enrichment.

    >12,100 PGs from 34 fractions at 300 SPD in 2.7 h vs 10,247 proteins in a 1 h single shot (24 SPD); twofold faster than the Orbitrap Astral (~4.5 h) and fourteenfold faster than Q Exactive HF-X; 2089 phosphosites (~85% Ser, ~15% Thr, <1% Tyr); 935 splicing events, 743 single amino acid variants (SAVs), 11 indels. · Optimized Fractionation Strategy for Rapid and…

  4. ShownIn a human/yeast/E. coli label-free quantification (LFQ) benchmark, faster scanning added protein identifications and did not degrade ratio accuracy relative to the Orbitrap Astral.

    50 SPD, 2 Th, IT capped at 1.5 ms on both instruments (270 vs 220 Hz), 200 ng, technical triplicates: Zoom averaged ~14,000 PGs, >1,000 more than Astral. Median SD of measured vs theoretical ratios was equal or lower. Zoom also outperformed Astral when both were capped at 2.5 ms. · Evaluation of the Accuracy and Precision of LFQ…; Figure legends

  5. ShownEnhanced dynamic range (eDR) MS1 splits the survey scan into segments with different accumulation times. It raises DDA precursor identifications well beyond the modest gains that Zoom hardware alone gives in DDA.

    MS1 signal-to-noise up to 20-fold higher; median MS1 peaks rose from 3742 to 4571. In 100 SPD DDA of 10 ng HEK293, eDR gave ~15,000 and ~30,000 more precursors than standard and wide-window DDA on the Astral, sustaining ~150 MS/MS scans/s and >80k spectra at 3 ms MS2 IT. · Optimized ion processing stage enables increased…; Enhancing MS1 Dynamic Range via Multiplexed Ion…; Figure legends

Methods, data and reproducibility

A prototype Orbitrap Astral Zoom adds faster quadrupoles, bent-trap ion pre-accumulation that runs in parallel with downstream steps, faster ion transfer, enhanced spectral processing (ESP) and segmented-injection MS1 (eDR). Features are benchmarked against the Orbitrap Astral with HEK293 and three-species digests in DIA and DDA, and a 32-cell-line panel is profiled on the Zoom at 180 and 300 SPD.

  • HEK293 tryptic digest (mostly 200 ng; 20 to 250 ng dilutions; 10 ng for DDA) for injection-time titration, feature dissection and 300/500 SPD runs · human · n = three (300 SPD) and four (500 SPD) technical replicates
  • Three-species LFQ benchmark: HeLa (Pierce 88328), yeast (Promega V7461) and E. coli (Waters 186003196) digests mixed as E05-H50-Y45, E25-H50-Y25, E45-H50-Y05 · human; yeast; E. coli · n = three mixing ratios, technical triplicates, 200 ng
  • HEK293 peptides (200 µg) separated by offline high-pH reversed-phase HPLC into 24, 34 and 46 fractions, each analyzed at 180 or 300 SPD · human · n = 24, 34 and 46 fractions; 200 ng per fraction
  • Panel of human cell lines representing cell types of the major organs, profiled by DIA at 180 and 300 SPD · human · n = 32 cell lines, technical quadruplicates
  • Earlier Orbitrap Astral 34-fraction HEK293 dataset at 180 SPD (Guzman et al., 2024), used as comparator · human · n = 34 fractions
  • UniProt human reference proteome (2024 release) plus common contaminants; separate three-species database for LFQ searches · human (plus yeast and E. coli in LFQ database) · n = 20,417 + 246 sequences; 31,657 sequences (three-species)
  • In-house empirical phosphopeptide library from 12 HpH fractions of phosphopeptide-enriched material, used to search the 34-fraction data · n = 119,793 precursors
  • Protein variant database translated from HEK293 transcriptomic variant information, used for splicing, SAV and indel searches · human
Methods & tools
Orbitrap Astral Zoom MS prototype (Pre-Ac, HAR, ESP, eDR), Orbitrap Astral MS (baseline), Narrow-window DIA, DDA and wide-window DDA acquisition, Vanquish Neo UHPLC; Evosep One with Eno module, Offline high-pH reversed-phase fractionation (UltiMate 3000 HPLC), Protein Aggregation Capture digestion on KingFisher, DIA-NN 1.9.2, Spectronaut v.19 (directDIA+), MaxLFQ via iq R package, MaxQuant v.1.6.7.0 and v.1.6.14.0 (MS1 features), Perseus peptide-collapse plugin; Protein Coverage Summarizer, clusterProfiler with org.Hs.eg.db
Reagents & environment
DIA-NN 1.9.2; Spectronaut v.19 (directDIA+); MaxQuant v.1.6.7.0 and v.1.6.14.0; R v.4.2.2 with RStudio v.2022.12.0 Build 353; UniProt human FASTA, 2024 release (20,417 sequences) plus 246 contaminants.
Reproducibility gaps
No raw-data repository accession or code link is given. In-house Python scripts (Thermo Rawfile Reader) are not shared. LC columns and methods are deferred to an unnumbered Supplementary Table X. All Zoom measurements used a prototype instrument.
Limitations (5)
  • Ultra-short gradients such as 300 and 500 SPD may miss low-abundance proteins (e.g., cell type-specific transcription factors) in complex samples, and the narrow precursor peaks of short chromatography can hinder precise quantification. · stated by the authors
  • Making use of the higher scan rate depends on downstream analysis software that can handle the larger data volumes produced. · stated by the authors
  • Performance on sample types beyond standard cell lines and model organisms, PTM analysis with ultra-fast scanning, and longitudinal studies of dynamic proteomes remain open questions the authors say need fuller evaluation. · stated by the authors
  • The instrument is a prototype, benchmarked mainly against its predecessor on commercial or cell-line digests by a team that includes manufacturer employees. Independent or cross-vendor comparisons are absent. · noted in this summary
  • Reporting inconsistencies hinder replication: methods cite an unnumbered 'Supplementary Table X', replicate counts differ (n = 4 in methods vs n = 5 in a figure legend), and the introduction claims ~1,600 splicing variants while results report 935 splicing events. · noted in this summary

Who should read:Read this if you run or plan high-throughput LC-MS proteomics cohorts and want to know how Orbitrap Astral Zoom pre-accumulation, scan speed and spectral processing affect depth at 300 to 500 samples per day.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISHsu '25Stewart '23Harking '25Heil '23Sandow '21Schebesta '25
79% semantic matchBiochemistry
Chris Hsu, Nicholas Shulman, Hamish I. Stewart+14 more·Journal of Proteome Research·Oct 22, 2025· 21 citations