What Is SEQUEST? Tandem MS Search Engine Guide

SEQUEST is a tandem mass spectrometry (MS/MS) database search algorithm and software engine used in bottom-up proteomics to identify peptides and proteins from experimental fragmentation spectra. Developed in 1994 by Jimmy Eng, Ashley McCormack, and John Yates III at the Scripps Research Institute and the University of Washington, it correlates observed mass spectra with theoretical fragmentation patterns derived from protein sequence databases.

Before examining the proteomics search engine in detail, researchers frequently encounter the term “sequestration” across other academic and administrative domains. In simple terms, “sequester” means to isolate, separate, or set apart. Common distinct uses include federal budget sequestration (statutory, across-the-board spending reductions enacted under the Budget Control Act), legal sequestration (isolating a jury or assets during litigation), and biological or medical sequestration (pathological tissue isolation such as pulmonary sequestration, or biochemical isolation of ions and carbon). In analytical chemistry and proteomics, however, SEQUEST refers strictly to the algorithmic software used for peptide-spectrum matching.

What It Is Used for in Research

Modern bottom-up proteomics relies on liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) to analyze complex biological samples. In these experiments, proteins are extracted from cells or tissues, chemically reduced and alkylated, and digested into peptides using a sequence-specific protease such as trypsin. As the peptide mixture elutes through liquid chromatography, the mass spectrometer performs two sequential measurements:

• MS1 Full Scan: The instrument measures the intact mass-to-charge ratio (m/z) and intensity of precursor peptide ions entering the mass analyzer.
• MS2 Tandem Scan: Selected precursor ions are isolated, fragmented (typically through collision-induced dissociation or higher-energy collisional dissociation), and analyzed to record the masses of the resulting fragment ions.

A single high-throughput LC-MS/MS experiment generates tens of thousands of individual tandem mass spectra. Because manual spectral interpretation is impractical, automated search engines like SEQUEST match experimental spectra against in silico digested protein sequence databases (typically supplied in FASTA format). The algorithm determines which peptide sequence best matches each observed fragmentation pattern, establishing a peptide-spectrum match (PSM).

SEQUEST is used across biomedical research to identify post-translational modifications (PTMs), characterize protein complexes, quantify relative protein abundance across cellular states, and validate recombinant protein expression. Downstream of SEQUEST, identified protein accessions are mapped into analytical tools like the STRING database for protein-protein interaction networks and recorded within institutional scientific data management systems (SDMS). To learn more about modern mass spectrometry workflows, see our guide on protein sequencing by mass spectrometry.

Install or Access (Official Source Only)

Researchers access the SEQUEST algorithm through commercial and open-source distributions:

• Thermo Scientific Proteome Discoverer: The commercial implementation of the algorithm, designated SEQUEST HT, is integrated as a core search node within Thermo Scientific Proteome Discoverer software. Access requires obtaining Proteome Discoverer through official Thermo Fisher Scientific distribution channels. The software runs in 64-bit Microsoft Windows environments and requires a multi-core workstation processor with sufficient system RAM to support in-memory FASTA database indexing and multithreaded search execution.
• Comet (Open-Source Reference Implementation): Researchers requiring an open-source, scriptable, or cross-platform deployment can access Comet, developed by original SEQUEST co-author Jimmy K. Eng at the University of Washington Proteomics Resource. Comet is freely distributed under the Apache License 2.0 and available on its official GitHub repository (https://github.com/UWPR/Comet). It runs natively on Linux, macOS, and Windows, accepting standard mzML and mzXML data formats.

Step-by-Step Task: Running a SEQUEST HT Search in Proteome Discoverer

The following numbered workflow details how to configure and execute a database search using the Sequest HT node in Proteome Discoverer:

1. Create a Study: Open Proteome Discoverer and select File > New Study. Name the study and assign an experiment directory.
2. Add Raw Data: In the Study Definition panel, open the Input Files tab. Click Add Files and load the target instrument raw data files (.raw).
3. Open Workflow Editor: Click the Workflows tab and open the Processing Workflow Editor.
4. Add Search Engine Node: In the left-hand tool palette, expand the Search Engines category. Drag the Sequest HT node onto the workflow design canvas and connect its input port to the Spectrum Selector node.
5. Select Sequence Database: In the Sequest HT parameters pane, locate the Input Data group. Set the Protein Database parameter to the target organism FASTA database (for example, a UniProtKB/Swiss-Prot canonical reference proteome).
6. Define Digestion Parameters: Set Enzyme Name to the protease used during sample preparation (such as Trypsin (Full)). Set Max. Missed Cleavage Sites to 2 (adjustable between 0 and 12 depending on digestion completeness). For detailed comparative cleavage properties, refer to our comparison of trypsin vs chymotrypsin.
7. Configure Mass Tolerances: Under Tolerances, configure instrument-specific windows:
   • Set Precursor Mass Tolerance to 10 ppm (for high-resolution Orbitrap MS1 scans) or the appropriate dalton window for low-resolution instruments.
   • Set Fragment Mass Tolerance to 0.02 Da for high-energy collisional dissociation (HCD) measured in an Orbitrap analyzer, or 0.6 Da for collision-induced dissociation (CID) measured in an ion trap.
8. Specify Chemical Modifications: In the Modifications section:
   • Add Static Modifications: Configure Carbamidomethyl / +57.021 Da on Cysteine (C) if samples were alkylated with iodoacetamide.
   • Add Dynamic Modifications: Configure Oxidation / +15.995 Da on Methionine (M) to account for spontaneous sample oxidation.
9. Connect Decoy Validation: From the Post-Processing node palette, drag the Percolator node onto the canvas and connect the output of Sequest HT to Percolator to execute target-decoy statistical scoring.
10. Execute the Search: Click Workflow > Run (or the green execution arrow in the top toolbar) to begin the automated database search.

Output and How to Read It

Upon search completion, Proteome Discoverer generates an interactive results report organized into hierarchical tables for Protein Groups, Peptide Groups, and Peptide-Spectrum Matches (PSMs). Understanding SEQUEST output requires interpreting its core mathematical scoring metrics:

• XCorr (Cross-Correlation Score): The primary numerical metric generated by SEQUEST. XCorr measures the statistical correlation between the experimental MS/MS spectrum and the theoretical fragment ion spectrum (principally b- and y-ion series) predicted from the candidate sequence. A higher XCorr indicates greater spectral agreement. In legacy workflows, static thresholds were applied based on precursor charge state (historically XCorr ≥ 1.5 for 1+, ≥ 2.0 for 2+, and ≥ 2.5 for 3+). In modern pipelines, raw XCorr values serve as features for machine-learning validation algorithms.
• DeltaCn (ΔCn): The normalized difference in cross-correlation score between the top-ranked candidate peptide and the second-highest candidate peptide for a given spectrum:

  $$Deltatext{Cn} = frac{text{XCorr}_{text{top}} – text{XCorr}_{text{second}}}{text{XCorr}_{text{top}}}$$
  
  A ΔCn value close to 1.0 indicates an unambiguous match where the top candidate scored far higher than any alternative. A ΔCn value approaching 0.0 indicates that multiple distinct sequences in the database produced nearly identical scores, highlighting ambiguity often caused by sequence homology or isobaric modifications.

• Sp (Preliminary Score) and RSp (Preliminary Score Rank): SEQUEST calculates an initial preliminary score (Sp) using a rapid dot-product heuristic based on the number of matching peaks and ion intensities. Only the top-ranking candidate peptides (typically the top 500) advance to the computationally intensive cross-correlation phase.
• FDR and q-Value: Modern proteomics standards reject arbitrary XCorr cutoffs in favor of empirical False Discovery Rate (FDR) control. Downstream validation algorithms such as Percolator use semi-supervised machine learning (support vector machines) trained on target and reversed decoy database matches. Percolator integrates XCorr, ΔCn, mass error, and peptide length to assign posterior error probabilities and q-values, allowing researchers to filter identifications at a rigorous 1% FDR (q ≤ 0.01).

Understanding these scoring principles is fundamental to both bioinformatics and cheminformatics workflows.

Troubleshooting Common Search Issues

When running SEQUEST HT searches, laboratory researchers often encounter specific pipeline errors or suboptimal match rates. The table below outlines common diagnostic scenarios and their operational solutions:

• Excessively Long Run Times or Pipeline Freezes:
  • Cause: Specifying too many dynamic modifications (each dynamic modification exponentially expands the theoretical search space) or selecting an uncurated database containing redundant entries.
  • Remedy: Restrict dynamic modifications to 2–3 biologically verified alterations. Use a curated, non-redundant FASTA file (such as Swiss-Prot rather than TrEMBL or NCBI nr), and ensure precursor tolerances are constrained to narrow ppm ranges on high-resolution instruments.
• Low Identification Rate (Low Number of PSMs):
  • Cause: Incorrect mass tolerance units (e.g., specifying 10 Da instead of 10 ppm), incorrect protease assignment, or omitted static modifications.
  • Remedy: Confirm that precursor mass tolerance matches instrument accuracy (ppm for Orbitrap/TOF, Da for quadrupole/ion trap). Check that alkylation modifications match wet-lab chemistry (carbamidomethylation on C). Verify that missed cleavage allowances reflect actual sample digestion efficiency.
• Elevated Decoy Match Rate (FDR Failure):
  • Cause: Fragment ion tolerance set too wide, allowing random decoy matches to score high cross-correlations, or searching without recalibrated precursor masses.
  • Remedy: Narrow fragment mass tolerance to instrument specifications (e.g., 0.02 Da for high-resolution HCD). Enable spectrum recalibration nodes in Proteome Discoverer prior to database searching.

Free Alternatives to SEQUEST

While SEQUEST HT is a commercial industry standard within the Thermo Scientific software ecosystem, multiple free and open-source database search engines are widely utilized across proteomics research. The following table compares primary verified alternatives:

Software / Engine	Developer / Institution	License / Access Model	Search Mechanism & Key Advantage	Primary Source / Repository
Comet	Jimmy K. Eng / Univ. of Washington Proteomics Resource	Open Source (Apache License 2.0)	Direct open-source re-implementation of the SEQUEST cross-correlation algorithm; fast multithreaded processing for mzML/mzXML data.	https://github.com/UWPR/Comet
X!Tandem	Ron Beavis / The Global Proteome Machine (GPM)	Open Source (CPAL)	Two-pass search strategy utilizing k-score and hypergeometric scoring; refines potential PTMs on previously matched peptides.	https://www.thegpm.org/tandem/
MSFragger	Alexey Nesvizhskii Lab / University of Michigan	Free for Academic & Non-Commercial Use	Ultrafast fragment-indexing algorithm; enables open database searching for unconstrained modification and mass-shift discovery.	https://fragpipe.nesvilab.org/
Andromeda	Matthias Mann Lab / Max Planck Institute of Biochemistry	Free Academic Software (integrated in MaxQuant)	Probability-based peptide score calculated from binomial distribution of matching fragment ions; natively integrated with MaxQuant LFQ.	https://www.maxquant.org/

Frequently Asked Questions

What is sequestration in simple terms?

In simple terms, sequestration means the act of isolating, separating, or setting something apart from others. Depending on the discipline, it can refer to financial isolation (budget cuts), legal isolation (protecting a jury), biological isolation (a non-functioning tissue segment), or computational proteomics (the SEQUEST algorithm matching mass spectra).

What is sequestration in the federal government?

In the United States federal government, sequestration refers to automatic, across-the-board spending cuts applied to federal agency budgets. Established under budgetary legislation such as the Balanced Budget and Emergency Deficit Control Act of 1985 and updated under the Budget Control Act of 2011, it is triggered when Congress fails to enact spending legislation within statutory deficit ceilings.

What does “sequestration” mean in medical terms?

In medical science, sequestration refers to the abnormal separation or isolation of a tissue segment from its normal anatomical connection. A classic example is bronchopulmonary sequestration, a congenital anomaly where a mass of non-functioning lung tissue develops without a connection to the bronchial tree and receives its arterial blood supply from the systemic circulation rather than pulmonary arteries.

What does “sequestered” mean in legal terms?

In legal practice, to sequester means to isolate property, assets, or persons by court order. A sequestered jury is kept isolated from the public, news media, and outside communication during a high-profile trial to prevent external bias. Sequestration also refers to holding disputed assets or property in court custody pending the resolution of litigation.

What is the difference between SEQUEST and Comet?

SEQUEST (specifically SEQUEST HT) is the commercial, proprietary implementation of the cross-correlation search engine licensed and integrated within Thermo Scientific Proteome Discoverer. Comet is an open-source, standalone implementation developed by original SEQUEST co-author Jimmy K. Eng under the Apache 2.0 license, allowing researchers to run SEQUEST-style scoring on Linux clusters and open data formats without commercial software licenses.

What is an acceptable XCorr score in SEQUEST?

Historically, researchers applied rule-of-thumb thresholds based on precursor charge: XCorr ≥ 1.5 for singly charged (1+) peptides, ≥ 2.0 for doubly charged (2+) peptides, and ≥ 2.5 for triply charged (3+) peptides. In modern proteomics workflows, isolated XCorr values are no longer used as standalone cutoffs; instead, raw scores are evaluated alongside ΔCn and mass accuracy through machine-learning tools like Percolator to achieve a rigorous false discovery rate below 1% (q ≤ 0.01).

Sources

• Eng, J. K., McCormack, A. L., & Yates, J. R. (1994). An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. Journal of the American Society for Mass Spectrometry, 5(11), 976–989. DOI: 10.1016/1044-0305(94)80016-2. Read 2026-11-07. https://pubs.acs.org/doi/10.1016/1044-0305%2894%2980016-2
• Thermo Fisher Scientific. (2024). Proteome Discoverer User Guide: Search Engines — Sequest HT Node Parameters. Thermo Fisher Scientific technical documentation. Read 2026-11-07. https://www.thermofisher.com
• University of Washington Proteomics Resource (UWPR). (2026). Comet Open Source MS/MS Search Engine (Jimmy K. Eng). Apache License 2.0. Read 2026-11-07. https://github.com/UWPR/Comet
• The Global Proteome Machine Organization (GPM). (2024). X!Tandem Open Source Tandem Mass Spectrometry Peptide Identification Software. Read 2026-11-07. https://www.thegpm.org/tandem/
• Nesvizhskii Lab, University of Michigan. (2024). MSFragger Ultrafast Peptide Identification Engine. Read 2026-11-07. https://fragpipe.nesvilab.org/
• Max Planck Institute of Biochemistry. (2024). Andromeda Peptide Search Engine (MaxQuant). Read 2026-11-07. https://www.maxquant.org/