FragPipe: Proteomics Analysis Platform Built on MSFragger

FragPipe is a free-to-download proteomics data analysis suite from the Nesvilab: a graphical and command-line front end that chains a search engine, statistical validation and quantification tools into one workflow. Its search engine is MSFragger, so “FragPipe” and “MSFragger” name the platform and its core engine respectively. This guide is for researchers who have raw mass-spectrometry files and a protein FASTA database and need to get from installation to a readable protein table.

What FragPipe is used for in research

FragPipe is described by its developers as “a cross-platform proteomics data analysis suite” that runs on Windows and Linux, in the cloud, and through either a GUI or the command line. It is powered by MSFragger, which supports conventional and open (mass-offset) database searches. The suite bundles the tools below, so one run can go from spectra to quantified proteins:

• Search and rescoring: MSFragger, Percolator, MSBooster
• Validation and protein inference: Philosopher (PeptideProphet, iProphet, ProteinProphet)
• Quantification: IonQuant, TMT-Integrator
• Modification analysis: PTM-Shepherd, Crystal-C
• Data-independent acquisition (DIA): DIA-Umpire, diaTracer, EasyPQP, DIA-NN, Skyline

Concepts behind these steps are covered in protein identification by database search, DDA and DIA proteomics and what proteomics is. For an older search engine in the same family of tools, see SEQUEST.

FragPipe versus MSFragger

MSFragger is the database search engine, distributed as a Java program that needs Java 17 or later (the MSFragger wiki recommends an Adoptium JRE 17) and, per the wiki, at least 8-16 GB of memory. FragPipe is the wrapper that configures MSFragger and the downstream tools. The MSFragger wiki has separate pages for an academic licence and a commercial licence; read those pages for terms, which this guide does not summarise.

Install or access (official source only)

Download releases from the FragPipe GitHub repository (github.com/Nesvilab/FragPipe); documentation lives at fragpipe.nesvilab.org.

• Windows: run FragPipe-x.x-Installer.exe, then FragPipe-x.x.exe.
• Linux: run the fragpipe shell script.
• macOS: command-line use only, through the Docker image published on Docker Hub.

Some inputs need extra components:

Situation	Requirement stated in the FragPipe documentation
Any step	mzML is accepted by all steps
Bruker .d or .mgf input	Accepted by fewer steps than mzML
timsTOF .d files on Windows	Visual C++ Redistributable for Visual Studio 2017
DIA-NN on Thermo .raw files	Thermo MS File Reader
Overlapping or staggered DIA windows	mzML with demultiplexing
Database splitting, spectral library generation (optional)	Python 3.8-3.11

Run an analysis step by step

The steps follow the official FragPipe tutorial (fragpipe.nesvilab.org/docs/tutorial_fragpipe.html, read 2026-11-08). Check the tutorial for your installed version, since tab layouts change between releases.

1. Config tab: use Browse to point to the folder containing MSFragger.jar, IonQuant.jar and diaTracer.jar, or use Download/Update. Optionally set the Python location.
2. Workflow tab: choose a workflow from the dropdown and press Load. “Default” is the choice for a simple closed search. Set memory and cores (RAM 0 means automatic).
3. Choose Regular MS, or IM-MS for timsTOF PASEF data.
4. Add the spectra with Add files, Add Folder Recursively, or drag and drop. Set the data type for each file: DDA, DDA+, DIA, DIA-Quant or DIA-Lib.
5. Annotate the Experiment and Bioreplicate columns, then save the table as a manifest so the setup can be reused.
6. Database tab: Browse to a FASTA file or Download from UniProt. For a custom FASTA, use Add decoys.
7. MSFragger tab: the Calibration and Optimization default is “Mass Calibration, Parameter Optimization”. The Validation, PTMs and quantification tabs each have a Load defaults button.
8. Browse to an output folder and press RUN.

For unattended runs, the documentation provides separate headless, batch and Docker tutorials, plus workflow tutorials for label-free quantification (LFQ), TMT, DIA and PTM discovery.

Output and how to read it

Based on the FragPipe outputs tutorial (fragpipe.nesvilab.org/docs/tutorial_fragpipe_outputs.html, read 2026-11-08):

File or folder	What it contains
psm.tsv	One row per peptide-spectrum match (PSM), filtered by FDR
ion.tsv	PSMs collapsed per peptide, charge and modification state
peptide.tsv	Peptide-level results
protein.tsv	Protein-group results
combined_ion.tsv, combined_modified_peptide.tsv, combined_peptide.tsv, combined_protein.tsv	The same levels combined across experiment groups
ptm-shepherd-output	global.profile.tsv, global.diagmine.tsv, global.modsummary.tsv
tmt-report	TMT-Integrator output
ion/peptide/protein_label_quant.tsv	IonQuant SILAC results
MSstats.csv, library.tsv, protein.fas	Export for MSstats, spectral library, protein database used
filter.log	Counts remaining after filtering
log_[timestamp].txt	Run log

For glycan searches, PSMs that pass the peptide FDR are reported even when they do not pass the glycan FDR, so filter on a glycan q-value (for example below 0.01) when you need glycan-level confidence. Scores and thresholds in the outputs are defined in the tutorial; confirm them there rather than assuming a default.

Troubleshooting

• timsTOF .d files fail on Windows: install the Visual C++ Redistributable for Visual Studio 2017.
• DIA-NN cannot read Thermo .raw: install Thermo MS File Reader.
• A step rejects your file format: convert to mzML, which all steps accept; staggered DIA windows need demultiplexing during conversion.
• MSFragger will not start: check that Java 17 or later is installed and a 64-bit operating system is in use; the wiki notes a 64-bit OS is needed for more than 4 GB.
• Memory errors: the wiki lists at least 8-16 GB for MSFragger; set the Workflow tab memory value explicitly if automatic (0) is not suitable for your machine.

Free alternatives and related tools

Tool	What its own site says
MaxQuant (maxquant.org, read 2026-11-08)	“A quantitative proteomics software package designed for analyzing large mass-spectrometric data sets”; “freely available and can be downloaded from this site”; the download includes the Andromeda search engine; installation guides for Windows and Linux
OpenMS (openms.org, read 2026-11-08)	Open-source C++ framework with Python bindings and 200+ command-line (TOPP) tools for LC/MS data; three-clause BSD licence; Windows, macOS and Linux

For a commercial DIA tool, see Spectronaut; for a downstream application, see phosphoproteomics.

FAQ

What is the difference between FragPipe and MSFragger?

MSFragger is the database search engine; FragPipe is the suite and interface that runs MSFragger together with validation, quantification and modification-analysis tools.

Which operating systems does FragPipe support?

Windows and Linux with a GUI or command line; on macOS it is command-line only through Docker.

What file formats can I use?

mzML works with all steps. Bruker .d and .mgf are accepted by fewer steps.

Sources

• FragPipe documentation and tutorials, fragpipe.nesvilab.org (read 2026-11-08)
• FragPipe repository, github.com/Nesvilab/FragPipe (read 2026-11-08)
• MSFragger wiki, github.com/Nesvilab/MSFragger/wiki (read 2026-11-08)
• MaxQuant, maxquant.org; OpenMS, openms.org (read 2026-11-08)