From mass spec runs to identified proteins.

Raw spectra converted, peaks picked, features found and matched against a protein database, with the tables ready for comparison.

Join the waitlist

proteomics.flow5 steps

What is Proteomics analysis?

Proteomics analysis identifies and quantifies the proteins in a sample from mass spectrometry data. A typical discovery workflow converts raw spectra to the open mzML format, picks peaks, groups them into peptide features, and matches those against a protein database at a controlled false discovery rate, using tools from the OpenMS platform.

In discovery (shotgun) proteomics, proteins are digested into peptides, separated by liquid chromatography and measured by a mass spectrometer. The instrument records spectra: the masses of the peptides and of their fragments. The analysis works backwards from those spectra to the peptides, and from the peptides to the proteins they came from.

Raw files are vendor-specific, so the first step converts them to mzML, the community standard that open tools read. Peak picking reduces each spectrum to its peaks; feature finding groups the peaks that belong to the same peptide across retention time and charge states; identification matches fragment spectra against theoretical spectra generated from a protein database such as UniProt.

An identification is only useful with an error rate you can state. The standard approach searches real (target) and reversed or shuffled (decoy) sequences together: hits to decoys estimate how many target hits are false, which sets the false discovery rate at which identifications are accepted (Elias and Gygi, 2007).

What happens to your data

For labs running discovery proteomics alongside their other experiments.

Questions it answers

Which proteins are in this sample?
How does abundance change between conditions?
Did the pull-down enrich what it should?

1
Peak picking
Finds the peaks in each spectrum.
Look at: The number of spectra per run. A run with far fewer is worth a look.
2
Feature finding
Groups peaks into features that correspond to peptides.
Look at: Feature counts across runs, which should be broadly similar.
3
Protein identification
Matches features against a protein database such as UniProt, with a false discovery rate.
Look at: The number of proteins identified at your chosen FDR.

What you take to lab meeting

A table of identified proteins per run
Abundance estimates ready for comparison
Run-level QC summaries

Common questions

What is mzML?

An open, vendor-neutral file format for mass spectrometry data, developed by the Proteomics Standards Initiative (Martens et al., 2011). Converting raw instrument files to mzML lets open-source tools read data from any vendor's instrument.

What false discovery rate should I use?

1% at the peptide and protein level is a common choice in discovery proteomics. Whichever you choose, apply it consistently and report it with the results.

Which protein database should I search?

The reference proteome for your organism, for example from UniProt, plus common contaminants such as keratins and the digestion enzyme. A database much larger than the sample needs inflates the search space and makes the false discovery rate harder to control.

Can I compare protein abundance between conditions?

Yes. Feature intensities give label-free abundance estimates for each run, which can be normalised and compared across conditions. Labelled methods such as TMT need their own quantification steps.

What do I need to know about the instrument method?

The essentials: the enzyme used for digestion, the precursor and fragment mass tolerances, and any fixed or variable modifications. They are on the method sheet, and they decide which peptides can be matched.

Sources

The papers behind the tools in this pipeline, and the advice on this page. Checked 25 October 2026.

Röst HL, et al. (2016). OpenMS: a flexible open-source software platform for mass spectrometry data analysis. Nature Methods 13, 741–748.
Martens L, et al. (2011). mzML: a community standard for mass spectrometry data. Molecular & Cellular Proteomics 10, R110.000133.
Elias JE, Gygi SP (2007). Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nature Methods 4, 207–214.

Read before you run it

Bioinformaticians say “pipeline” the way bench scientists say “protocol”. Here is the translation, and why the analogy holds better than you might expect.

What a pipeline actually is, for people who have never written one

Explainers, 7 October 2026

Other starter pipelines

Run Proteomics on your own data.

OpenBionode is in early access. Tell us what you analyse and we’ll make sure the steps you need are ready when your seat opens.