Getting started

Setup

Required software dependencies

• NextFlow 25.10.4, together with Java >= 17, <= 25
  (older NextFlow down to 22.10 should also work)

• Apptainer, Podman, Docker, Singularity or Guix

Ubuntu 24.04 LTS

sudo apt install openjdk-25-jre-headless

# change the default Java if currently selected one is older than version 17
sudo update-alternatives --config java

curl -s https://get.nextflow.io | bash
sudo mv nextflow /usr/local/bin/

sudo apt update -y
sudo apt install software-properties-common -y
sudo add-apt-repository ppa:apptainer/ppa -y
sudo apt update
sudo apt install apptainer -y

CSC

• Puhti and Mahti offer Apptainer and LUMI offers Singularity out of the box.

• nextflow executable can be downloaded with curl -s https://get.nextflow.io | bash and placed somewhere in the PATH. Java is available as a module eg. module add Java/17.0.7.

Get the elolab/glaDIAtor-nf

git clone https://github.com/elolab/glaDIAtor-nf.git

Tutorial

The workflow can be run at any location. The main entrypoint for the NextFlow is gladiator-nf/workflow/gladiator.nf.

Let’s run this tutorial in a tutorial directory next to gladiator-nf

mkdir tutorial
cd tutorial

Download example input files

mkdir raw-dia-spectra
wget --no-directories --directory-prefix=raw-dia-spectra 'ftp://massive-ftp.ucsd.edu/v05/MSV000090837/raw/Exp03_repeat_90minGradient_from_Exp02_same_Mastermix/210820_Grad090_LFQ_'{A,B}'_01.raw'

mkdir protein-sequences
wget --directory-prefix=protein-sequences 'ftp://massive-ftp.ucsd.edu:/v05/MSV000090837/sequence/fasta/*.fasta'

Only one pair of technical replicates (10820_Grad090_LFQ_A_01.raw and 210820_Grad090_LFQ_B_01.raw) is selected from MSV000090837 dataset to lower the storage and processing requirements. .mzML format is also available but .raw is used to demontstrate the conversion.

Conversion to mzML and peak picking

Docker

MSCONVERT_CMD='find . -iname '"'"'raw-dia-spectra/*.raw'"'"' -print0 | xargs -P5 -0 -i wine msconvert {} --filter '"'"'titleMaker <RunId>.<ScanNumber>.<ScanNumber>.<ChargeState> File:"<SourcePath>", NativeID:"<Id>"'"'"' -o dia-spectra/'
echo "${MSCONVERT_CMD}" > ./msconvert.sh
chmod +x msconvert.sh

PWIZ_IMAGE_ID="$(docker pull chambm/pwiz-skyline-i-agree-to-the-vendor-licenses:3.0.21354-9ee14c7)"
docker run -it  -v $PWD:$PWD -w $PWD $PWIZ_IMAGE_ID /bin/bash ./msconvert.sh

Podman

MSCONVERT_CMD='find . -iname '"'"'raw-dia-spectra/*.raw'"'"' -print0 | xargs -P5 -0 -i wine msconvert {} --filter '"'"'titleMaker <RunId>.<ScanNumber>.<ScanNumber>.<ChargeState> File:"<SourcePath>", NativeID:"<Id>"'"'"' -o dia-spectra/'
echo "${MSCONVERT_CMD}" > ./msconvert.sh
chmod +x msconvert.sh

PWIZ_IMAGE_ID="$(podman pull docker://chambm/pwiz-skyline-i-agree-to-the-vendor-licenses:3.0.21354-9ee14c7)"
podman run -it  -v $PWD:$PWD -w $PWD $PWIZ_IMAGE_ID /bin/bash ./msconvert.sh

Apptainer

MSCONVERT_CMD='find . -iname '"'"'raw-dia-spectra/*.raw'"'"' -print0 | xargs -P5 -0 -i wine msconvert {} --filter '"'"'titleMaker <RunId>.<ScanNumber>.<ScanNumber>.<ChargeState> File:"<SourcePath>", NativeID:"<Id>"'"'"' -o dia-spectra/'
echo "${MSCONVERT_CMD}" > ./msconvert.sh
chmod +x msconvert.sh

apptainer pull pwiz.sif docker://chambm/pwiz-skyline-i-agree-to-the-vendor-licenses:3.0.21354-9ee14c7
apptainer exec --bind $PWD pwiz.sif bash msconvert.sh

Native ProteoWizard builds

Many proprietary mass spectrometry formats can only be read using Windows-only converter blobs. These don’t run on Linux, and the Linux build of ProteoWizard doesn’t ship with them. Because of that, people are often forced to run the Windows build using Wine, inside a container to get full format support.

Otherwise convertion command would look like this

Ubuntu 24.04

sudo apt install libpwiz-tools

find . -iname 'raw-dia-spectra/*.raw' -print0 | xargs -P5 -0 -i msconvert {} --filter 'titleMaker <RunId>.<ScanNumber>.<ScanNumber>.<ChargeState> File:"<SourcePath>", NativeID:"<Id>"' -o dia-spectra/

GlaDIAtor-nf requires input data to be peak picked before further processing. Material and Methods document attached to the example files mention that centroid pick picking was already performed. Otherwise prepend additional --filter parameter to the msconvert command, typically --filter "peakPicking true 1-".

Configuration

Paste following to a new file, it can be called gladiator-config.nf

params {
    fragment_mass_tolerance = 0.02  // fragment mass tolerance in Dalton, 0.02 is a sensible default
    precursor_mass_tolerance = 10  // in ppm (parts per million)
    protFDR = 0.01  // passed to mayu as cutoffrate for finding the peptide probability
    max_missed_cleavages = 1 
    pyprophet_subsample_ratio = null  // when set to null, it resolves to subsample ratio of 1 / {number of samples}

    // DIA-Umpire, Comet and X! Tandem configuration can be customized by making copies
    // of 'config/diaumpire.params', 'config/comet.params' or 'config/xtandem.xml'.
    //
    // The custom files are not used until their location is specified, for example:
    //
    //   diaumpireconfig = "./diaumpire.params"
    //   comet_template = "./comet.params"
    //   xtandem_template = "./xtandem.xml"
}

Run glaDIAtor-nf

Docker

gladiator_location="../gladiator-nf"

NXF_VER="25.10.4" nextflow -c gladiator-config.nf -c "${gladiator_location}/config/containers/docker.config" \
    run "${gladiator_location}/workflow/gladiator.nf" --diafiles='dia-spectra/*.mzML' --fastafiles='protein-sequences/*.fasta'

Podman

gladiator_location="../gladiator-nf"

NXF_VER="25.10.4" nextflow -c gladiator-config.nf -c "${gladiator_location}/config/containers/podman.config" \
    run "${gladiator_location}/workflow/gladiator.nf" --diafiles='dia-spectra/*.mzML' --fastafiles='protein-sequences/*.fasta'

Apptainer and Singularity

gladiator_location="../gladiator-nf"

NXF_VER="25.10.4" nextflow -c gladiator-config.nf -c "${gladiator_location}/config/containers/singularity.config" \
    run "${gladiator_location}/workflow/gladiator.nf" --diafiles='dia-spectra/*.mzML' --fastafiles='protein-sequences/*.fasta'

Results

Two files should be available in the ./results directory

• dia/DIA-peptide-matrix.tsv contains the intensitities on the peptide level

  | ProteinName_FullPeptideName                       | 210820_Grad090_LFQ_A_01.mzML   | 210820_Grad090_LFQ_B_01.mzML   |
  |---------------------------------------------------|--------------------------------|--------------------------------|
  | 1/A2I7N3_LAVSHVIHK                                | 624895                         | 723064                         |
  | 3/sp|P08729|K2C7_HUMAN/Q3KNV1/P08729_VDALNDEINFLR | 2037142                        | 1283280                        |
    ...
  

• dia/DIA-protein-matrix.tsv contains intensities are on the protein group level

  | ProteinName                                   | 210820_Grad090_LFQ_A_01.mzML | 210820_Grad090_LFQ_B_01.mzML |
  |-----------------------------------------------|------------------------------|------------------------------|
  | 1/sp|O14561|ACPM_HUMAN                        | 1410100                      | 1359820                      |
  | 2/sp|P0CX49|RL18A_YEAST/sp|P0CX50|RL18B_YEAST | 99249410                     | 44970930                     |
    ...
  

A different output directory can be selected with --outdir parameter.