fragpipe on Biowulf

Quick Links

FragPipe is a Java Graphical User Interface (GUI) and CLI workflow tool for a suite of computational tools enabling comprehensive analysis of mass spectrometry-based proteomics data. It is powered by MSFragger - an ultrafast proteomic search engine suitable for both conventional and “open” (wide precursor mass tolerance) peptide identification.

Documentation

fragpipe Main Site

Important Notes

GUI use of this application requires a graphical connection using NX

Beginning with version 18.0, fragpipe may be run in headless mode. See the Batch Job section for more details.
Module Name: fragpipe (see the modules page for more information)
Environment variables set
- FRAGPIPE_HOME
- FRAGPIPE_TAR
fragpipe saves configuration and session information in the same directory where it is installed. If you want your data to persist across sessions, you need to copy this program to your own space. See below.

Interactive job

The fragpipe GUI must be run in an Interactive job. The underlying programs that it uses like msfragger and philosopher can be run as batch or swarm jobs.

First, start a nomachine desktop session on the Biowulf login node. Within your nx session, allocate an interactive session and run fragpipe.
Sample session (user input in bold):

[user@biowulf]$ sinteractive
salloc.exe: Pending job allocation 46116226
salloc.exe: job 46116226 queued and waiting for resources
salloc.exe: job 46116226 has been allocated resources
salloc.exe: Granted job allocation 46116226
salloc.exe: Waiting for resource configuration
salloc.exe: Nodes cn3144 are ready for job

[user@cn3144 ~]$ module load fragpipe

[user@cn3144 ~]$ fragpipe

You should now see the Java Graphical User Interface (GUI).

If you run fragpipe from this central location, you will be unable to save data (like newly created pipelines) and recover them in a new session. You can copy fragpipe to your own space using the following commands or similar:

[user@cn3144]$ module load fragpipe

[user@cn3144]$ export INSTALL_DIR=/data/${USER}/fragpipe

[user@cn3144]$ mkdir -p $INSTALL_DIR && cd $INSTALL_DIR

[user@cn3144]$ tar xvf $FRAGPIPE_TAR

These steps need only be carried out once. In future sessions fragpipe will already be installed in your space. You can now run it using the full path (where <ver> is the version of fragpipe that you copied). Fragpipe and associated tools require some extra modules that are loaded as prerequisites for the fragpipe module (Java, Dotnet, etc). You need to load these in every session for Fragpipe to run correctly, and may do so most easily by loading the matching fragpipe module for your installation.

[user@cn3144]$ module load fragpipe/<ver>
[user@cn3144]$ /data/${USER}/fragpipe/<ver>/bin/fragpipe

Or you can add the directory to your path to open more easily. It is important to export the PATH variable after loading the module, or you will run the centrally-installed version instead.

[user@cn3144]$ module load fragpipe/<ver>
[user@cn3144]$ export PATH=/data/${USER}/fragpipe/<ver>/bin:$PATH

[user@cn3144]$ fragpipe

In either case, your saved data will now persist across sessions. When running fragpipe from your own space, it's still a good idea to load the fragpipe module first to properly prepare your environment.

Batch job

Most jobs should be run as batch jobs.

FragPipe versions 18.0+ may now be run in headless mode to submit as part of an interactive or batch job.

Workflow and manifest files should be created using the GUI either in an interactive session or on your own computer.

Create a batch input file (e.g. msfragger.sh). For example:

#!/bin/bash
set -e
module load fragpipe/18.0
cd /path/to/data/and/conf

fragpipe --headless \
  --workflow <path to workflow file> \
  --manifest <path to manifest file> \
  --workdir <path to result directory>