JMod

JMod is an open and flexible software for increasing the throughput of sensitive proteomics by supporting multiplexing in the mass and time domains

Abstract

The throughput of mass spectrometry (MS) proteomics can be increased substantially by multiplexing that enables parallelization of data acquisition. Such parallelization in the mass domain (plexDIA) and the time domain (timePlex) increases the density of mass spectra and the overlap between ions originating from different precursors, potentially complicating their analysis. To enhance sequence identification and quantification from such spectra, we developed an open source software for Joint Modeling of mass spectra: JMod. It uses the intrinsic structure in the spectra and explicitly models overlapping peaks as linear superpositions of their components. This modeling enabled performing 9-plexDIA using 2 Da offset PSMtags by deconvolving the resulting overlapping isotopic envelopes in both MS1 and MS2 space. The results demonstrate 9-fold higher throughput with preserved quantitative accuracy and coverage depth. This support for smaller mass offsets increases multiplexing capacity and thus proteomic throughput for a given plexDIA tag, and we demonstrate this generalizability with diethyl labeling. By supporting enhanced decoding of DIA spectra multiplexed in the mass and time domains, JMod provides an open and flexible software that enables increasing the throughput of sensitive proteomics.

---

Reference

JMod: Joint modeling of mass spectra for empowering multiplexed DIA proteomics Kevin McDonnell, Nathan Wamsley, Jason Derks, Sarah Sipe, Maddy Yeh, Harrison Specht, Nikolai Slavov bioRxiv 2025.05.22.655512; doi: 10.1101/2025.05.22.655512

---

Deployment

Table of Contents

• Environment Setup
  • Linux/MacOS
  • Windows
• Outputs
• Running a Search
  • File Conversion
  • With Command Line Args
  • With JSON Configuration File
• Search Parameters
• Running Tests

Environment Setup

Linux/MacOS

To set up a Conda environment to run JMod in, please download the data/jmod_env.yml file.

In a dedicated terminal, type:

1. conda env create -f jmod_env.yml -n $env_name
2. conda activate $env_name
3. python run_jmod.py <args>

If run successfully, the inputted configurations alongside "Loading library..." should be printed.

Windows

If on a Windows machine, please use the data/jmod_env_windows.yml file to set up your environment. Set up will follow the same steps as Linux/MacOS.

Outputs

• filtered_IDs.csv – list of precursor identifications that are FDR filtered
• all_IDs.csv – list of peptide identifications that are not FDR filtered
• decoylibsearch_coeffs.csv – list of JMod MS2 precursor coefficients in each
• spectrumfirstSearch.csv – Precursor IDs from the preliminary search
• params.txt – list of parameters used in the JMod search
• summary.txt – summary of the precursor and protein IDs from the search

Running a Search

File Conversion

JMod supports .mzML files. When converting files to .mzML, the data should be centroided. This can be done with MSconvert through the command --filter peakPicking true 1-

With Command Line Args

python path/to/run_jmod.py -l path/to/library.tsv -i path/to/file_to_search.mzML

With JSON Configuration File

Example config.json in ./data/default_config.json

python path/to/run_jmod.py --config_json path/to/config.json

Search Parameters

-i, --mzml
  Input file in mzML format
-l, --speclib
  Spectrum library in DIANN output format (must be .tsv)
-r, --use_rt
  Use retention time filtering during search
  default = False
-f, --use_features
  Use peptide-like features in the preliminary search. mzML file must have associated Dinosaur/Biosaur output
  default = True
-m --atleast_m
  Required number of fragments matched from top N fragments (N=10)
  default = 3
-p --ppm
  MS2 matching tolerance in parts per million.
  default = 10
--iso
  Use MS2 isotopes in search.
  default = False
--num_iso
  Number of MS2 isotopes to consider if using them
  default = 2
--tag
  Tag used in the experiment, if any. See mass_tags.py for details.
  default = None
--use_emp_rt
  Force use of library retention time for alignment.
  default = False
--user_rt_tol
  Force use of provided retention time tolerance.
  default = False
--rt_tol
  User provided retention time tolerance.
--no_ms1_req
  Don't require observation of an MS1 peak for consideration in the search.
  default = False
--ms1_ppm
  User provided MS1 ppm error tolerance.

Running Tests

python run_tests.py -c