INTEGRATOR
Robust, reproducible peak detection and integration for large-scale targeted LC-MS-MRM analyses
INTEGRATOR is the stand-alone peak-integration module of the MRMhub framework. MS raw data are imported as mzML files, and MRMhub is then run as a console application in four steps: (1) import and conversion of mzML files, (2) cross-sample peak detection, (3) peak integration, and (4) plotting of chromatograms with integrated peaks as PDFs. Peak-area data are saved as long- and wide-format CSV files.
The workflow
INTEGRATOR performs the peak integration; QUANT (dashed) is a separate MRMhub module.
Download and setup
INTEGRATOR is distributed as a self-contained executable for macOS and Windows via the GitHub releases page; no installation and no separate runtimes are required. A copy of the executable is kept with each analysis project, which ensures reproducibility. On first launch, typically a one-time OS security prompt must be cleared. Detailed instructions are given in Installing MRMhub-INTEGRATOR.
Getting started
- Run the demo — the bundled dataset with all required files can be processed end-to-end with INTEGRATOR. It ships within the latest release and is documented in
readme.txt. - Quick Start — a short tutorial for running a new peak integration project.
- Browse real analyses — Read details on peak integration of two large-scale lipidomics studies.
- Manual — detailed instructions and tool reference for every stage; also available from the Manual menu above.
- FAQ & Troubleshooting — common problems preparing input files and running INTEGRATOR, and how to resolve them.
Post-processing of integration results
The results produced by INTEGRATOR can be used directly by the QUANT R package (mrmhub) for post-processing — normalisation, drift/batch correction, calibration, quality control, and reporting. See the QUANT manual ↗ and the corresponding Demo.
Limitations
INTEGRATOR derives its speed from surveying peak features across all samples at once and sharing that information during integration, rather than treating each sample in isolation; this cross-sample approach does not compromise integration quality.
Its limitations follow from this design and its intended scope. INTEGRATOR processes unit-resolution LC-MRM data from triple-quadrupole instruments only, not high-resolution (HRMS/PRM) acquisitions. It fits a straight-line baseline and applies no peak-shape deconvolution, so specialised or vendor software may integrate individual low-abundance or co-eluting peaks more accurately. The cross-sample retention-time alignment can fail on very noisy or irregular chromatograms. And because features are defined by consensus across samples, a peak present in only a few samples may be missed.
Links
License
- AGPLv3 (non-commercial use)
- Commercial licensing — contact Jonathan Tan