IsoGroup: Isotopic Grouping for mass spectrometry labeling experiments

Welcome to IsoGroup documentation!

IsoGroup is a scientific software tool dedicated to the processing of isotopic data from untargeted mass spectrometry (MS) labeling experiments. It is suitable for diverse applications including annotation, isotopic profiling, quantification, and fluxomics experiments.

IsoGroup supports two complementary modes of data processing:

In targeted mode, IsoGroup performs annotation of isotopic clusters based on a user-provided database of compounds. Mass features (m/z, retention time, intensity) are matched to known metabolites and isotopologues according to their exact mass and retention time. A database of isotopologues specific to the selected tracer is generated and used for the annotation.
In untargeted mode, IsoGroup performs clustering by grouping MS features identified as isotopologues into isotopic clusters without prior knowledge, based solely on the features mass differences induced by the tracer element and constrained by the user-defined retention time window.

IsoGroup generates an output file that catalogs all identified isotopic clusters, including the constituent features and their associated m/z values, retention times, intensities, and isotopologue information.

The code is open-source, and available on GitHub under a GPLv3 license.

This documentation is available on Read the Docs (https://isogroup.readthedocs.io) and can be downloaded as a PDF file.

Key features

Targeted mode: annotate isotopic clusters using a user-provided compound database,
Targeted mode: generate a database with isotopologues for a given tracer element and use it for annotation,
Targeted mode: calculate exact mass and retention time errors for annotated clusters,
Untargeted mode: group MS features into isotopic clusters without prior knowledge,
Untargeted mode: provide options to improve isotopologue annotation inside clusters,
Untargeted mode: Untargeted mode: provides options to reduce redundancy at both cluster and feature levels, including filtering of features with identical isotopologue indices,
Compatible with any tracer element,
Handle multiple sample types (unlabeled, fully labeled, Pascal Triangle, etc)
Open-source, free and easy to install everywhere where Python 3 and pip run.