MetGem

MetGem constructs molecular networks from tandem mass spectrometry (MS/MS) data and visualizes high-dimensional spectral relationships to support interpretation of metabolomics, proteomics, and GC-EI-MS analyses.

Key Features:

• Molecular Networking with EI-MS Data: Extends molecular networking to electron ionization (EI) mass spectrometry data, enabling network generation from GC-EI-MS datasets that lack clear parent ion selection.
• Spectral Similarity Organization: Computes spectral similarities using a cosine-score metric on peak lists containing fragments and neutral losses to organize related spectra.
• Integration with MZmine2: Supports integration with MZmine2 for combined data processing workflows.
• t-SNE Based Projections: Applies t-SNE (t-distributed Stochastic Neighbor Embedding) to project high-dimensional spectral data into lower-dimensional spaces for visualization.

Scientific Applications:

• Metabolomics: Enables exploration and annotation of metabolite relationships from MS/MS and GC-EI-MS data.
• Proteomics: Supports visualization of peptide- and protein-related spectral networks derived from MS/MS datasets.
• Chemical Ecology: Facilitates comparative analysis of small molecules and chemical profiles relevant to chemical ecology studies.

Methodology:

Calculates cosine-scores from peak lists of fragments and neutral losses, constructs molecular networks from those similarity scores, and applies t-SNE projections for visualization while supporting EI-MS data.