PALS-V

PALS-V scores pathway activity levels in metabolomics datasets by decomposing and ranking metabolite-set contributions to detect differential pathway activity using the PLAGE (Pathway Level Analysis of Gene Expression) approach.

Key Features:

• Pathway activity decomposition: Implements PLAGE to decompose pathway activity levels into constituent metabolite-set contributions.
• Robustness to missing data and noise: Demonstrates improved handling of missing peaks and noisy signals compared to Over-Representation Analysis (ORA) and Gene Set Enrichment Analysis (GSEA).
• Flexible metabolite grouping: Supports pathway-based groupings and structure-based groupings using MS-MS fragmentation spectral similarities and shared chemical substructures.
• Ranking-based significance detection: Uses a ranking system to identify significantly changing metabolite sets across experimental conditions.
• Applicability to untargeted metabolomics: Designed to accommodate the noise and missing-data characteristics typical of untargeted metabolomics datasets.

Scientific Applications:

• Metabolomics research: Enables detection and interpretation of differential pathway activity relevant to disease mechanisms such as Human African Trypanosomiasis and to ecological and evolutionary studies within plant clades.
• Comparative metabolomics: Facilitates comparison of metabolite sets between groups (e.g., Rhamnaceae clades) using tandem MS/MS fragmentation structures and has been applied to human microbiome datasets such as the American Gut Project.

Methodology:

Implements PLAGE to decompose pathway-level activity and applies a ranking system on pathway-based and MS-MS/chemical substructure-based metabolite groupings, with emphasis on robustness to missing peaks and noisy signals.