PALS

PALS computes pathway-level activity scores to rank significantly changing metabolite sets and interpret metabolic pathway alterations in metabolomics studies.

Key Features:

• mPLAGE (adapted PLAGE): Implements an adaptation of the pathway level analysis of gene expression (PLAGE) method, referred to as mPLAGE, specifically tailored for metabolomics data.
• Metabolite-set ranking: Ranks significantly changing metabolite sets across experimental conditions to identify altered pathway activity.
• Decomposition of pathway activity: Decomposes activity levels within metabolic pathways to provide pathway-level scores.
• Grouping by pathways and substructures: Groups related metabolites into sets based on participation in metabolic pathways or shared chemical substructures using tandem mass spectrometry fragmentation patterns.
• Robustness to noisy peak data: Designed to handle untargeted metabolomics peak data with prevalent noise and missing peaks and reported to outperform overrepresentation analysis (ORA) and gene set enrichment analysis (GSEA).
• Normalization impact framework: Provides a framework for investigating the impact of normalization on pathway analysis results.

Scientific Applications:

• Untargeted metabolomics studies: Identifies pathway-level changes and ranks metabolite sets in untargeted metabolomics datasets.
• Pathway-level interpretation: Interprets alterations in metabolic pathways by associating decomposed activity levels with metabolite sets.
• MS/MS substructure analysis: Analyzes coordinated changes among metabolites grouped by tandem mass spectrometry fragmentation patterns to detect chemically related set-level alterations.
• Case study analyses: Has been applied to datasets from human African trypanosomiasis, Rhamnaceae species, and the American Gut Project.
• Normalization assessment: Used to assess how different normalization strategies affect pathway-level results.

Methodology:

Adapts PLAGE into mPLAGE to decompose activity levels within metabolic pathways, groups metabolites by pathway membership or tandem MS fragmentation-based substructures, ranks significantly changing metabolite sets, and provides a framework to evaluate the impact of normalization on pathway analysis results.