MetaProViz

MetaProViz integrates prior knowledge with metabolomics and exometabolomics data to generate mechanistic hypotheses through pre-processing, functional analysis, and visualization.

Key Features:

• Data Pre-processing: Performs feature filtering, missing value imputation, normalisation, and outlier detection on standard metabolomics and exometabolomics data.
• Differential Metabolite Analysis (DMA): Conducts statistical analysis to identify differentially abundant metabolites between conditions.
• Metabolite Clustering Analysis (MCA): Implements clustering based on regulatory rules to group metabolites by functional relationships.
• Metabolism Signature Database (MetSigDB): Provides curated metabolite sets including classical pathway-metabolite sets, metabolite-receptor sets, metabolite-transporter sets, and chemical class-metabolite sets.
• Gene-to-Metabolite Set Conversion: Enables conversion of gene sets to metabolite sets using enzyme-metabolic reaction associations.
• Metabolite Identifier Translation: Translates between metabolite identifiers across commonly used databases, analyzes mapping ambiguities, and completes missing annotations.
• Visualization: Generates biologically interpretable graphs in publication-ready formats.

Scientific Applications:

• Intracellular Metabolomics: Analyzes metabolite profiles from cell extracts and tissue samples to infer metabolic states.
• Exometabolomics: Processes cell-culture media and other extracellular samples to study metabolic exchange and nutrient utilization.
• Functional Enrichment Analysis: Applies classical enrichment methods using MetSigDB to identify overrepresented metabolic pathways and processes.
• Biologically Informed Clustering: Groups metabolites based on prior knowledge of regulatory relationships to reveal functional modules.
• Sample Metadata Analysis: Integrates experimental metadata to explore associations between phenotypic variables and metabolic profiles.
• Cancer Metabolism Research: Identifies altered metabolic pathways in disease contexts, such as linking methionine usage to DNA hypermethylation in clear-cell renal cell carcinoma.

Methodology:

MetaProViz implements a modular framework consisting of five components: processing, differential analysis, prior knowledge integration, functional analysis, and visualization. The package uses enzyme-metabolic reaction associations to convert gene sets to metabolite sets and translates between metabolite identifiers of commonly used databases while analyzing mapping ambiguities and completing missing annotations.