Tooluniverse Metabolomics Analysis

What it does

This skill enables AI agents to perform comprehensive analyses of metabolomic data, encompassing metabolite identification and quantification, statistical analysis, pathway interpretation, and integration with other omics datasets. It utilizes various tools for tasks such as normalization (TIC or internal standard), handling missing values, and accounting for batch effects. The tool emphasizes retrieving information from established databases like HMDB, KEGG, CTD, and spectral libraries rather than relying on assumptions.

When to use it

• You have metabolomics data acquired through LC-MS, GC-MS, or NMR techniques.
• You need to identify differential metabolites between different conditions.
• You want to understand which metabolic pathways are dysregulated.
• You're looking for metabolite biomarkers related to disease classification.
• You require integration of metabolomics data with other omics layers like transcriptomics.

Key capabilities

• Data Import: Supports LC-MS, GC-MS, NMR, and both targeted/untargeted platforms.
• Metabolite Identification: Matches metabolites to HMDB, KEGG, PubChem, and spectral libraries.
• Quality Control: Includes peak quality assessment, blank subtraction, and internal standard normalization.
• Normalization: Offers probabilistic quotient, total ion current (TIC), and internal standards normalization methods.
• Statistical Analysis: Performs univariate and multivariate analyses like PCA, PLS-DA, and OPLS-DA.
• Differential Analysis: Identifies significant changes in metabolite abundance.
• Pathway Enrichment: Analyzes metabolic pathways using KEGG, Reactome, and BioCyc.
• Metabolite-Enzyme Integration: Correlates metabolite levels with enzyme expression data.
• Flux Analysis: Performs metabolic flux balance analysis (FBA).
• Biomarker Discovery: Identifies multi-metabolite signatures for biomarker discovery.

Example prompts

• "Analyze this LC-MS metabolomics data for differential metabolites."
• "Which metabolic pathways are dysregulated between conditions?"
• "Identify metabolite biomarkers for disease classification."

Tips & gotchas

• Always confirm metabolite identities using Metabolite_search and Metabolite_get_info; do not assume identity based solely on m/z values.
• Apply defined QC criteria (CV < 30%, blank ratio > 3x) – avoid overriding these with estimations.
• Normalization methods (TIC vs. internal standards) are crucial for accurate quantification and depend on the experimental design.