Lilikoi

Lilikoi performs personalized pathway-based classification modeling of metabolomics data to convert metabolite measurements into pathway-level profiles for disease-related analysis.

Key Features:

• Feature Mapping Module: Standardizes user-provided metabolite names and maps metabolites to relevant biological pathways.
• Dimension Transformation Module: Computes pathway deregulation scores and transforms raw metabolomic profiles into personalized pathway-based profiles.
• Feature Selection Module: Identifies significant pathway features that are associated with disease phenotypes.
• Classification and Prediction Module: Applies a variety of machine learning algorithms for classification and predictive modeling using personalized pathway profiles.

Scientific Applications:

• Personalized disease classification: Builds classification models to distinguish phenotypes or disease states based on pathway-level profiles.
• Pathway-level biomarker discovery: Identifies pathway features linked to disease mechanisms and potential therapeutic targets.
• Integration with systems biology: Integrates metabolomics data with biological pathways for pathway-centric analyses.
• Precision medicine research: Supports analyses aimed at understanding metabolic perturbations at the individual level for precision medicine.

Methodology:

Standardization and mapping of metabolite names to pathways; calculation of pathway deregulation scores; transformation of raw metabolomic data into personalized pathway-based profiles; feature selection of significant pathways; application of machine learning algorithms for classification and prediction.