LipidSig

LipidSig performs comprehensive analysis of lipidomic data to profile lipid species and characteristics (class, chain length, unsaturation, hydroxyl groups, and fatty acid composition) and to support differential expression, correlation, network, and machine learning analyses.

Key Features:

• Profiling: Generate detailed profiles of lipid species within datasets to identify key lipids associated with specific biological processes or conditions.
• Differential Expression: Identify significant changes in lipid expression between experimental conditions or phenotypes.
• Correlation: Assess relationships between lipid species and other variables to uncover potential interactions and co-regulation.
• Network Analysis: Construct networks to visualize complex interactions among lipids and with other biomolecules.
• Machine Learning: Apply machine learning algorithms to predict lipid functions or classify samples based on lipidomic profiles.
• Characteristic-based conversion: Convert between lipid species and user-defined characteristics using a customizable table to enable subgroup-level analyses.
• Fatty acid and multi-characteristic analysis: Perform analyses focused on fatty acid properties and on multiple lipid characteristics simultaneously.

Scientific Applications:

• Biomarker discovery: Identify lipids or lipid signatures associated with biological conditions or disease states.
• Comparative lipidomics: Compare lipid expression across experimental conditions or phenotypes to investigate lipid roles in cellular functions and disease mechanisms.
• Interaction and co-regulation analysis: Explore correlations and network relationships to uncover lipid interactions and co-regulatory patterns.
• Pathway and systems analysis: Visualize lipid networks to assess broader impacts of lipids on cellular pathways and molecular interactions.
• Predictive modeling: Use machine learning-based classification or prediction of sample phenotypes from lipidomic profiles.

Methodology:

Computational methods explicitly include lipid profiling, differential expression testing, correlation analysis, network construction, application of machine learning algorithms, conversion between lipid species and user-defined characteristics via a customizable table, and analyses focused on fatty acid properties and multiple lipid characteristics simultaneously.