LiPydomics

LiPydomics analyzes hydrophilic interaction liquid chromatography–ion mobility–mass spectrometry (HILIC-IM-MS) lipidomics data to identify and profile lipid species and to predict collision cross sections (CCS) and HILIC retention times for downstream biomarker and mechanism studies.

Key Features:

• HILIC-IM-MS data processing: Processes lipidomics data acquired by hydrophilic interaction liquid chromatography–ion mobility–mass spectrometry (HILIC-IM-MS).
• Statistical and multivariate analysis (stats module): Performs statistical analyses and multivariate statistics via a "stats" module.
• Plotting (plotting module): Generates visualizations and plots via a "plotting" module.
• Lipid identification (identification module): Identifies lipid species at multiple confidence levels via an "identification" module.
• Experimental database: Includes mass-to-charge ratios (m/z) and collision cross sections (CCS) for 45 lipid classes, with retention times available for 23 of those classes.
• Predictive models and predicted database: Provides predictive models trained on the experimental dataset to predict CCS and HILIC retention times and to generate a predicted lipid database of over 145,388 entries.

Scientific Applications:

• Biomarker discovery and disease mechanism studies: Supports identification of potential biomarkers and elucidation of disease-related lipidomic alterations through comprehensive lipid profiling.
• Complex sample analysis: Applicable to analysis of complex biological samples, including antimicrobial-resistant Staphylococcus aureus strains.
• Expanded lipid class coverage: Extends lipid class coverage and aids identification by integrating experimentally characterized lipid classes with a large predicted lipid database.

Methodology:

Processes HILIC-IM-MS lipidomics data and applies a "stats" module for statistical and multivariate analyses, a "plotting" module for visualization, and an "identification" module for multi-confidence lipid identification; uses an experimental database of m/z and CCS for 45 lipid classes (retention times for 23) and predictive models trained on that dataset to predict CCS and HILIC retention times and generate a predicted lipid database (>145,388 entries).