LORA, Lipid Over-Representation Analysis Based on Structural Information.

Area of Science:

• Lipidomics
• Bioinformatics
• Systems Biology

Background:

• Lipidomic studies are growing, necessitating efficient and automated data analysis tools.
• Standardized lipid nomenclature is crucial for accurate statistical evaluation of lipidomic data.
• Existing methods often struggle with the complexity and variability of lipid names.

Purpose of the Study:

• To introduce the Lipid Over-Representation Analysis (LORA) web application for automated lipidomic data analysis.
• To address the challenge of lipid nomenclature standardization in large-scale lipidomic studies.
• To facilitate the interpretation of epilipidomics data by identifying statistically significant lipid features.

Main Methods:

• The LORA web application utilizes the Goslin framework for lipid name standardization.
• Users upload experimental and reference datasets for analysis.
• Lipid names are normalized using selected grammars, enabling over-representation analysis.
• Results are interactively explored and visualized based on lipidome hierarchy.

Main Results:

• LORA successfully translates lipid names into standardized nomenclature, including detailed structural information.
• The tool identifies statistically over-represented lipid terms and defines Very Important Lipids (VILs).
• Example analysis shows enrichment of "glycerophospholipids" with specific acyl chains at the "sn-2 position".

Conclusions:

• LORA provides an efficient and automated solution for analyzing complex lipidomic data.
• The tool enhances the interpretation of epilipidomics datasets by focusing on structurally defined lipid features.
• LORA facilitates the identification of key lipids and their biological relevance.