LipidIN: a comprehensive repository for flash platform-independent annotation and reverse lipidomics

Area of Science:

• Biochemistry
• Computational Biology
• Analytical Chemistry

Background:

• Traditional lipid annotation faces challenges in accuracy, coverage, speed, and depth.
• Accurate lipid profiling is crucial for understanding biological processes and disease mechanisms.

Purpose of the Study:

• To introduce LipidIN, an advanced, platform-independent framework for lipid annotation.
• To improve the accuracy, coverage, speed, and depth of lipid profiling.
• To facilitate lipid biomarker discovery in clinical cohorts.

Main Methods:

• Development of a 168.5-million lipid fragmentation hierarchical library.
• Implementation of an expeditious querying module with speeds exceeding 100 billion queries/second.
• Creation of a lipid categories intelligence model using relative retention time rules.
• Integration of a Wide-spectrum Modeling Yield network for fragment fingerprint regeneration.

Main Results:

• LipidIN achieves high-speed querying across extensive mass spectral libraries.
• The intelligence model reduces false positives and predicts unannotated lipids with a 5.7% false discovery rate, covering 8923 lipids.
• The Wide-spectrum Modeling Yield network boosts recall by an estimated 20%, enhancing accuracy and coverage.
• Demonstrated utility in lipid annotation and biomarker discovery tasks.

Conclusions:

• LipidIN offers a significant advancement in lipid annotation technology.
• The framework provides a robust solution for comprehensive lipid profiling.
• LipidIN is valuable for lipid biomarker discovery in clinical applications.