BATL

BATL annotates targeted lipidomics LC-ESI-MS/MS data acquired in SRM or MRM modes to assign lipid identities using Bayesian statistical inference for reproducible, high-throughput annotation.

Key Features:

• Supported data types: Processes liquid chromatography-electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS) data acquired in selected reaction monitoring (SRM) or multiple reaction monitoring (MRM) modes.
• Gaussian Naïve Bayes Classifier: Employs a Gaussian naïve Bayes classifier tailored for targeted lipidomics to annotate peak identities.
• Eight input features: Evaluates eight critical features related to retention time, intensity, and peak shape for each peak.
• Feature distribution modeling: Models the distributions of the eight input features across biological conditions to capture variability.
• Joint posterior optimization: Maximizes the joint posterior probabilities of all peak identities at a given transition to generate annotations.
• Performance on lipid classes: Demonstrated rapid and accurate identification with over 95% of peaks correctly annotated in sphingolipid and glycerophosphocholine SRM datasets.
• High-throughput reproducibility: Provides probabilistic, scalable annotation suitable for large targeted lipidomics datasets.

Scientific Applications:

• Targeted lipid identification and quantification: Assigns identities to peaks in targeted lipidomics workflows to support downstream quantification.
• Sphingolipid and glycerophosphocholine analysis: Applied to SRM datasets of sphingolipids and glycerophosphocholines for high-accuracy annotation.
• Biological and disease mechanism studies: Enables large-scale lipid profiling to inform studies of biological processes and disease mechanisms.

Methodology:

Applies a Gaussian naïve Bayes classifier that evaluates eight input features related to retention time, intensity, and peak shape, models their distributions across biological conditions, and maximizes joint posterior probabilities to assign peak identities at each transition.