Multi-profile Bayesian alignment model for LC-MS data analysis with integration of internal standards.

Area of Science:

• Biomolecular profiling studies
• Omics data analysis
• Analytical chemistry

Background:

• Liquid chromatography-mass spectrometry (LC-MS) is crucial for omics studies (proteomics, metabolomics, glycomics).
• Accurate LC-MS data preprocessing, particularly retention time (RT) alignment, is essential for reliable biological group comparisons.
• Existing RT alignment methods often fail to utilize complementary information within the entire LC-MS dataset.

Purpose of the Study:

• To develop an advanced Bayesian alignment model for LC-MS data.
• To enhance the accuracy and reliability of RT alignment in omics studies.
• To provide robust uncertainty measures for RT variability estimates.

Main Methods:

• A novel Bayesian alignment model was developed for LC-MS data.
• The model integrates multiple information sources, including internal standards and clustered chromatograms.
• The approach offers a mathematically rigorous framework for data integration.

Main Results:

• The Bayesian alignment model effectively estimates RT variability with associated uncertainty.
• Application to metabolomic, proteomic, and glycomic data demonstrated significant improvements in RT alignment.
• Performance evaluation using ground-truth data confirmed enhanced correlation of variation, reduced RT differences, and improved peak matching.

Conclusions:

• The proposed Bayesian alignment model offers a superior approach to RT alignment in LC-MS data analysis.
• By integrating diverse data sources, the model enhances the precision and accuracy of omics profiling.
• This method provides a valuable tool for researchers in proteomics, metabolomics, and glycomics.