A Novel Two-Stage Alignment Method for Liquid Chromatography Mass Spectrometry-Based Metabolomics

Abstract

We report a novel two-stage alignment algorithm that contains full alignment and partial alignment, for the analysis of LC-MS based metabolomics data. The purpose of full alignment is to detect landmark peaks that present in all peak lists to be aligned. These peaks were first selected based on m/z value and isotopic peak profile matching. After removing peaks with large Euclidian distance of retention time from the potential landmark peaks, a mixture score was calculated to measure the matching quality of each landmark peak pair between reference peak list and a test peak list. After optimizing the weight factor in the mixture score, the value of minimum mixture score of all landmark peaks was used as the threshold for peak matching in the partial alignment. A local optimization based retention time correction method was used to correct the retention time changes between peak lists during partial alignment. The two-stage alignment method was used to analyze a spiked-in experimental data and further compared with literature reported algorithm RANSAC implemented in MZmine.

This work was supported by NIH grant 1RC2AA019385 through the National Institute on Alcohol Abuse and Alcoholism. This work was also partially supported by National Institute of Health (NIH) grant 1RO1GM087735 through the National Institute of General Medical Sciences (NIGMS).