Statistical Assessment of Time and Mass Alignment Quality in Liquid Chromatography-Mass Spectrometry

Abstract

This research evaluated the efficacy of an alignment quality algorithm and follows its development. Proteomics research frequently involves liquid chromatography-mass spectrometry (LC-MS) methods for data collection. To correct for systematic errors, researchers often apply alignment algorithms to these data; the quality of these alignment procedures is often overlooked and needs to be assessed to offer confidence in results derived from LC-MS data. The data we worked with was aligned by a dynamic time warping (DTW) alignment algorithm. We developed an assessment algorithm based on a null hypothesis significance testing method applied to a generalized regression of particular LC-MS data. We found that the assessment algorithm alone was an insufficient indicator of the quality of alignment and could in some cases fail, but there is potential for it to be valuable as an aide with other information to make judgments on the alignment quality.