Parametric time warping in metabolomics

In many omics fields, experimental data are collected using hyphenated techniques coupling detection methods like mass spectrometry (MS) and nuclear magnetic resonance (NMR) to chromatographic separation methods. Unfortunately, chromatographic retention times can be quite variable, and as a consequence it may be hard to determine which features are actually related across the different samples. Parametric time warping (PTW), a polynomial transformation of the time axis to maximize the agreement between time profiles in different samples, has been applied successfully as a restricted, yet powerful method to align features in such a situation. This presentation discusses the different aspects of applying PTW in the context of MS-based metabolomics. In particular, recent improvements increasing the speed and quality of the alignment will be investigated. Software is available as the open-source ptw package for R and in the form of matlab scripts