A geometric approach for the alignment of liquid chromatography—mass spectrometry data

Abstract

This algorithm is implemented as part of the OpenMS software library for shotgun proteomics and available under the Lesser GNU Public License (LGPL) at www.openms.de.

Generic Software Frameworks for GC-MS Based Metabolomics

“…After binning, the data is stored as a rectangular array of values, with the first dimension representing time, the second dimension representing the approximate bin mass values, and the third dimension representing the intensity corresponding to each measured ion. This process is also often described as resampling (Lange et al, 2007).…”

“…These methods can also be applied to interpolate values where gaps are present in the original data. The top-hat filter (Bertsch et al, 2008;Lange et al, 2007) is used to remove a varying baseline from the signal. More refined methods use signal decomposition and reconstruction methods, such as Fourier transform and continuous wavelet transform (CWT) (Du et al, 2006;Fredriksson et al, 2009;Tautenhahn et al, 2008) in order to remove noise and baseline contributions from the signal and simultaneously find peaks.…”

“…The other available method is based on the Obi-Warp dynamic time warping (Prince & Marcotte, 2006) algorithm and is capable of correcting large non-linear retention time distortions. It uses the peak set with the highest number of features as alignment reference, which is comparable to the approach used by Lange et al (2007). However, it is much more computationally demanding then the LOESS-based alignment.…”

“…Another distinction in alignment algorithms is the requirement of an explicit reference for alignment. Some methods apply clustering techniques to select one chromatogram that is most similar to all others (Hoffmann & Stoye, 2009;Smilde & Horvatovich, 2008), while other methods choose such a reference based on the number of features contained in a chromatogram (Lange et al, 2007) or by manual user choice (Chae et al, 2008;Clifford et al, 2009). For high-throughput applications, alignments should be fast to calculate and reference selection should be automatic.…”

Quantitative strategies to fuel the merger of discovery and hypothesis-driven shotgun proteomics

“…This alignment is performed by algorithms such as dynamic time warping or correlation optimized warping which find an optimal mapping of retention times between runs that maximizes their similarity [34][35][36]. The second class matches detected features (such as peptide isotope distributions or MS/MS spectra) between runs, and applies algorithms such as regression to fit a time correction function to the matched markers [29,37,38]. Additionally, variations in signal intensity can be addressed by normalization.…”