SpectralWorks Technical Note
SpectralWorks Ltd, United Kingdom
First Published: Scottish Metabolomics Network Newsletter August 2018.
GC-MS and LC-MS play an important part in the metabolomics workflow. Deriving features from these analytical techniques is a significant part of this workflow.
Peak overlapping or partial co-elution is a common problem in any chromatographic separation technique. Where a detector which produces spectral characteristics is used, such as a mass spectrometer, the deconvolution of partially overlapping peaks can be achieved without any assumptions being made regarding the peak shape or underlying spectra of the individual components. Figure 1 shows the purpose of chromatographic deconvolution as the resolution between two components is reduced. From fully resolved to unresolved, AnalyzerPro is able to determine the two components in each case.
Figure 1. Deconvolution of resolved, partially resolved and unresolved components.
The benefits of deconvolution can be realised in two ways. The number of features determined in samples may be increased and/or the required chromatographic resolution may be reduced. The effect of the latter is to help reduce analysis time.
Deconvolution brings with it the question of how many data points need to be acquired across a peak? If we reduce the issues of peak fronting and tailing and take the peak width at Full Width Half Maximum (FWHM), typically 8 to 10 sample points would be expected in order to be able to reproducibly generate peak data. Even in the simplest case of partial co-elution, a sampling rate of 20 – 25 data points would not be unreasonable. The closer that the co-eluting components are together, then the more data points are required to generate reliable deconvoluted peak data.
Chromatographic peak deconvolution is very effective at determining multiple components in your samples.