Quantification and deconvolution of asymmetric LC-MS peaks using the bi-Gaussian mixture model and statistical model selection - PubMed (original) (raw)

Quantification and deconvolution of asymmetric LC-MS peaks using the bi-Gaussian mixture model and statistical model selection

Tianwei Yu et al. BMC Bioinformatics. 2010.

Abstract

Background: Liquid chromatography-mass spectrometry (LC-MS) is one of the major techniques for the quantification of metabolites in complex biological samples. Peak modeling is one of the key components in LC-MS data pre-processing.

Results: To quantify asymmetric peaks with high noise level, we developed an estimation procedure using the bi-Gaussian function. In addition, to accurately quantify partially overlapping peaks, we developed a deconvolution method using the bi-Gaussian mixture model combined with statistical model selection.

Conclusions: Using extensive simulations and real data, we demonstrated the advantage of the bi-Gaussian mixture model over the Gaussian mixture model and the method of kernel smoothing combined with signal summation in peak quantification and deconvolution. The method is implemented in the R package apLCMS: http://www.sph.emory.edu/apLCMS/.

PubMed Disclaimer

Figures

Figure 1

Figure 1

The characteristics of the bi-Gaussian function. (a) the four parameters that define the bi-Guassian function; (b) The function A(τ)-B(τ) used in our estimation. Different σ1/σ2 ratios are plotted.

Figure 2

Figure 2

Comparison of the rate of successfully selecting the correct number of components between the bi-Gaussian mixture model and the Gaussian mixture model. Each sub-plot corresponds to a different degree of asymmetry, as shown in the titles of the sub-plots (ratios between the right- and left- standard deviations). Each dot represents a simulated situation. The values were obtained by averaging the results from 100 simulations. The color represents the level of overlaps between the simulated peaks. The size of the dot represents the amount of noise added to the data. The fill of the dot represents the percentage of values missing in the ion trace.

Figure 3

Figure 3

Comparison of the accuracy in peak size quantification between the bi-Gaussian mixture model and the Gaussian mixture model. Each sub-plot corresponds to a different degree of asymmetry, as shown in the titles of the sub-plots (ratios between the right- and left- standard deviations). Each dot represents a simulated situation. The values were obtained by averaging the results from 100 simulations. The color represents the level of overlaps between the simulated peaks. The size of the dot represents the amount of noise added to the data. The fill of the dot represents the percentage of values missing in the ion trace.

Figure 4

Figure 4

Comparison of the accuracy in peak size quantification between the bi-Gaussian mixture model and the method of kernel smoother combined with signal summation. Each sub-plot corresponds to a different degree of asymmetry, as shown in the titles of the sub-plots (ratios between the right- and left- standard deviations). Each dot represents a simulated situation. The values were obtained by averaging the results from 100 simulations. The color represents the level of overlaps between the simulated peaks. The size of the dot represents the amount of noise added to the data. The fill of the dot represents the percentage of values missing in the ion trace.

Figure 5

Figure 5

Comparison of the fit of the bi-Gaussian mixture model and the Gaussian mixture model to real asymmetric peaks. (a) The ion trace at m/z = 446.8913. (b) The ion trace at m/z = 301.1409. Colored curves: fitted components; black curve: summation of the signal from all the components. Upper-panel: the bi-Gaussian mixture fit; lower-panel: the Gaussian mixture fit.

Similar articles

Cited by

References

    1. Issaq HJ, Van QN, Waybright TJ, Muschik GM, Veenstra TD. Analytical and statistical approaches to metabolomics research. J Sep Sci. 2009;32(13):2183–2199. doi: 10.1002/jssc.200900152. - DOI - PubMed
    1. Dettmer K, Aronov PA, Hammock BD. Mass spectrometry-based metabolomics. Mass Spectrom Rev. 2007;26(1):51–78. doi: 10.1002/mas.20108. - DOI - PMC - PubMed
    1. Dunn WB. Current trends and future requirements for the mass spectrometric investigation of microbial, mammalian and plant metabolomes. Phys Biol. 2008;5(1):11001. doi: 10.1088/1478-3975/5/1/011001. - DOI - PubMed
    1. Griffin JL, Kauppinen RA. A metabolomics perspective of human brain tumours. Febs J. 2007;274(5):1132–1139. doi: 10.1111/j.1742-4658.2007.05676.x. - DOI - PubMed
    1. Chen G, Pramanik BN. Application of LC/MS to proteomics studies: current status and future prospects. Drug Discov Today. 2009;14(9-10):465–471. doi: 10.1016/j.drudis.2009.02.007. - DOI - PubMed

Publication types

MeSH terms

LinkOut - more resources