MetSign: a computational platform for high-resolution mass spectrometry-based metabolomics

Abstract

Data analysis in metabolomics is currently a major challenge, particularly when large sample sets are analyzed. Herein, we present a novel computational platform entitled MetSign for high-resolution mass spectrometry-based metabolomics. By converting the instrument raw data into mzXML format as its input data, MetSign provides a suite of bioinformatics tools to perform raw data deconvolution, metabolite putative assignment, peak list alignment, normalization, statistical significance tests, unsupervised pattern recognition, and time course analysis. MetSign uses a modular design and an interactive visual data mining approach to enable efficient extraction of useful patterns from data sets. Analysis steps, designed as containers, are presented with a wizard for the user to follow analyses. Each analysis step might contain multiple analysis procedures and/or methods and serves as a pausing point where users can interact with the system to review the results, to shape the next steps, and to return to previous steps to repeat them with different methods or parameter settings. Analysis of metabolite extract of mouse liver with spiked-in acid standards shows that MetSign outperforms the existing publically available software packages. MetSign has also been successfully applied to investigate the regulation and time course trajectory of metabolites in hepatic liver.

Figure 1

The workflow of the MetSign software.

Figure 2

An example of spectrum deconvolution for direct infusion experiment by second-order polynomial fitting (A) and Gaussian mixture model fitting (B). The points with the red circle are the original experimental data while the points with green triangle are the fitted data.

Figure 3

Results of deconvoluting the isotopic peaks by intensity MSE method for the tentative metabolite assignment. (A) is a segment of the centralized MS spectrum. Based on the match of the m/z values of all metabolites recorded in the MetSign database, the m/z values of the isotopic peaks of three metabolites were matched. (B), (C) and (D) are three deconvoluted isotopic peak profiles of the three metabolites, respectively. In (B), the metabolite is C₄₄H₇₆N₁O₈P₁-[M+H⁺], by fitting {m+1} and {m+2} isotopic peaks with the theoretical intensity, the fitted experimental intensity is given as solid lines. Pearson s correlation coefficient between the deconvoluted isotopic peaks and the theoretical isotopic peaks is 0.95621, indicating a high confidence of tentative metabolite assignment. In (C) and (D), the metabolites are C₄₃H₇₀O₁₀-[M+Na⁺]-²H-10 and C₄₂H₈₀N₁O₈P₁-[M+Na⁺], and the similarities of isotopic peak profiles are 0.99971 and 0.99946, respectively.

Figure 4

Regulation changes of two metabolites in two different physiological conditions. (A) shows the regulation change of a metabolite with a MetSign identity of C₃₇H₅₈O₁₀-[M+Na⁺]-²H-1. The abundance test (pair-wise t-test) shows that the regulation of this metabolite in the sample group DE is up-regulated with a fold change of 2.7 and a p-value of 0.0116. (B) shows the regulation change of a metabolite with a MetSign identity of C₄₅H₈₈N₁O₁₃-[M+K⁺]-²H-12. This metabolite was not detected in the sample group D and the abundance test could not be applied. The Fisher s exact test shows that this metabolite has different regulation between the sample group D and the sample group DE with p-value of 0.0008.

Figure 5

A sample of an unsupervised clustering result for the analysis of metabolite profiles of 10 samples from the sample group D and 15 samples from the sample group DE. The threshold of the appearance frequency of each metabolite detected in all samples was set as f_t = 75% and the PCA method was employed to reduce the data dimensionality. Fluctuations in the heat map are illustrated in the concentration of metabolites between treatments by color-scale. Two-dimensional hierarchical clustering analysis combined with heat map indicates the trends both in treatments and variables. The metabolite regulation grouped with same color in variables clusters reveals the similar attributes among them.

Figure 6

The time course trajectories of a metabolite (MetSign identity: C₆₃H₁₀₀O₆-[M+H⁺]-²H-3) in the sample group D and the sample group DE. This metabolite was labeled with three ²H and had an adduct ion of H⁺ in all samples. The black points, the upper bound and the lower bound of each bar represent the mean, the maximum, and the minimum of the response of this metabolite in all samples at a certain time, respectively.