Skip to main content

Advertisement

Figure 3 | BMC Plant Biology

Figure 3

From: MeRy-B: a web knowledgebase for the storage, visualization, analysis and annotation of plant NMR metabolomic profiles

Figure 3

Examples of Visualization and Statistical Analysis results for the tomato project T06002. Screenshots from the various visualization and statistical tools. The user selected the tomato project T06002 (a), the composition overview of the samples (b), visualization of the NMR spectra according to tissue criteria (c), visualization of the statistical analysis results (d) and a zoom on one specific spectrum (e). MeRy-B provides statistical analysis facilities within each project. First, the experimental factors and individual samples (rows) and the spectral region variables (columns) for construction of the initial data matrix must be chosen. Second, a statistical analysis workflow must be selected from a list of proposals. Workflow typically begins with standardization of the data, followed by data reduction by analysis of variance (ANOVA) to select the meaningful variables (p-value threshold 0.05). An unsupervised method, such as principal component analysis (PCA), can then be used, if desired, to determine a set of variables from the inputs that can be used to classify the samples into factor groups. An ANOVA test can then be applied to each variable of the set, generating box and whisker plots making it possible to check the relevance of the discrimination. If variables are of the analytical type, it may be important to ensure that they are not affected by an analytical artifact (such as chemical shift). Such checks can be carried out with the Spectra overlay tool, which can be used to visualize all the spectra of an experiment, overlaid in a single graph.

Back to article page