Table 1.
A comparison between BioNetApp, Mzmine2, Graphle, and Biological Networks.
Fig 1.
(A) Time points available for this experiment (in our case study we chose all three time points). (B) Analysis type to perform. (C) Molecules (peaks) frequency filter across all samples (cross-board filter) and also per group of selected features (group-based filter).
Fig 2.
Correlation analysis window with single circle layout.
(A) Molecular IDs. (B) Interactive graphic display of molecular correlations in a single circle display while applying the Pearson correlation calculation. (C) Information panel display for rich set of information about the selected (highlighted) molecules in the graph. Color scheme is used to show the correlation directions: red indicates positive correlation while blue indicates negative correlation. The encircled color areas readily demonstrate clusters of strong molecular correlation.
Fig 3.
Correlation analysis window with multiple circle layout.
Three-circle network graph display illustrating the up-regulated molecules in green borer nodes(A), down-regulated in red border (B), and other (not differentially expressed) molecules with no border (C). The fold change threshold is set at 2, meaning that molecules regulation that don’t meet this threshold are grouped in the third neutral circle (C). The fold change value can be adjusted accordingly.
Fig 4.
(A) Meta information for molecule 901.7314. (B) List of correlated molecules. Further molecule details can be invoked by highlighting a molecule in this list and choosing wither the “Show Element” or “Show Correlation” buttons. (C) Molecular concentration levels across all samples. (D) Concentration details for molecule 901.7314. (E) Concentration details for correlated molecule 903.7465. (F) Correlation graph showing the expression data of the two molecules. Each point represents the expression levels of both molecules (x-axis and y-axis) in the same sample.
Fig 5.
Comparative and distribution analysis.
(A) A tree describing the molecules and their corresponding samples and time points. (B) Box plot displaying the concentration of a selected molecule across time points while applying the Robust Linear fitting. (C) Concentration distribution plot of the molecule across all samples. (D) and (E) display the same plots but for two sample groups (control and test) side-by-side for ease of comparison.
Fig 6.
(A) Clustering averages across all samples, (B) Cluster 2 molecules concentration details across all samples.