文摘
Metabolomics studies generate increasingly complex datatables, which are hard to summarize and visualize withoutappropriate tools. The use of chemometrics tools, e.g.,principal component analysis (PCA), partial least-squaresto latent structures (PLS), and orthogonal PLS (OPLS),is therefore of great importance as these include efficient,validated, and robust methods for modeling information-rich chemical and biological data. Here the S-plot isproposed as a tool for visualization and interpretation ofmultivariate classification models, e.g., OPLS discriminate analysis, having two or more classes. The S-plotvisualizes both the covariance and correlation between themetabolites and the modeled class designation. Therebythe S-plot helps identifying statistically significant andpotentially biochemically significant metabolites, basedboth on contributions to the model and their reliability.An extension of the S-plot, the SUS-plot (shared andunique structure), is applied to compare the outcome ofmultiple classification models compared to a commonreference, e.g., control. The used example is a gaschromatography coupled mass spectroscopy based metabolomics study in plant biology where two differenttransgenic poplar lines are compared to wild type. Byusing OPLS, an improved visualization and discriminationof interesting metabolites could be demonstrated.