Metabolite-based clustering and visualization of mass spectrometry data using one-dimensional self-organizing maps

2008

Clustering and Visualization of Metabolite Data Using Self-Organizing Maps

Sample size: 837 publication 10 minutes Evidence: moderate

Author Information

Author(s): Meinicke Peter, Lingner Thomas, Kaever Alexander, Feussner Kirstin, Göbel Cornelia, Feussner Ivo, Karlovsky Petr, Morgenstern Burkhard

Primary Institution: University of Göttingen

Can one-dimensional self-organizing maps effectively identify and visualize metabolic markers in mass spectrometry data?

The proposed self-organizing maps provide a useful visualization tool for identifying relevant groups of metabolites in complex datasets.

The study identified 837 high-quality marker candidates based on a conservative p-value threshold.
One-dimensional self-organizing maps effectively visualize complex metabolite data.
The approach allows for the identification of both expected and unexpected metabolic markers.
Clustering results were validated through comparison with traditional methods like hierarchical clustering.
Significant metabolic changes were observed in response to wounding in Arabidopsis thaliana.
Markers related to jasmonic acid biosynthesis were specifically highlighted in the analysis.
The method demonstrated robustness against variations in data quality.
Visualizations provided insights into the temporal dynamics of metabolite accumulation.

This study shows a new way to group and visualize metabolites from plant samples, helping scientists find important markers more easily.

The study used one-dimensional self-organizing maps to cluster and visualize metabolite intensity profiles from mass spectrometry data.

Potential biases may arise from the selection of marker candidates based on statistical significance.

The method relies on the quality of the mass spectrometry data and may not perform well with low-quality markers.

The study focused on two plant lines: wild type Arabidopsis thaliana and a jasmonic acid-deficient mutant.

10^-6

p<0.000001

Access the complete publication on the publisher's website