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

Area of Science:

• Metabolomics
• Data Mining
• Bioinformatics

Background:

• Global metabolomic analysis aims to identify metabolic markers from high-throughput data.
• Metabolite clustering groups similar concentration profiles but lacks predefined cluster numbers.
• Identifying true biological groups within complex datasets remains a challenge.

Purpose of the Study:

• To present a data mining approach for analyzing metabolite intensity profiles.
• To utilize one-dimensional self-organizing maps for metabolite-based clustering and visualization.
• To identify relevant marker candidates from complex metabolomic data.

Main Methods:

• Employed one-dimensional self-organizing maps (SOMs) for unsupervised clustering of metabolite profiles.
• Applied the method to mass spectrometry data from Arabidopsis thaliana wound response.
• Utilized clustering and visualization capabilities for marker identification.

Main Results:

• Demonstrated the effectiveness of SOMs in clustering metabolite intensity profiles.
• Successfully identified relevant groups of markers in the Arabidopsis thaliana case study.
• Showcased the visualization capabilities for analyzing marker candidates.

Conclusions:

• Specialized SOMs provide insight into complex pattern variations in metabolite profiles.
• The visualization approach facilitates identification of metabolite groups by offering an overview.
• This method effectively supports researchers in analyzing numerous clusters when the number of biological groups is unknown.