Related Experiment Video
Updated: May 30, 2026
08:59
Morphology-Based Distinction Between Healthy and Pathological Cells Utilizing Fourier Transforms and Self-Organizing Maps
Published on: October 28, 2018
Expression cartography of human tissues using self organizing maps.
Henry Wirth1, Markus Löffler, Martin von Bergen
1Interdisciplinary Centre for Bioinformatics of Leipzig University, D-4107 Leipzig, Härtelstr. 16-18, Germany. henry.wirth@ufz.de
BMC Bioinformatics
|July 29, 2011
Summary
Self-organizing maps (SOMs) simplify complex genomic data by reducing thousands of genes to key metagenes. This machine learning approach offers a clearer view of gene expression patterns across human tissues.
Area of Science:
- Genomics
- Bioinformatics
- Machine Learning
Background:
- High-throughput sequencing and microarray experiments generate vast multidimensional genomic data.
- Analyzing this data requires effective dimension reduction and visualization techniques.
- Self-organizing maps (SOMs) are a machine learning approach for analyzing high-dimensional data.
Purpose of the Study:
- To bridge the gap between SOMs' dimension reduction capabilities and practical gene expression analysis.
- To apply SOMs for characterizing whole-genome expression profiles across diverse human tissues.
- To enhance the visualization and analysis of complex genomic datasets.
Main Methods:
- Application of self-organizing maps (SOMs) to whole-genome expression data from 67 healthy human tissues.
- Reduction of tens of thousands of genes to a few thousand metagenes, representing co-regulated gene clusters.
- Functional enrichment analysis of identified gene modules (spots) using pre-defined gene sets.
Main Results:
- SOM mapping reduced gene expression data to metagenes, revealing tissue-specific and shared molecular properties.
- Tissue-related spots contained enriched gene populations linked to specific molecular processes.
- Metagene-based clustering aggregated tissues into three main groups: nervous, immune, and others, with improved signal-to-noise ratios compared to gene-level analysis.
Conclusions:
- Self-organizing maps provide a more intuitive and informative global view of gene expression modules.
- SOMs offer a superior method for analyzing correlated and differentially expressed genes compared to individual gene analysis.
- The oposSOM R-package is available for implementing these SOM-based analyses.
