eSOMet

Motivation

Modern high-throughput techniques like GC-MS facilitate the identification and quantification of hundreds of metabolites of a biological system, thus covering central parts of the metabolome. Due to the amount and complexity of obtained data, there is an increasing demand for the development of appropriate computational analysis methods. Here, we present a novel analysis pipeline specifically designed for high-throughput based metabolomics data, which enables the detection of hierarchical relationships within different metabolic patterns measured under various environmental conditions and for different genetic backgrounds.

A novel analysis pipeline for metabolome data analyses

eSOMet is a new software that combines established algorithms like hierarchical cluster analysis (HCA) with modern methods like emergent self-organizing maps (ESOMs) and integrates those in an analysis pipeline for metabolome analysis and visualization. Underlying relationships within a series of metabolomics data are deduced. The functionality of the tool covers the ability of biomarker discovery, detection of statistical outliers and the automatic mapping of detected metabolic differences onto KEGG metabolic pathway maps Biomarkers are metabolites with a characteristic concentration pattern for an analyzed environmental condition or genetic background.

Webstart

What is the easiest way to launch eSOMet? Simply click on the Webstart launcher on the right side. If you have Java 1.6 (or higher) properly installed, eSOMet should launch automatically. Java-Webstart will inform you that eSOMet has been signed by a not trustworthy source. However, the certificate that was used for signing eSOMet is placed under the hierarchy of the Deutsche Telekom Root CA 2 which is not classified as trustworthy yet. Although we cannot provide any warranty, it is most unlikely that eSOMet does any harm to your system. Therefore, you should ignore this Webstart warning.

Get the Jar

If you do not want to use the Java Webstart technology, you can download the application as an executable Jar file. Depending on your system configuration you should be able to run eSOMet by simply double-clicking the downloaded Jar file. For details, we refer to the user manual.

You want to look behind the scenes? The Jar file contains not only the binaries but also the source code.

Test case data

In cooperation with the Department of Bioinformatics and Biochemistry at the Technische Universität Braunschweig we analyzed a time resolved metabolomics datasets containing in total 126 metabolic patterns of Corynebacterium glutamicum cells grown with different carbon sources. The complete analysis can be downloaded here. The zip-file containes an *.eso-file that can directly be loaded into eSOMet. Additionally, the plain data in *.cvs-format and an image of the relavant parts of the KEGG database, suitable for usage with eSOMet are provided. Note that you may only use the KEGG-database if you agree to their terms of conditions.

Printed version

A detailed user manual is available as PDF-file. Simply click on the button to start the download.

Poster

A very comprising overview about eSOMet can be gained by having a look on our poster, that we presented on the ICSB 2008.

Publication

Isam Haddad*, Karsten Hiller*, Eliane Frimmersdorf, Beatrice Benkert, Dietmar Schomburg and Dieter Jahn, (2009)
An emergent self-organizing map based analysis pipeline for comparative metabolome studies, In Silico Biology 9, 149 - 162
Link | pdf | Pubmed | BibTeX

* Both authors contributed equally

License

eSOMet and the source codes are released under the the GNU Lesser GPL 3.0 . The design of this website and the poster, as well as the content of the manual are licensed under Creative Commons by-nc-sa. If these conditions do not fullfill your needs, feel free to contact us in order to find a solution that fits your purposes.

Authors

eSOMet was developed by Karsten Hiller and Isam Haddad at the Institute of Microbiology at the Technische Universität Braunschweig, Germany, headed by Dieter Jahn.

Contact (Impressum)

For any comments, questions, suggestions or trouble, please feel free to contact us:

Isam Haddad

℅ Inst. f. Microbiology
Technische Universität Braunschweig
Spielmannstraße 7
38106 Braunschweig
Germany

Phone: +49-531-391 5810
eMail: i.haddad@tu-bs.de

Acknowledgements

This work was funded by the German Federal Ministry of Education and Research (BMBF), Grant No. 031U110A/031U210A and Grant No. 0313980D.