Elsevier

Biosystems

Volume 105, Issue 2, August 2011, Pages 154-161
Biosystems

Visualizing multi-omics data in metabolic networks with the software Omix—A case study

https://doi.org/10.1016/j.biosystems.2011.04.003Get rights and content

Abstract

Systems Biology is a multi-disciplinary research field with the aim of understanding the function of complex processes in living organisms. These intracellular processes are described by biochemical networks. Experimental studies in alliance with computer simulation lead to a continually increasing amount of data in liaison with different layers of biochemical networks. Thus, visualization is very important for getting an overview of data in association with the network components.

Omix is a software for the visualization of any data in biochemical networks. The unique feature of Omix is: the software is programmable by a scripting language called Omix Visualization Language (OVL). In Omix, the visualization of data coming from experiment or simulation is completely performed by the software user realized in concise OVL scripts. By this, visualization becomes most flexible and adaptable to the requirements of the user and can be adapted to new application fields.

We present four case studies of visualizing data of diverse kind in biochemical networks on metabolic level by using Omix and the OVL scripting language. These worked examples demonstrate the power of OVL in conjunction with pleasing visualization, an important requirement for successful interdisciplinary communication in the interface between more experimental and more theoretical researchers.

Introduction

Systems Biology aspires towards a holistic understanding of intracellular processes of living organisms. In this context, the neologism “omics” evolved as a broad term for various scientific approaches to analyzing biological systems (Lederberg and Mccray, 2001). This includes, for instance, genomics: the study of the genomes of an organism, metabolomics: the study of small-molecule fingerprints that specific cellular processes leave behind, proteomics: the large-scale study of the structure and function of proteins, transcriptomics: the study of all RNA (mRNA, rRNA, tRNA and non-coding RNA) molecules produced by the cell, and many more.

Ideally, increasing knowledge about the biological system is gained in a close loop cycle of collecting data from experimental studies, formulating models describing the processes, iteratively fitting model parameters until simulated data correspond to experimental measurements, adapting the experimental designs and elaborating next-generation experiments. Today, high-throughput experimental approaches applied to mutant collections and high-performance simulation lead to a huge and continually increasing amount of multi-omics data concerning the biological system (Covert et al., 2004, Ishii et al., 2007, Weitzel et al., 2007). This leads to an urgent requirement of information visualization techniques in order to integrate available data into holistic models of intracellular processes.

A widespread approach is to describe biochemical systems as networks. Here, the network nodes represent the actors of intracellular processes (enzymes, proteins, compounds, genes, etc.) and edges connecting the nodes describe relations between these actors (conversion, synthesis, degradation, transcription, regulation, etc.). There are different types of biochemical networks that occur as different layers within the whole system. Gene regulation networks, for instance, describe the gene transcription processes. Protein–protein networks describe in which manner polypeptides are combined to form quarternary structures. Signal transduction networks, furthermore, show the complex stimulus-response system of an external stimuli towards the alteration of intracellular molecules by protein and enzyme activation and inactivation.

In this contribution, we are focused on metabolic networks describing the metabolic processes of an organism. Metabolic networks consist of enzymatically catalyzed reactions that transform biochemical compounds (metabolites) into each other. In metabolic networks, certain series of reactions are assembled in metabolic pathways representing special functional modules in the system, for instance, synthesis or degradation of amino acids or secondary metabolites, lipids or energy storage.

Omics data usually stand in direct relation to nodes and edges of metabolic networks. Hence, a network-integrated visualization helps to gain a system wide overview of data in contrast to tables, histograms and function graphs. Here, the software “Omix” is a valuable tool providing extensive and highly customizable information visualization features. As we previously stated in Droste et al., 2009, Droste et al., 2010, Omix is a network drawing tool combined with a programmable visualization framework. Fig. 1 shows a screen shot of the Omix main window. The software is freely available for non-commercial academic purpose (www.13cflux.net/omix).

This contribution will discuss four case study of visualizing omics-data with the software Omix. We choose different kinds of metabolic networks to demonstrate the visualization of metabolome, fluxome and transcriptome data as well as the animation of a time series of multi-omics data. First of all, the visualization abilities of Omix are presented.

Section snippets

Visualization with Omix

In Omix, metabolic networks are formulated as bipartite graph in which reactions and metabolites are represented by two different node symbols as shown in Fig. 2. Likewise, two kinds of edges are used to indicate different types of metabolite-reaction relations. Here, edges with an arrowhead indicate a flux relation, i.e. the metabolite is educt or product of a reaction, called flux edge. The second edge type, the effector edge, represents regulatory effects of a metabolite on a reaction. An

Case Studies on Visualization

During the last years collecting experiences with the programmability of Omix many OVL scripts have been developed representing solutions for visualization of metabolome, fluxome, transcriptome, regulatory effects, carbon labeling states, literature references and thermodynamic data in the context of metabolic networks. This section presents four case studies pointing out the flexibility and impact of Omix.

Conclusions

Systems Biology is concerned with studying biochemical networks. Here, data-intensive high-throughput experimental techniques and high-performance computer simulation lead to a continually growing multitude of data. Hence, a strong need emerged for advanced network visualizations supporting visual analysis. Metabolic networks represent the current knowledge how metabolites interact to accomplish biological functions and respond to environmental stimuli. The diversity of these interactions takes

Acknowledgements

We are grateful to Stephan Noack for scientific discussion and providing data. Furthermore, Evonik Industries deserves out thanks for financial support as part of the BMBF co-funded SysMAP project (project no. 0313704) and the EU-funded SysInBio project (project no. 212766).

References (27)

  • Droste, P., 2011b. Omix Visualization Language—Technical Manual. Institute of Bio- and Geosciences, IBG-1...
  • P. Droste et al.

    Customizable visualization of multi-omics data in the context of biochemical networks

  • P. Droste et al.

    Customizable Visualization on Demand for Hierarchically Organized Information in Biochemical Networks

  • Cited by (82)

    • Interactive Multiresolution Visualization of Cellular Network Processes

      2020, iScience
      Citation Excerpt :

      Multiple tools exist for static visualization of biological networks. Some of the tools used to visualize reaction-based models include Dynetica (Eidum et al., 2014), COPASI (Bergmann et al., 2017), CySBML (König et al., 2012), and Omix (Droste et al., 2011). Although these tools provide useful visualizations of biochemical models, they are implemented as Graphical User Interfaces, which can hinder the creations of pipelines for model creation, visualization, and analysis.

    • Amino acids are key substrates to Escherichia coli BW25113 for achieving high specific growth rate

      2020, Research in Microbiology
      Citation Excerpt :

      Protein synthesis cost was calculated similarly to work of Peebo and colleagues [6] by multiplying protein copy number with sequence length and ATP needed for peptide bond synthesis. Proteome analysis results were visualised with proteomaps [31] and the figure depicting pathways was built with Omix [32]. We started our analysis with comparing growth rate and substrate consumption values between minimal and nutrient-rich conditions.

    View all citing articles on Scopus
    View full text