3d International ICST Conference on Bio-Inspired Models of Network, Information, and Computing Systems

Research Article

Stability cannot be derived from local structure in biochemical networks

Download226 downloads
  • @INPROCEEDINGS{10.4108/ICST.BIONETICS2008.4687,
        author={Domenico Bellomo and Peter van Nes and Marcel J. T.  Reinders and Dick de Ridder},
        title={Stability cannot be derived from local structure in biochemical networks},
        proceedings={3d International ICST Conference on Bio-Inspired Models of Network, Information, and Computing Systems},
        publisher={ICST},
        proceedings_a={BIONETICS},
        year={2010},
        month={5},
        keywords={Biochemical networks structure dynamics network motifs structural stability.},
        doi={10.4108/ICST.BIONETICS2008.4687}
    }
    
  • Domenico Bellomo
    Peter van Nes
    Marcel J. T. Reinders
    Dick de Ridder
    Year: 2010
    Stability cannot be derived from local structure in biochemical networks
    BIONETICS
    ICST
    DOI: 10.4108/ICST.BIONETICS2008.4687
Domenico Bellomo1,2,*, Peter van Nes3,*, Marcel J. T. Reinders3,2,*, Dick de Ridder3,2,*
  • 1: Inf. & Comm. Theory Group/ Bioprocess Tech. Group, Delft University of Technology, The Netherlands.
  • 2: Kluyver Centre for Genomics of Industrial Fermentation
  • 3: Inf. & Comm. Theory Group, Delft University of Technology, The Netherlands.
*Contact email: d.bellomo@tudelft.nl, pvannes@gmail.com, m.j.t.reinders@tudelft.nl, d.deridder@tudelft.nl

Abstract

In recent literature, the relationship between structure and dynamics in biochemical networks has been intensively investigated. In fact, the scarcity of information about such networks has led to attempts to predict some of their dynamic features based exclusively on more easily available structural information. A recent finding relating structure to dynamics is that network motifs (a structural feature) that are structurally stable (a dynamic feature) are enriched in some biochemical networks [Prill 05]. In this work, we systematically investigate the method in [Prill 05] and the assumptions it relies on. Our findings suggest that the conclusions drawn on the considered biological networks (over-representation of structurally stable motifs) cannot be generalized, as they critically depend on a user-defined choice of null model in the motif enrichment analysis. We have further applied the method in [Prill 05] to metabolic networks, which provide an excellent test-bed, as a relatively large amount of information is available about the type, strength and activity of metabolic interactions. For metabolic networks we arrive at the same conclusion: stability cannot be derived from local structure.