Regulatory networks and connected components of the neutral space*
A look at functional islands
Bioinformatics Group, Dept. of Computer Science, University of Leipzig, Leipzig, Germany
Revised: 1 April 2010
Published online: 17 June 2010
The functioning of a living cell is largely determined by the structure of its regulatory network, comprising non-linear interactions between regulatory genes. An important factor for the stability and evolvability of such regulatory systems is neutrality – typically a large number of alternative network structures give rise to the necessary dynamics. Here we study the discretized regulatory dynamics of the yeast cell cycle [Li et al., PNAS, 2004] and the set of networks capable of reproducing it, which we call functional. Among these, the empirical yeast wildtype network is close to optimal with respect to sparse wiring. Under point mutations, which establish or delete single interactions, the neutral space of functional networks is fragmented into ≈ 4.7 × 108 components. One of the smaller ones contains the wildtype network. On average, functional networks reachable from the wildtype by mutations are sparser, have higher noise resilience and fewer fixed point attractors as compared with networks outside of this wildtype component.
© EDP Sciences, Società Italiana di Fisica, Springer-Verlag, 2010