Information Technology Reference
In-Depth Information
Chapter 1
Introduction
The topic is organized in three main parts. Part I deals with networks at the level
of the cell, the basic unit of any living organism. In the cell, all kinds of structural
and functional processes are ruled by the intricate interaction of genes, proteins and
other molecules. It is then easy to understand why the complex network approach
has become a common and useful paradigm to investigate, model and understand
cellular processes. The rst part of the topic covers various kinds of cellular net-
works such as: gene networks, protein networks, metabolic networks, protein-folding
networks, as well as signaling networks.
The rst Chapter, by Brilli and Lio, is a general overview on the structure and
function of regulatory networks, gene coexpression networks and protein-protein
interaction networks. Here, the reader will nd a review of the main properties
of a network, and the basic measures to characterize its topology, such as degree
distributions, average shortest path lengths, clustering coecient, assortativity and
network motifs.
The rst step to understand the behavior of gene regulatory systems, is to study
the dynamics of small gene networks. Historically, one of the rst examples of
articial gene circuits to be investigated was the repressilator, a synthetic network
of three genes whose products inhibit the transcription of each other cyclically. In
the second Chapter, Ullner et al. show how to describe the dynamics of coupled
small synthetic gene networks through a set of ordinary dierential equations. In
particular, two dierent types of genetic oscillators, the repressilator and a relaxator
oscillator, with two dierent types of coupling (namely a phase-attractive and a
phase-repulsive coupling) are considered. This is enough to produce rich dynamical
scenarios that include multistability, oscillation death, and quantized cycling.
The drawback of the dierential equation-based approach is the necessity to
know the kinetic details of molecular and cellular interactions, an information that is
rarely available. There is increasing evidence, however, that the input-output curves
of many regulatory relationships are strongly sigmoidal and can be approximated
by step functions. In addition to this, regulatory networks often maintain their
function even when faced with uctuations in components and reaction rates. This
explains the diusion of coarse-grained methods, such as Boolean models. In a
Boolean network the state of each node can assume one of two possible values, and
the interaction between nodes is ruled by a set of Boolean functions. The Chapter
1
