approaches, such as MCL [33], NEMO [34], and MINE [35],
have been proposed to identify highly interconnected gene sets.
After obtaining modules of GRNs, an in-depth analysis is carried
out to find functional motifs, which are overrepresented patterns
when compared to random networks with equivalent size. Func-
tional motifs are elementary unit of modules, and they are closely
correlated to regulatory framework of GRNs, for example, bi-fan
regulating motif, feed-forward loop regulatory unit, and dense
overlapping regulons [31, 36]. In the past decade, some motif
detection tools, such as FANMOD [37], MFINDER [31], and
NetMODE [38], have been developed in the research field of
computational biology. After identifying modules and motifs,
gene ontology (GO) and pathway enrichment analysis are con-
ducted to help biologist understand organization of genes within
modules and motifs and their combinatory functional effect, which
promotes mechanism investigation of cellular process in biological
system.3.3 Interpretation
of Hierarchical
Structure in Networks
For some GRNs, they have hierarchical structure, and they can be
divided into topological frameworks in multilevels [29]. As for a
network, topology modules, containing functional motifs, are ele-
mentary structure units, which are assembled into larger module
frameworks. Next, these frameworks are combined in a hierarchical
fashion to form the complete network. For these networks, they
have not only scale-free property but also hierarchical characteris-
tics, in which the mean value distribution of clustering coefficient is
proportional to the reciprocal value of linksn(c(n)~n^1 ), while in
random and scale-free networks, the mean value distribution of
clustering coefficient is independent to links. According to the
hierarchical characteristics of gene regulatory networks, behavior
of the whole biological system can be predicted de novo through
interpreting topological structures in multilevel of GRNs and
explaining interactions of molecules within a hierarchical scale
over the full range of cellular compartments [39].3.4 Comparative
Investigation
of Networks
After reconstructing the gene regulatory networks, comparison
study of GRNs across different species is carried out to reveal
conservative subnetworks in evolution, which is thought to provide
more insights into evolutional mechanism than gene sequence
comparison study [40, 41]. These conservative subnetworks are
essential modules, which are involved in many significant cellular
processes and thought to be response units of internal and external
stimuli in living cell. On the other hand, comparative investigation
of GRNs between different organs for an organism is conducted to
reveal organ-specific subnetworks, which are regarded as key factors
controlling morphological formation and playing important roles
in the organ [42].The Reconstruction and Analysis of Gene Regulatory Networks 145