Genome-Scale Models of Metabolic and Regulatory Networks - Systems Biology: Networks, Models, and Applications - page 250

to study the stimulus-dependent binding of TFs [97], opening up the
possibility of using this technique to map combinatorial interactions
between TFs on a genome-wide scale.
The combination of expression profiling with ChIP-chip as well as
promoter sequence motif analysis has been shown to allow generation
of hypothetical regulatory network structures using a variety of data
integration methods [85,93,98-101]. However, deriving full regulatory
network structures based solely on experimental data appears to be
challenging due to the large quantities of high-quality data that would
be required for such a task. One alternative to this purely data-driven
approach would be to utilize regulatory network structures derived
from databases and primary literature as a starting point for expanding
the network based on high-throughput data. Such a combination of
knowledge-driven and data-driven regulatory network reconstruction
strategies has already been shown to accelerate network reconstruction
in E. coli [57] and yeast [102].
Both knowledge- and data-driven network reconstruction strate-
gies have so far been primarily applied to the two best characterized
microbial organisms, E. coli and S. cerevisiae . Existing genome-scale
regulatory network reconstructions and data collections for these
organisms are summarized in table 8.3.

Integration of Regulation and Metabolism
Regulatory network reconstructions can be represented at different
levels of detail, ranging from connectivity diagrams to kinetic descrip-
tions, depending on the intended application of the resulting network
model. Different in silico modeling approaches have been extensively
reviewed elsewhere [103] and will not be discussed in detail here.

Table 8.3 Examples of reconstructed transcriptional regulatory network
structures and data sets for reconstruction in E. coli and
S. cerevisiae

E. coli full E. coli S. cerevisiae S. cerevisiae

metabolic database metabolic database S. cerevisiae

[57] [80] [102] [124] ChIP-chip [94]

Regulatory 104 123 55 109 203

genes
Target genes 451 762 a 348 418 1296 b

Regulatory — 1468 a 775 945 3353 b

interactions
Regulated 555 — — —

reactions
a Counting each gene in operon separately.
b Includes only high confidence interactions.