Mapping the Genotype–Phenotype Relationship in Cellular Signaling Networks: Building Bridges Over the Unknown - Systems Biology: Networks, Models, and Applications - page 160

previously, cDNA and oligonucleotide microarrays provide gene
expression profiles that help provide information about the underlying
network structure. However, characterizing large-scale networks is
much more challenging, simply because of the computationally intensive
task that is associated with cataloging the activities of the thousands of
genes within the genome. Furthermore, certain assumptions, including
the hypothesis that genes that perform similar functions are expressed
together, while likely to hold locally, are much more suspect when
applied to global network function [91]. What if mRNA and protein
expressions are not correlated? Singular value decomposition (SVD) is
a technique that has been applied to biological network reconstruction
with the aim of resolving these issues [91]. SVD is capable of using
microarray data and generating a set of candidate biological networks.
Further, based on the empirical observation that gene regulatory net-
works are usually large and sparse, the sparsest such candidate
network is chosen as the likely biological outcome. A recent approach
of reverse-engineering was successfully applied to gene regulatory
networks [91].
CONCLUDING REMARKS
Cellular signaling network functions can be discretized into three distinct
tiers, namely, genotypes, “intermediate phenotypes,” and “endpoint
phenotypes.” The first tier involves the transcriptional activity of a
genome, that is, all the proteins that result from the transcription and
translation of genes. Indeed, the genotype effectively delineates the
“parts list” of a given cellular network. The next tier, called “interme-
diate phenotypes,” is composed of the chemical transformations in
which these “parts” are involved, including, for example, the various
protein-protein interactions that mediate responses to signaling events.
Finally, “endpoint phenotypes” are the subsequent transcriptional
and cellular events, such as apoptosis, that ultimately drive cellular
phenotypes.
Numerous experimental technologies have emerged to decipher
cellular signaling networks, and these are easily classified according
to the three-tiered structure. For instance, ChIP-chip methodologies
explain the genotype by identifying which genes are regulated by
various regulatory proteins under specific conditions. Intermediate
phenotypes are characterized by methods such as FRET-based tech-
nologies that delineate which proteins interact with each other.
Techniques like RNAi-based approaches, which indicate the pheno-
typic function of a specific transcript, elucidate how particular
components of a network affect global cellular responses.