Mapping the Genotype–Phenotype Relationship in Cellular Signaling Networks: Building Bridges Over the Unknown - Systems Biology: Networks, Models, and Applications - page 137

with fluorescence resonance energy transfer (FRET)-based technologies,
two-hybrid assays, RNA interference (RNAi)-based methodologies,
and other approaches that delineate the chemical reactions that com-
prise signaling networks. Endpoint phenotypes are measured by
migration and growth assays, flow cytometry, and other such tech-
niques that quantify global cell behaviors.
Although each of these technologies generates specific types of data,
these resulting data sets are not yet continuous. For example, there
exist data on what genes are differentially regulated by interferon alpha,
beta, and gamma stimulation, and on what intracellular kinases are
activated by the corresponding cell surface receptors [1]. However, how
combinations of these stimuli affect cellular phenotype or how different
cell types respond to the same stimuli remains to be discovered. Thus,
mathematical frameworks that can quickly assess a variety of conditions
and account for inherent “unknowability” would be of tremendous value.
This chapter describes the promise and limitations of the experimen-
tal technologies that are used to characterize each of these levels of the
genotype-phenotype relationship in cellular signaling networks. In
addition, the need for mathematical techniques to both integrate these
data sets and to connect the disconnected parts of a network recon-
struction will be discussed.
DECIPHERING GENOTYPES
A genome annotation provides a “parts list” of all the potential protein
components in a cellular system. Thus, the genotype places constraints
on a signaling network by restricting what proteins are even present.
The differentiated states of a cell further restrict which genes are
expressed and thus which associated proteins are synthesized. To date,
many cellular signaling network reconstructions are therefore “gener-
alities” in which a given cell type may only possess a subset. Table 5.1
indicates various experimental techniques that are used to characterize
the genotype, including the advantages and disadvantages of each.
Existing genome sequencing techniques have led to a tremendous
advance in genomic data with the genomes of more than 250 organi-
sms now available [17]. Further advances in sequencing techniques,
called Ultra-Low Cost Sequencing (ULCS) methods, will result in an
explosion of even more sequence data [18]. Such methods will lead to
an increased resolution in cellular signaling network reconstructions.
For example, cost-effective sequencing may allow a researcher to
evaluate the genotype of a variety of cancer tissue samples leading to
a catalog of all mutations that generate cancerous phenotypes. As
another example particularly relevant to cellular signaling, such ULCS
methods would also allow for the profiling of B-cell and T-cell receptor
diversity.