Mapping the Genotype–Phenotype Relationship in Cellular Signaling Networks: Building Bridges Over the Unknown - Systems Biology: Networks, Models, and Applications - page 161

Despite the enormous wealth of data that each of these technologies
generates, these data are not yet continuous. For example, although it
is possible to measure the function of an individual protein and per-
haps even how that protein affects network-level behavior, it remains a
challenge to describe how the network is a function of the simultane-
ous interaction of thousands of such proteins. Indeed, the
understanding of integrated functions remains elusive. Furthermore,
systems biology, which attempts to model the integrated properties of
cellular networks, cannot adequately generate all the possible combi-
nations of all the possible experimental conditions; realizing the point
at which the repertoire of tested experimental conditions becomes
sufficient is therefore an essential property that is not yet defined. The
evolvability of biological systems, an important property of livelihood,
is similarly not well characterized.
Nevertheless, a number of approaches are being developed in an
attempt to bridge these knowledge gaps. The simplest of these have
involved enhancing the experimental technologies that are used to
decipher the genotype-phenotype mapping. Improving microarrays so
that they map signaling networks and their individual components, for
instance, is an ongoing research avenue. More significantly, to compen-
sate for the types of information that may never be deciphered from
experiment, several mathematical approaches that account for the
degree of unknowability have been developed. For instance, it has
been shown that the spectrum of possible cellular phenotypes can be
narrowed by virtue of a “constraint-based” approach, which involves
identifying constraints and stating them mathematically. In addition,
reverse-engineering biological networks (e.g., generating a set of can-
didate networks from microarray data) has had success in describing
metabolism and regulation and has shown signs of initial promise in
signaling networks.
Nonetheless, direct relations between external factors, or even
specific internal factors, and internal responses are rare. For instance,
although specific factors are believed to uniquely specify particular
responses in the cell-fate control of stem cells, few such factors have
been identified [92]. Instead, stem cell-fate control relies principally
on common, developmentally conserved signaling processes that are
involved in multiple stem cell-fate decisions. Thus, the challenge
remains to discriminate the combinations of inputs to signaling net-
works and the resulting phenotypes that drive higher-level behavior.
ACKNOWLEDGMENTS
We thank the National Institutes of Health for financial support.