Biology Reference
In-Depth Information
systems biology proper does, the systeomic project degrades network dynamics
to another source of large data sets. This critique, however, applies only to
attempts of 'realistic' modeling that aim at doing without any simplification. All
extant systems-biological models do justice to the explanatory attitude of model
building. They gain an explanatory status by idealization of the modeled system,
which allows an exploration of which components and relations actually bring
about the dynamics of the network under investigation.
9. CONCLUSION
We have reviewed the field of systems biology and its variegated roots to char-
acterize its structure, and have addressed some epistemological and ontological
issues regarding concepts on which the field relies and that seem to us to require
clarification. Systems biology is largely tributary to genomics and other 'omic'
disciplines that generate vast amounts of structural data. 'Omics', however, lack
a theoretical framework that would allow to use these data sets as such (rather
than just tiny bits that are extracted by advanced data-mining techniques) to
build explanatory models that help understand physiological processes. Systems
biology provides such a framework by adding a dynamic dimension to merely
structural 'omics', making use of bottom-up and top-down models. The former
are based on data about systems components and the latter on systems-level
data. We traced back both modeling strategies (which are often used to delineate
two branches of the field) to the modeling of metabolic and signaling pathways
in the bottom-up case, and to biological cybernetics and systems theory in the
top-down case. We identified three roots of systems biology: pathway modeling,
biological cybernetics, and 'omics'. The data richness one encounters in both
branches of systems biology stems from the 'omic' disciplines. Both other roots
were found to be comparatively data-poor but strong in model building. There-
fore, we regard systems biology as merging modeling strategies (supplemented
by new mathematical procedures) from data-poor fields with the data supply
from a field that is quite deficient in explanatory modeling.
An epistemologically important problem of the top-down approach to sys-
tems biology arises from the fact that the systems under investigation are large
and must be decomposed into subsystems in order to model them in a way
that allows to explain their dynamics. Decomposition into modules is performed
according to either functional or structural criteria. While the latter way, usually
called 'unbiased modularization', distorts the picture of the system to a lesser
degree than functional decomposition, it gives up the established link between
physiology and function. We have argued that functionality had to be reintro-
duced if one is interested in statements about the biological 'meaning' of the
dynamics of a system. As we view scientific explanation not as mere depiction
