Biology Reference
In-Depth Information
root of the bottom-up branch of systems biology. We will discuss in Sections 6.1
and 7 the differences between pathway models and the related kind of systems
biological models.
This sketch regards model building as a process starting from given kinetic
data. However, the data must be collected first. Biologists often start by sin-
gling out a metabolic or sensory capacity that is to be analyzed and eventually
modeled. The next step, after the characterization of the capacity, is the bio-
chemical or physiological analysis of the pathway, yielding the components
of the pathway. Third, these components have to be isolated, characterized,
and their kinetic parameters must be determined. This yields the data set from
which mathematical modeling can start. While modeling itself is bottom-up,
the whole process of gathering the data is a top-down process, starting from
a higher-level capacity and proceeding to the contributions of the lower-level
components of the system. (For an analysis of a historical case see Darden &
Craver, 2002.)
To summarize, the process of model building in the tradition of metabolic
and signaling pathway research typically consists of the following steps: (i)
singling out a capacity of the organism or cell; (ii) experimental breakdown of
the underlying system into its components, providing the data for the modeling;
(iii) synthesizing a model from subsets of the data, which may involve several
steps of simplification and refinement; (iv) analyzing the model analytically or
numerically, i.e., by running simulations; and (v) assessing the reliability of the
results that the model generates.
4. THE SECOND ROOT OF SYSTEMS BIOLOGY: BIOLOGICAL
CYBERNETICS AND MATHEMATICAL SYSTEMS ANALYSIS
Biological cybernetics concentrates on the modeling of regulatory processes that
occur in complex systems. Like pathway modeling, it starts from the description
of systemic capacities, most often ones that are related to signal and stimulus
processing. But in contrast to pathway modeling, it does not aim for an empirical
analysis of the investigated capacity at the molecular level before modeling the
system. Molecular and cellular processes are black-boxed and empirical data are
collected at the systems level by input-output analysis. Biological cybernetics
thus tries to identify the regulatory processes that are required to produce the
observed systemic response at an abstract level. Relating the systems output to
controlled stimulus input allows inferences with respect to frequency filtering,
signal delay, and the amplification characteristics of the system. The regulatory
instances found by such an analysis are often depicted as equivalent electric
