Biology Reference
In-Depth Information
biology (Mayr, 1961). For proper examples of design explanations see Teusink
(Teusink et al., 1998), Bakker (Bakker et al., 2000), and Wouters (1999).
Having read about these various types of explanation and their potential for
systems biology in this topic, we note that the precise conditions under which
they can be used are incompletely defined. The philosophical foundations of
systems biology may be more diverse than those of other sciences; we believe
that they should still be well specified, in order for the quality of this new science
to become outstanding and its results reliable.
4. DESCRIPTION OF MOLECULAR MECHANISMS
USING MODELS
The detailed description of the organization and inner workings of molecu-
lar mechanisms in terms of mathematical models can take many forms, often
depending on the purpose of the description. In more applied settings, where
models are supposed to have interpolative power, the exact details of the molec-
ular mechanisms may not matter too much, as long as the mechanisms are
described to such an extent that the model is a successful predictor. When
constructing such models, the structure of the network, description, and the
parameterization of its processes can be phenomenological. On the other end
of the scale one finds detailed models - 'silicon cell' models - that are care-
ful reconstructions of molecular mechanisms characterized mostly
in vitro
.
Such models aim at being or becoming 'replicas' of cells. In between these
extreme approaches various models can be distinguished that combine the advan-
tages of both approaches. These models often attempt to mimic the network
structure and process characteristics precisely, while parameterization is more
phenomenological and derives from fitting the model parameters to experimen-
tal system data. This 'hybrid' approach also exemplifies the present lack of
knowledge of the kinetic description of many intracellular processes, except for
processes occurring in the central metabolism in yeast,
Trypanosoma brucei
,
and
E. coli
.
Because of its dependence on complex nonlinear systems, systems biology
depends on modeling. The above modeling approaches have therefore been
discussed from an epistemological perspective in some of the chapters in this
topic (Hofmeyr; Schaffner; Westerhoff & Kell; Wolkenhauer & Ullah). Where
some authors emphasize that some of these approaches are less valuable than
others, we conclude from these chapters that systems biology will profit from
as wide a variety of modeling approaches as possible. These should include
various hybrid approaches, e.g., between continuous and discrete models, and
an optimum combination of spatio-temporal-chemical modeling for the systems
that are stiff in temporal, spatial and chemical-concentration dimensions.
