Biology Reference
In-Depth Information
molecular constitution, which he described by a mathematical model accounting
for the data obtained by input-output analysis.
4.2. Contents of the black boxes and the 'direction' of cybernetic
modeling
Biological cybernetics uses data from input-output analysis, i.e., from systems
dynamics under controlled conditions, to infer to regulatory instances within the
black-boxed system. These instances are specified primarily on an abstract level,
without reference to their physical realization. On this level, then, the content of
the black box is analyzed as the functional structure of the system that the model
describes - more precisely, a possible functional structure. What is searched for
is a model of the lowest possible mathematical complexity that can explain the
behavior of the system. This criterion is ambiguous; it neither needs to single
out exactly one possible model, nor can one decide in all cases which one of
two models is the less complex (Varjú, 1967). The biological realization might
be more complex than the minimal model anyway. By beginning the analysis
of the systems dynamics from data on systems-level behavior and inferring the
functional structure on a level below, biological cybernetics follows a top-down
modeling strategy.
Although cybernetic models only describe the functional structure of a sys-
tem, cybernetics does not ignore the components of a system; it aims to relate
the low-level regulatory units to cellular and molecular structures. This is done
in a second step that supplements the regulatory analysis of a system. Only
anatomical, physiological, and biochemical analyses can reveal the actual phys-
ical structure that realizes the regulatory functions. As mentioned, cybernetic
and molecular analysis went hand in hand in the case of the squid axon model.
Another case in which such a mixed strategy was applied is the biological
clock. Initially, the molecular basis of the oscillator was completely unknown;
Winfree's model stated minimal requirements for the regulatory mechanism.
With increasing biochemical knowledge, molecular data were obtained, allow-
ing the model to be refined. Modeling is continued within a systems-biological
framework, the latest version by Leloup and Goldbeter (2003) being a prime
example of the success of this research strategy. Similarly, an early cybernetic
model of
E. coli
chemotaxis (Spudich & Koshland, 1975) was supplemented later
on by the mechanistic models cited in Section 3.1. Having in mind such combi-
nations of cybernetic modeling and molecular analysis, it is worth considering
the direction of modeling again. Initially it was top-down, from the system-
level response to the functional structure of the system. At this point, molecular
analysis came in; its results could now be related to the model, the molecules
being regarded as instantiations of the functional components of the system.
