Biology Reference
In-Depth Information
An interesting problem then arises: 'How should one speak about causality in
complex systems involving many interacting factors?'. Also because cells are
complex systems acting as integrated wholes, linear chains of causes and effects
are virtually absent and 'circular' causality abounds (this volume Hofmeyr;
Westerhoff & Hofmeyr in Alberghina & Westerhoff, 2005). Our intuitive under-
standing of phenomena cannot cope with the overwhelming complexity of
nonlinear interactions among cellular components. The use of mathematical
models is then an indispensable tool to tackle this complexity systematically
(Westerhoff & Kell, this volume; Wolkenhauer & Ullah, this volume; Bruggeman
et al. in Kriete & Eils, 2006). Such models may also be a much more appro-
priate domain of an investigation into the nature of causality than description
in everyday language (e.g. Wagner, 1999). When computations are done with
in silico
models of living systems, integrating all the known molecular properties
of such systems, surprising counterintuitive results appear, results that would not
have been anticipated without the integration of the molecular properties into
mathematical models (Noble, 2002; Westerhoff & Kell, this volume). Examples
of such models can be found in the silicon cell's live model base on the Internet
(Snoep & Olivier, www.siliconcell.net). A similar statement can be made for the
counterintuitive finding of metabolic control analysis that there is not necessarily
a 'rate-limiting' step in a metabolic pathway as was generally believed, but that
control tends to be distributed among various steps (enzymes) of the pathway
(Fell, this volume). Reconstruction of system's behaviour from the behaviours
of the constituent parts can, in principle, be done in two ways:
in vitro
reconsti-
tution and
in silico
modelling. Although impressive results have been obtained,
in vitro
reconstruction of system's behaviour is experimentally rather difficult
and laborious.
In silico
modelling through computational integration of the prop-
erties of and interactions among parts is more promising (Westerhoff & Kell,
this volume; Wolkenhauer & Ullah, this volume). The crux here is that for
integrative computation, binding and kinetic data of all the parts should be deter-
mined such that their values equal, or at least approach, their values
in vivo
,at
the operating state (see also above). Experimental data that are obtained
in vitro
are not necessarily the same as those that are valid
in vivo
, unless the condi-
tions and the states the components are in, are identical. Models may be used
as a hypothesis generator or predictor for desired perturbations, but become
much more powerful after validation through combined experimental-theoretical
approaches.
4.6. What is life?
Biology is the science of living systems and many, if not all, scientists and
philosophers will agree to the statement that 'the cell is the smallest unit of
life'. Although molecules constitute living systems, they are by themselves not