Biology Reference
In-Depth Information
Gandrillon, CNRS, UCB-Lyon; Jean-Jacques Kupiec, INSERM,
ENS-Paris; Andras Paldi, EPHE, Généthon-Evry).
6.4
Simulation of the Darwinian model
Computer simulation is another tool with which the validity of a
theoretical model can be tested. We used this approach and created
a computer program which brings into play virtual cells subjected
to the rules of the Darwinian model (Laforge
et al
., 2005; Glisse
et al
., in press). This work was done in cooperation with physicists
of the Pierre and Marie Curie University (Paris). Before describing
our simulations, we need to explain a point of methodology.
Modelling does not consist in reproducing all aspects of reality,
but is bound to be a simplification and abstraction of it. Otherwise,
it is not modelling. Let us take the example of cell modelling
which concerns us here. If we were to create a computer program which
reproduced all the aspects and all the details of a real cell, it would no
longer be a model but a virtual copy of the cell. Since it would be as
complex as the original, this copy would be just as difficult to analyse,
and would therefore be of little use. A model should only seek to
capture one aspect of the reality and help us understand the con-
tribution of this aspect to the real phenomenon which is necessar-
ily always more complex. In our simulations, our aim was not
therefore to mimic a particular situation by precisely describing all
the parameters of a cell, but to analyse the general properties of the
Darwinian model. The question we asked was as follows. If in a real
complex cell there is a Darwinian mechanism based on chance and
selection, what is this mechanism's contribution to the behaviour of
the cell? More specifically we wanted to know whether it is in a
position to create organised tissues reproducibly despite its proba-
bilistic nature. We also wanted possibly to bring to light its non-
trivial properties. Our simulations are, to some extent, thought
experiments assisted by computer, allowing us to explore the model
by going beyond what it is possible to do simply by reflecting.
To perform them, we constructed the very simplest idealised cell
