Jeudi 24 janvier 2013
The development of an organism from egg to adult has fascinated scientists for centuries.
In the past several decades, genetics has been very successful in attaching
a function to hundreds of genes whose mutation can lead to developmental defects.
This makes the task of quantitative modeling effectively impossible. We have
instead used geometric methods from dynamical systems theory plus genetics to
posit simple models of phenotype that collapse the activity of many genes onto a
few parameters, that in favorable circumstances can all be fit to data. An implication
is that the space of phenotypes is not impossibly vast, but ’canalized’ into a few
channels. Examples will be given for some well studied developmental paradigms
in model organisms.