next up previous
Next: Lymphocyte dynamics Up: Overview of research 1993-1998 Previous: Whole Heart modeling

Research Area's : Spatio temporal dynamics Whole Heart modeling Evolution Lymphocyte dynamics Development

Publications: PhD theses (D1-8) 1993 (1-13) 1994 (14-25) 1995 (26-41) 1996 (42-51) 1997 (52-70) 1998 (71-97)

Evolution

Eco-evolutionary dynamics


Evolutionary models have long neglected to look explicitly to the evolution of complexity. Examining what did happen in evolution Maynard Smith and Sathmary postulated that the occurrence of multiple levels of selection is crucial. From the early nineties we pioneered studying the evolutionary consequences of space in simple, individual based ecological interaction models, and studied what does happen. We have shown that new levels of selection automatically arise, in the form of generic spatio-temporal patterns which enslave the dynamics of the self-replicating entities (D4,D7,5,27,28,65,66). The direction of evolution of the replicators tends to be such as to stabilize the patterns in which they find themselves. This leads to many counterintuitive results, because the spatio-temporal dynamics leads to long term information integration, e.g. variants which are less fit over each time-slice of many generations, can nevertheless persist or take over in the long run. Thus, selection at multiple levels provides counterexamples for the general held 'taboo', that only immediate benefits should be considered in evolution. Within such multilevel systems processes as division of labor, and conflicts between levels of selection occur and we have shown that pattern formation is a a prerequisite for exploitation of long term information storage (19,20). The latter work was the first to use as modeling strategy the focus on side-effect of evolutionary optimization towards a 'trivial' target.
see review paper

Three mechanisms of sympatric speciation were elucidated: (a.) in coevolving predator/prey or host parasitoid systems sympatric speciation happens due to oscillatory dynamics and interlocking ecological and evolutionary time scales (D5,36), (b.) the long standing question of the stability of the sexual continuum was finally solved analytically in terms of Turing and resonance type pattern formation in phenotype space (61), and (c.) sexual selection can lead to speciation provided the trade preference mapping has certain properties. These properties are realized in below water color perception, and the model nicely explains speciation and the current loss of species in cychlid fishes.

Molecular Evolution


Evolution is often portrayed as an optimization process in a rugged fitness landscape, with many local peaks. Our work (with Schuster's group, Vienna) on RNA evolution, incorporating the (computable) non-linear mapping between RNA primary and RNA secondary structure, has led to a modification of this (spin-glass like) picture: RNA forms a rugged fitness landscape with long neutral paths connecting the peaks. In collaboration with Santa Fe Institute we characterize the dynamics on these landscapes analytically. The resulting evolution reconciles adaptionist and neutralists views on evolution. Evolutionary trajectories on these neutral paths lead in a stationary landscape to flatter parts (i.e. relative insensitive to mutations) while in coevolution they lead to more rugged parts (D2,6,21) In other coevolutionary models this effect occurs as well and we have shown that robustness to mutations and robustness to environmental differences are inversely correlated (p).

Evolution of Immune repertoire


We have shown in several models that the diversity of vertebrate immune repertoires does not seem to reflect primarily the diversity of pathogens. It can be explained as evolutionary enhanced side-effect of somatic recombination (p) or in terms of constraints on self-tolerance avoiding auto-immunity (3,p).


next up previous
Next: Lymphocyte dynamics Up: Overview of research 1993-1998 Previous: Whole Heart modeling
Paulien Hogeweg
1999-03-05