M B Cronhjort | PubMed Health

PubMed1995

Hypercycles versus parasites in the origin of life: model dependence in spatial hypercycle systems.

Spatial hypercycle systems can be modelled by means of cellular automata or partial differential equations. In either model, two dimensional spirals or clusters can be formed. Different models give rise to slightly different spatial structures, but the response to parasites is fundamentally different: In cellular automata the hypercycle is resistant to parasites that are fatal in a partial differential equations model. In three dimensions scroll rings correspond to the two dimensional spirals. Numerical simulations on a partial differential equations model indicate that the scroll rings are unstable: The contract by a power law and disappear. Therefore, in three dimensions clusters seem to be the best candidate for the hypercycle resistant to parasites.

Computer Simulation↗

PubMed1992

Intron evolution: a statistical comparison of two models.

The two most frequently occurring explanations for the existence and distribution of introns in the genes of different species are: (1) introns are remnants of the original genetic material. (2) Introns were introduced during evolution. We construct mathematical models corresponding to these two explanations, and calculate the probabilities that the intron distribution in genes from different species coding for actin, alpha-tubulin, triosephosphate isomerase and superoxide dismutase are described by these models. In both models, the branch lengths as well as the structure of the corresponding evolutionary tree is taken into account. Every branch in the evolutionary tree is assumed to have its own individual rate of loss of introns for the first model and rate of gain of introns for the second model. These rate constants are estimated from the actual number of introns. Using the rate constants we stimulate the intron evolution and calculate the probabilities that the actual intron arrangements are produced. The results for actin and alpha-tubulin, which are the two genes we have the most data for, favor the model corresponding conjecture (1), i.e. the idea that introns are old. This contradicts the results from an earlier attempt to model intron evolution where almost the same data was used (Dibb & Newman, 1989, EMBO J. 8, 2015-2021).

Actins↗