392163 Random-model of simulated evolution (Pj) (SoSe 2009)

This project is about the simulation of evolution.

Based on Darwin's fundamental observations we have the following general scheme:

1. The individual organisms in a population vary in some random way.
2. They overproduce (if the available resources allow).
3. Natural selection favors the reproduction of those individuals that are best adapted to the environment.
4. Some of the variations are inherited to the offspring.

Here, an individual may be represented by its DNA or protein sequence,
its genome, or any other collection of characteristic data items, even
bitmaps showing some type of animals are conceivable. The production of
new individuals may happen by random events, through recombination with
other individuals (a.k.a. sex), or a combination of both. A fitness
function determines which individuals will form the next generation.

The goal of this project is to write a software that simulates this
scheme through the definition of individuals, species, mutational
operators, fitness functions, and a control structure working on top of
all this. The resulting tool may be used to create data sets for the
experimental assessment of multiple sequence alignment, genomic
distance, and phylogenetic tree reconstruction algorithms.
Technically, the implementation will be done in Java, following a
modular, service-oriented software architecture. The user interface
shall be graphical, and a visual output is desired.

An extension of the project may be the implementation of a game where
the goal of the players is to define individuals with properties such
that they produce a maximum amount of offspring in the course of
evolution. Obviously, many particular instances of such a game are
possible ...

Requirements for participation, required level

Prerequisites: Knowledge of Java programming, basic insights in evolutionary processes.

External comments page

http://wiki.techfak.uni-bielefeld.de/gi/Teaching/2009summer/RoseP