Chaos Automata
One of the biggest problems in bioinformatics today is determining the function of a protein. Since experimental techniques are still costly in terms of time and money, computational techniques are often used to infer protein function. This research explores the use of a model termed chaos automata in transforming sequence data into fractal visualizations. As more is learned about the behavior of chaos automata, it is the hope that this model can be used as a method to infer protein function from protein sequence characteristics.
Starting in 1990 using chaos games, Jeffrey showed that four cornered chaos games allowed patterns in DNA sequences to be visualized. In this
technique, each DNA nucleotide was associated with one of the four vertices of a square and the DNA sequence data was used to select the
vertices. Later, Ashlock extended the technique by developing a model termed chaos automata. Chaos automata combine iterated function systems with finite state automata where one function of an iterated function system is associated with each state in a finite state automaton. A chaos automaton functions by taking sequence data as input to drive the selection of states and thus can select which function of the iterated
function system to apply based upon patterns in the input sequences. Upon entering a state the function of the iterated function system is used to
calculate and output an (x, y) coordinate. The collective coordinates produce a fractal image. In his research, Ashlock used evolutionary
algorithms to evolve the chaos automata parameters to visually distinguish between the introns and exons of corn (Zea mays) DNA.
As chaos automata are rich models with many possibilities that have not been fully explored, this research explores the use of the chaos
automata model in transforming sequence data to fractal visualizations. The current goal is to learn more about the relationship between the fractal images produced by chaos automata and the properties of the input sequences. Like the previous research, the chaos automata will be evolved by evolutionary algorithm.
References
H. J. Jeffrey, "Chaos game representation of gene structure,"
Nucleic Acids Research, vol. 18, no. 8, pp. 2163-2170, 1990.
D. Ashlock, and J. Golden, "Chaos automata: iterated function systems with memory,"
Physica D, vol. 181, no. 3-4, pp. 274-285, July 2003.