Biosequence Motif Discovery using Evolutionary Computation
- Members: Michael Lones, Andy Tyrrell
- Website: http://www-users.york.ac.uk/~mal503/motif.htm
- Start Date: June 2006
We have developed an evolutionary algorithm based system for identifying regulatory motifs in the promoter regions of genes. Evolutionary algorithms provide an interesting approach for motif discovery problems since they offer both a global approach to search and flexibility in terms of representation and scoring function. Arguments along these lines can be found in this review paper:
The Evolutionary Computation Approach to Motif Discovery in Biological Sequences [PDF]
M. A. Lones and A. M. Tyrrell, Proc. Workshop on Biological Applications of Genetic and Evolutionary Computation (BioGEC),
GECCO2005, June 2005.
In our initial approach, we used within-population data clustering to promote search for multiple, diverse, transcription factor binding sites. Details are available in this paper:
Regulatory Motif Discovery Using a Population Clustering Evolutionary Algorithm [PDF] [link]
M. A. Lones and A. M. Tyrrell, IEEE/ACM Transactions on Computational Biology and Bioinformatics 4(3):403-414, July-September 2007.
We are currently looking at how co-evolution may be used to search for transcription factor binding sites and higher order regulatory motifs (such as cis-regulatory modules) in parallel. Our initial results suggest that this approach promotes search for weakly conserved transcription factor binding sites in long promoter sequences. Details appear in this paper:
A Co-Evolutionary Framework for Regulatory Motif Discovery [PDF]
M. A. Lones and A. M. Tyrrell, Proc. IEEE Congress on Evolutionary Computation, September 2007.
For further information about field of research, visit our:
Bibliography of Biosequence Applications of Evolutionary Algorithms.