IMPROVING A GREEDY DNA MOTIF SEARCH USING A MULTIPLE GENOMIC SELF-ADAPTATING GENETIC ALGORITHM

摘要

We consider the problem of combining a greedy motif search algorithm with a self-adapting genetic algorithm that employs multiple genomic representations in order to find high scoring substring patterns of size k in a set of t DNA sequences of size n. This improves the results of a stand-alone greedy motif search. The encoding schemes used insure feasibility after performing the operations of crossover and mutation and also ensure the feasibility of the initial randomly generated population. The GA's applied in solving this problem employ non-locality or locality representations when appropriate, that is, the GA's adapt to their current search needs. This makes the GA's more efficient.