Abstract
Recently, genetic algorithms (GAs) have received a lot of attention because of their easy-to-use features for solving many engineering problems. They are capable of locating a good approximation in extremely large search spaces with a reasonable amount of computational effort. In this study, we have developed DNA algorithms (DNAAs). The distinction between GAs and DNAAs comes from the fact that GAs take into account only exons whereas DNAAs are concerned about not only exons but also introns. Exons are parts of genes embedded in DNA. Introns are the segments of DNA remaining after the removal of genes. It is said that exons do not contain any information for protein synthesis. It speculated that introns might have some hidden roles. If all parts of DNA had important information, mutations or mal-duplication would be expected to cause fatal problems. It can be hypothesized that introns, therefore, serve as protection for exons against mutations. On basis of this hypothesis, we investigated the role of introns by using an artificial life. Then a string search problem and a knapsack problem were solved by DNAAs in order to evaluate performance of DNAAs. DNAAs performed robustly even under a fairly high ratio of mutation. It was also found that more introns were accumulated near exons whose role seemed more important than other exons.
