This paper presents a novel approach algorithm for bimolecular sequences alignment. Sequences comparison is the most important primitive operation in computational biology. There are many computational requirements for a alignment algorithm such as computer memory space requirement and computational complexity (computation time). To overcome the computational complexity of sequence alignment, the presented method first randomly divides the entire bimolecular sequences into several small sequences, and search for a partial near optima solution. After all of the partial near optima searching operations arc completed, the algorithm starts to search for better global optima by scan the new bimolecular sequences that are combined from the optimized small sequences. It allows pairwise alignment in each small sequence and does not apply dynamic programming at any optimization operation. The proposed algorithm also provides highly alignment efficient and very fast performance. Moreover, the proposed algorithm has been implemented in an x86 program, and used to verify the validity of the proposed algorithm and experiment on real DNA and protein datasets.
Cybernetics and Intelligent Systems, 2004 IEEE Conference on 1, pp.602-607