Hamiltonian Path Approach to Construct the Multiple DNA Sequence Alignments and Solve DNA Sequence Assembly with Bioinformatics Computing

• Main Authors: Feng Yu Chang, 張峰瑜
• Other Authors: Shen-Hui Ho
• Format: Others
• Language: en_US
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/88d7p7

碩士 === 國立臺北大學 === 電機工程學系 === 102 === In the past 20 years, DNA sequence alignment has been a big problem. As the number of candidate evolutionary trees grow exponentially with the large number of species, an exhaustive search over all trees on an electronic computer is not feasible. Based upon bioinformatics parallel computing, the use of multiple sequence alignment methods along with a newly developed DNA sequence reassembly algorithm constructs all evolutionary trees in the solution space. At first overlapping fragments of the DNA sequences are used to construct all overlapped paths, and through a Hamiltonian path and evolutionary trees. Secondly, multiple sequence alignments are applied, to find the optimal overlapping fragment path resulting in the DNA sequence reassembly. In other words, using Hamiltonian path and a corresponding optimal evolutionary tree can reassemble DNA sequences. An optimal bioinformatics computing containing massive storage and parallel computations is proposed for solving the bottlenecks of time complexity, improving the space and evolution of the trees to find the best fragments of overlapped DNA sequences overlap, and making it more efficient. The time complexity on using bioinformatics optimization algorithms to construct all evolutionary trees and Hamiltonian paths is estimated O(n2) polynomial bound.