Block Alignment: An Approach for Multiple Sequence Alignment Containing Clusters

• Main Authors: Bo Weng Xiao, 蕭博文
• Other Authors: Professor R. C. T. Lee
• Format: Others
• Language: en_US
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/19616460760964410139

碩士 === 國立暨南國際大學 === 資訊工程學系 === 92 === Multiple sequence alignment is a fundamental problem in computational molecular biology. It has been known as an NP-hard problem. To find its optimal solution will take a lot of time. For a reasonable wait and an acceptable solution, we have progressive methods. These methods perform pairwise alignment, and then combine them in to a multiple sequence alignment. In this thesis, we focus on multiple sequence alignment containing clusters. We try to take another view point to deal with sequence alignment. We use a matrix to present a sequence. Every sequence will be represented as a matrix. After two sequences (matrices) are aligned, the result of the alignment will again be represented by a matrix and then the original two sequences (matrices) will be discarded. That is, the result of aligning a set of sequences will always be considered as a block and represented by a matrix. This is thus different from the old ways in which only two sequences are aligned, not a group of aligned sequences and another group of aligned sequences. In this thesis, we will show some experimental results to test our proposed method. Block alignment outperforms those progressive methods for sequences containing clusters.