Genomic sequence search and clustering using Q-gram

Pao Yue-kong Library Electronic Theses Database

Genomic sequence search and clustering using Q-gram


Author: Yuen, Man-chun
Title: Genomic sequence search and clustering using Q-gram
Degree: M.Phil.
Year: 2007
Subject: Hong Kong Polytechnic University -- Dissertations.
Department: Dept. of Computing
Pages: viii, 85 leaves : ill. (some col.) ; 30 cm.
Language: English
InnoPac Record:
Abstract: With the advances in technologies, the amount of biological data such as DNA sequences and microarray data have been increased tremendously in the past decade. In order to obtain knowledge from the data, e.g., enhancing our understanding of the evolutionary changes and the causes of those severe diseases, one has to search for patterns from the databases of large size and high dimensionality. Information retrieval and data mining are powerful tools to extract information from the databases and/or information repositories. In the past several years, there have been attempts to apply these two branches of intelligent techniques to different bioinformatics applications. However, the performance of these existing techniques has not been optimized due to the characteristics of and requirements from biological data, e.g. extremely long genomic sequences with high dimensionality, and interpretable search/mining results. In this thesis, we focus on how to improve the searching and the clustering performance in genomic sequence databases. A Q-gram based genomic search (QgramSearch) algorithm and a Q-gram based genomic sequence clustering (QgramClust) algorithm are proposed. Our QgramSearch can efficiently search the homologous database sequences to a query sequence. It makes use of two novel hashing techniques to enhance the efficiency of indexing and retrieval. These two hashing techniques can better capture the overlapping characteristics in the Q-gram based index. As demonstrated by the experimental results, they run faster than the existing data structures. Besides, we measure the similarity of sequences based on the significance of Q-gram instead of the expensive sequence alignment. Thus, our search algorithm can run faster than the famous Blast algorithm. Following the idea of QgramSearch, a Q-gram based genomic sequence clustering (QgramClust) is proposed. In view of the challenge of expensive pairwise sequence comparison for large database sequences faced by the existing clustering algorithms, QgramClust employs the inverted index of Q-gram in sequence comparison so that the clustering process can be made efficient. Our clustering algorithm is a hybrid of partitioning method and hierarchical method. It quickly clusters a group of nearest neighbors and finally merges the clusters. Our experimental results show that QgramClust runs faster than BlastClust.

Files in this item

Files Size Format
b21657580.pdf 1.669Mb PDF
Copyright Undertaking
As a bona fide Library user, I declare that:
  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.


Quick Search


More Information