The process of short read generation for de novo Genome Assembly results in a loss of information about the position each read originates from in the source genome. State-of-the-art assembly algorithms typically process reads as an un-ordered set, attempting to reconstruct ordering information relatively late in the assembly process. This thesis proposes an alternative framework where an ordering is recovered prior to the start of assembly. The ordered reads are then divided into blocks, where each block may be assembled as an independent process. The ordering is produced by constructing a graph model of read affinities, which is used as an input to graph vertex ordering algorithms. The efficacy of several graph ordering algorithms for use in this framework is compared, and the effects of a “perfect” ordering of reads for use in a distributed assembly approach is demonstrated.
Metrics
19 File views/ downloads
22 Record Views
Details
Title
An Evaluation of Read Ordering for Genome Assembly
Creators
Patrick Bryan Keppler
Contributors
Ananth Kalyanaraman (Advisor)
Sriram Krishnamoorthy (Advisor)
Dingwen Tao (Committee Member)
Aravind Sukumaran Rajam (Committee Member)
Awarding Institution
Washington State University
Academic Unit
Electrical Engineering and Computer Science, School of
Theses and Dissertations
Master of Science (MS), Washington State University