Access Type

Open Access Dissertation

Date of Award

January 2015

Degree Type


Degree Name



Computer Science

First Advisor

Dongxiao Zhu


Recent progress in DNA amplification techniques, particularly multiple displacement

amplification (MDA), has made it possible to sequence and assemble bacterial

genomes from a single cell. However, the quality of single cell genome assembly has

not yet reached the quality of normal multi-cell genome assembly due to the coverage

bias (including uneven depth of coverage and region blackout) and errors caused by

MDA. Computational methods try to mitigates the amplification bias. In this document

we introduce a de novo co-assembly method using colored de Bruijn graph,

which can overcome the problem of blackout regions due to amplification bias. The

algorithm is implemented in a tool named HyDA (Hybrid De novo Assembler). HyDA

can assemble various genome of phylogenetically close species simultaneously whit a

high quality while it can determine the genomic relationship between each pair of

co-assembled dataset based on their common and exclusive contigs. Moreover, co-assembly can provide a high quality genome assembly from a number of guessed to

be identical single cell. Since our algorithm can detect the outlier, it can generate a

non-chimeric result with most recovered genome regions. Various techniques, such as

iterative assembly with multiple k, are employed in HyDA, which makes it a powerful

de novo assembly tool.