Computational Methods for De Novo Assembly of Next-Generation Sequencing Data

Rayan Chikhi

De novo assembly of genomes is challenging bioinformatics problem. It has
recently received a lot of attention due to the rapid development of
next-generation high-throughtput sequencing. Nevertheless, current
assemblers can only provide satisfying results for small genomes, and the
problem of high-quality mammalian-sized genome assembly remains open.
Therefore, new computational approaches are needed in order to solve large
instances (hundreds of millions of reads) in reasonable time, and new
statistics are needed to estimate lower and upper bounds on the quality of
results given realistic sequencing scenarios.
In this talk I will describe current methods for short reads assembly, and
give an overview of the underlying computational theory. Then I will
introduce my thesis work: how paired-end data can be used to improve
assembly both in terms of speed and results quality. A new experimental
paired-end assembler will be presented.