Determining plasmids from short read sequences

Abstract

In this thesis we present a novel method to determine the DNA-sequence of plasmids from a De Bruijn Graph. The method uses coverage data as well as estimates, based on the program mlplasmids, how likely a certain contig is plasmidal or chromosomal. The algorithm first thins the Bruijn Graph by removing all edges which are unlikely to be plasmidal. Next, all simple circular paths are determined using Johnson’s algorithm. Finally a Markov Chain Monte Carlo method is used to determine the circular paths which fit the observed coverage data well. The method discussed in this thesis could identify some small plasmids in a few real De Bruijn graphs of Enterococcus faecium and we have several suggestions to modify the method to find larger plasmids as well.