The diversity of microbial communities, or microbiomes, and their importance to both human health, for example the gut, and the environment, has only become apparent through the development of improved DNA sequencing technologies. These are now cost effective enough to enable sequencing of an entire community of microbes, i.e. the metagenome. However, the challenge of entirely resolving the genomes of the individual species and strains present in a community from metagenomics sequence has yet to be solved. Recently, long read technologies, such as Oxford Nanopore enable far longer reads than before but these sequences are noisy. This project proposes to develop methods that draw on Bayesian graphical models to integrate these long noisy reads with the short accurate reads that are available from standard next generation sequencing, this is termed a ‘hybrid’ assembly approach.
We have already developed methods for representing the key structure in genome assembly, the assembly graph, as a Bayesian graphical model and demonstrated that it can be solved for short reads with methods such as variational inference. Currently, though, our results are limited to local regions of the assembly graph enabling strain resolution on conserved genes but not the reconstruction of entire genomes. This project would build on this to develop novel graphical models that can represent both long and short reads. This project will be supervised by Dr Christopher Quince (University of Warwick), Dr Xavier Didelot (University of Warwick), Dr Alistair Darby (University of Liverpool) and/or Dr Tim Dallman (Public Health England).
We invite applications from candidates who have a 2.1 or 1st class degree in a relevant biological science or quantitative science discipline. A Masters degree in a relevant discipline would be an advantage. Applicants should send a full CV (including the names and email addresses of at least two academic referees), and personal statement to [email protected]. This should state: • An outline of how this programme of research and training will benefit from their past experience and impact upon their career aspirations.
The position may be based at either, The University of Liverpool, The University of Warwick or PHE (Colindale/ Porton Down), depending on the project. For some projects, students will be required to work at one of the partner Universities’ and spend time at PHE.
Thesis studentships cover research costs and tuition fees at the UK/EU rate for students who meet the residency requirements only.