Theoretical and empirical analysis of whole genome sequence data for routine surveillance of bacterial pathogens: focus on antibiotic resistance
Dr Brooks Pollock
Applications accepted all year round
Self-Funded PhD Students Only
Whole genome sequencing (WGS) has been used successfully to understand outbreaks of infection in hospitals e.g. MRSA and to reconstruct likely transmission routes e.g. Tuberculosis. Recently, the cost has reduced to the extent that routine sequencing of samples is now a feasible alternative to traditional lab methods of bacterial detection and characterization such as microscopy, culture, antibiotic resistance phenotyping and DNA-based methods such as multi-locus sequence typing (MLST). However there remain barriers to full implementation of whole genome sequencing of clinical samples from community acquired pathogens which make up the vast majority of infectious disease acquisition. Many novel methods for analysis of genomic data have been developed (e.g. Bayesian methods for analysis of genomic data) but these have not been validated and tested to demonstrate added value or cost-effectiveness in comparison with existing laboratory and surveillance systems in the UK.
Aims & Objectives
This PhD seeks to address these gaps through validation and testing of existing analytical methods for whole genome sequence analysis for application to routine surveillance of community acquired pathogens.
The student will need to flexibly combine computational simulation and empirical data analysis to compare existing strategies for understanding changes in species composition over time. We will develop microsimulations to derive theoretical populations of bacteria, transmitted on different contact networks and with different evolutionary strategies. We will use these simulated datasets to compare the performance of cutting-edge Bayesian statistical analytical tools to detect epidemiological changes over time and then apply these methods to datasets of WGS data from published and ongoing studies. We will use the datasets to compare different methods for selecting samples for routine whole genome sequencing and inform the optimal method, taking into account cost-considerations, for detecting population level changes in strain types. The initial focus will be on infections where antimicrobial resistance is a problem e.g. gonorrhoea, but the methods developed will be widely applicable.