Identifying Mycobacterium tuberculosis transmissions using whole-genome sequences

Tuberculosis has infected a third of the world's population (2 billion people) and is responsible for 1.7 million deaths a year.

We plan to take advantage of new rapid technology that will turn the total genetic code of greater than 4 million DNA molecules from each germ into a barcode. This will be enormously valuable to the detective work of investigating how germs are spreading from person to person. Once we can do that we'll benefit from being able to identify outbreaks of disease quicker so that earlier intervention can be arranged.

The same technique of looking at the germ's barcode can be used to answer wider scientific questions about its evolutionary success both nationally and internationally.

Aims:
I plan to use Whole Genome Sequencing (WGS) to undertake a genealogical investigation of M tuberculosis (MTB) isolates in a high incidence region of the UK in order to further refine clusters of disease as defined by the current gold standard 24-locus MIRU-VNTR typing. I have 3 aims:
(i) I will investigate the distribution of M tuberculosis genome-wide sequence diversity within active cases across space (isolates from different anatomical loci), time (sequential isolates from individuals with a persistent infection) and household outbreaks. This data will be used to estimate the range of times to the most recent common ancestor (MRCA) of these samples which are known to be related (akin to "transmission").
(ii) Using results from (i) to provide thresholds for determining whether recent transmission is plausible, I will estimate most likely times to the MRCA within groups defined by 24-locus MIRU-VNTR, and use the 95% credible intervals from Bayesian phylogenetic methods to subdivide groups into clusters credibly representing recent transmission of tuberculosis.
(iii) I will then compare these genealogically identified clusters with those identified by the public health team using standardised epidemiological and demographic questionnaires.

Methods:
(i) I will sequence further samples and use Bayesian statistics to estimate the range of time to the MRCA
(ii) I will analyse 4-5 years' worth of whole-genome sequenced samples (~ 2000 samples) from central Birmingham, and a further 250 samples from the MIRU-VNTR groups across the Midlands, using classical statistics.

Opportunities:
(a) Translational: This study will provide a pathway for the implementation of these techniques into routine service.
(b) Scientific: Robust genealogical analyses will provide a technique to better investigate the incubation period and transmission dynamics (including the possibility of super-spreaders), and provide a foundation for model base transmission analyses.

The range of beneficiaries from this research project will be wide.

1. Decision making bodies:
(a) In the future, it may yield valuable information about the burden of on-going transmission in any region world-wide.
Time: 5-10 years
(b) At a national level the results will be of interest to policy makers in the Department of Health and HPA (now re-named and incorporated into the DoH) as they will make decisions about resource allocation to public health
Time scale: 5 years
(c) At a regional level the health protection units are likely in the future to start using WGS data rather than MIRU-VNTR typing to map transmission
Time scale: 3-5 years

2. Practitioners
(a) TB cluster investigators will be able to guide their investigations on the basis of links drawn by the WGS data
Time scale: 3-5 years
(b) Clinicians' decisions in the future may be benefited from a wider application of WGS not only to transmission networks but to resistotyping.
Time scale: 5-10 years

3. Academics
(a) This will be of value to TB epidemiologists investigating patterns (e.g. incubation period) and routes of transmission
Time scale: Immediately
(b) These techniques could be turned to investigating the association between genotypes and phenotypes such as hyper-transmissibility or hyper-virulence.
Time scale: 5-10 years

4. Patients
(a) New cases will be prevented if transmission is interrupted
Time scale: 3-5 years
(b) When informing patients about how they got tuberculosis this will give reassuring levels of accuracy
Time scale: 3-5 years