Whole-genome sequence analysis of Mycobacteria tuberculosis bacteria to identify the genetic determinants and mechanisms of anti-tuberculosis drug

resistance.

Tuberculosis disease (TB) is a major public health problem. The emergence of Mycobacterium
tuberculosis strains resistant to the drugs used to treat the disease threatens to derail efforts to
control TB. Resistance to at least two of the first line drugs is denoted as multidrug-resistant
tuberculosis (MDR-TB), whilst additional resistance to any drug used to treat MDR-TB is known as
extensively drug-resistant tuberculosis (XDR-TB). The appearance of these strains has complicated
clinical drug selection, which along with the considerable toxicity and side effects of the drugs result
in poor outcomes, less compliance and ultimately can amplify resistance. For this reason, early
detection of resistant strains is crucial. Susceptibility tests of clinical samples have been traditionally
carried out by phenotypic methods, which are costly and can take weeks since they involve culture
and manipulation of highly infectious bacteria. For some first-line drugs there are also molecularbased
tests, but most have low sensitivity.
Drug resistance is conferred almost exclusively by accumulation of point mutations and insertions
and deletions in genes encoding targets or drug-activating enzymes. The study of genotypephenotype
associations has enabled the creation of a mutation library for drug resistance markers.
Therefore, the use of whole-genome sequence analysis of clinical samples in combination with a
mutation library can inform clinical drug management, in a more timely fashion than phenotypic
methods. Nevertheless, one of the challenges is implement the direct analysis of sputum samples,
for which culture of the bacteria prior to sequence is still necessary. A method known as selective
whole-genome amplification (SWGA) is a potential solution, and enables the amplification of the M.
tuberculosis target genome from a complex sample by using specific primers that do not anneal to
non- M. tuberculosis meta genomes in the sputum.
On this basis, the aims of this project are: (i) to identify novel resistance mutations and mechanisms
of anti-tuberculosis drug resistance from whole-genome sequence data by applying regression-
2
based models; (ii) to characterise the effects of mutations on protein stability and drug docking by
using protein structure models; (iii) to develop quantitative methods to rapidly predict drug
resistance from whole genome sequence data. In order to accomplish these aims, strong
bioinformatics and big data analysis skills will be gained and applied in the field of genomics.
Interdisciplinary skills will be gained including training in culture of M. tuberculosis and development
of the SWGA approach. Finally, the results of this project could lead to new diagnostic tools, assist
with the management of the TB, and eventually benefit public health. The participation of St.
George's Hospital and University will assist with the roll-out in a clinical setting of any sequencing
or laboratory-based tools developed..