An inability to standardize the bioinformatic data produced by wholegenome sequencing (WGS) has been a barrier to its widespread use in tuberculosis phylogenetics. The aim of this study was to carry out a phylogenetic analysis of tuberculosis in Wales, United Kingdom, using Ridom SeqSphere software for core genome multilocus sequence typing (cgMLST) analysis of whole-genome sequencing data. The phylogenetics of tuberculosis in Wales have not previously been studied. Sixty-six Mycobacterium tuberculosis isolates (including 42 outbreak-associated isolates) from south Wales were sequenced using an Illumina platform. Isolates were assigned to principal genetic groups, single nucleotide polymorphism (SNP) cluster groups, lineages, and sublineages using SNP-calling protocols. WGS data were submitted to the Ridom SeqSphere software for cgMLST analysis and analyzed alongside 179 previously lineage-defined isolates. The data set was dominated by the Euro-American lineage, with the sublineage composition being dominated by T, X, and Haarlem family strains. The cgMLST analysis successfully assigned 58 isolates to major lineages, and the results were consistent with those obtained by traditional SNP mapping methods. In addition, the cgMLST scheme was used to resolve an outbreak of tuberculosis occurring in the region. This study supports the use of a cgMLST method for standardized phylogenetic assignment of tuberculosis isolates and for outbreak resolution and provides the first insight into Welsh tuberculosis phylogenetics, identifying the presence of the Haarlem sublineage commonly associated with virulent traits.

This study describes the analysis of DNA from heat-killed (boilate) isolates of Mycobacterium tuberculosis from two UK outbreaks where DNA was of sub-optimal quality for the standard methodologies routinely used in microbial genomics. An Illumina library construction method developed for sequencing ancient DNA was successfully used to obtain whole genome sequences, allowing analysis of the outbreak by gene-by-gene MLST, SNP mapping and phylogenetic analysis. All cases were spoligotyped to the same Haarlem H1 sub-lineage. This is the first described application of ancient DNA library construction protocols to allow whole genome sequencing of a clinical tuberculosis outbreak. Using this method it is possible to obtain epidemiologically meaningful data even when DNA is of insufficient quality for standard methods.

Mycobacterium tuberculosis is the leading cause of death from an infectious disease worldwide. An understanding of tuberculosis transmission dynamics in outbreak settings is vital for its control. The advent of affordable whole genome sequencing (WGS) has provided scope for superior resolution of tuberculosis outbreaks, compared to previous methods. However, the challenge lies in standardising the vast quantities of resulting data in a structured manner which lends itself to easy comparison of isolates. Gene-by-gene Multi-Locus Sequence Typing (MLST) methods of analysing WGS data, as opposed to Single Nucleotide Polymorphism (SNP) mapping, have shown promise in providing a uniform platform for outbreak resolution. WGS was performed on clinical isolates from three M. tuberculosis outbreaks in South West Wales. Molecular typing by MIRU-VNTR and epidemiological investigation had resulted in conflicting conclusions. Outbreak analysis and phylogenetic typing of all isolates was carried out using the WGS gene-by-gene MLST analysis method of core genome MLST (cgMLST) and traditional WGS SNP mapping. Where DNA quality was unsatisfactory, an ancient DNA library preparation was used successfully. Provean and BEAST software analysis provided physiological