Therefore in order to obtain local support values for the branch split points the same data were used to produce an approximate ML tree with local support values using FastTree
2 [25]. This tree had almost identical topology to the RAxML tree and the majority of split points had local support values of > 0.8. The same sequence data used to generate the tree were clustered using three methodologies; eBurst, BAPS of allelic data and BAPS of sequence Torin 2 cost data (Figures 2, 3 and 4). Figure 2 Clusters as determined by eBURST mapped onto a Etomoxir in vitro radial phylogram generated by FastTree 2. STs not assigned to a cluster (singletons in eBURST) are coloured black. Figure 3 Clusters as determined by BAPS using allelic data mapped onto a radial phylogram generated by FastTree 2. Figure 4 Clusters as determined Batimastat by BAPS using linked sequence mapped
onto a radial phylogram generated by FastTree 2. STs that have significant admixture are coloured black. The clusters are labelled using the lowest ST number found within the cluster. eBurst analysis eBurst uses the BURST algorithm to identify mutually exclusive groups of related genotypes in the population, to identify the founding genotype of each group and to predict the descent from the predicted founding genotype to the other genotypes in the group [26]. The algorithm assumes that each allele is equally related to all other alleles of the same locus and as such assumes that recombination is a frequent event. eBurst clustering produced 55 groups, 31 of which contained just two STs, and 190 singletons. Bayesian Analysis of Population Aspartate Structure (BAPS) BAPS is a tool for the detection and representation of recombination between populations [27]. The BAPS mixture model is derived using novel Bayesian predictive classification theory, applied to the population genetics context. A variety of different prior assumptions about the data can be utilized in BAPS to
make inferences, however it does not require either a prior model of clonality versus recombination, or a pre-defined number of clusters. BAPS can be used to determine the population structure, to determine gene flow within a population, to determine the amount of admixture in an individual, and to divide the population into clusters [28, 29]. The data required for BAPS population analysis can be in several formats. The first analysis performed used allelelic data identical to that for the BURST analysis but saved in GENEPOP format. Those STs that had significant (p <0.05) admixture (genetic material from more than one genetic lineage) were not assigned to a cluster. With the maximum permissible number of clusters set at 20 clusters, the optimal partitioning of the 838 STs resolved them into 15 clusters with a mean number of STs of 55.9 and a standard deviation of 48.0. However 12 sequence types had significant admixture and were excluded from clusters. BAPS analysis was also performed using molecular sequence data.