Identifying the founder of a group and the patterns of evolutionary descent

The following description of eBURST focuses on groups identified using the stringent default setting, where all STs must be a single locus variant (SLV) of at least one other ST in the group, and eBURST produce a diagram where all STs in the group are linked (a clonal complex). The use of eBURST with the more relaxed group definition of 5/7 shared alleles is discussed later.

The primary founder of a group is defined as the ST that differs from the largest number of other STs at only a single locus (i.e. the ST that has the greatest number of single-locus variants; SLVs). This method of assigning the primary founder takes account of the way in which clones emerge and diversify (Figure 1A); the initial diversification of the founding genotype of a clonal complex will result in variants of the founder that differ at only one of the seven loci (i.e., SLVs of the founder). If two STs have the same number of associated SLVs, the one with the greatest number of double-locus variants (DLVs) is selected as the founding ST. In such cases the confidence in the assignment of the primary founder will be low, as reflected in the bootstrap values (see also here).

Note that the assignment of the founding ST does not take into account the number of isolates of each ST; this makes the procedure relatively robust to sampling bias. In many cases, however, the predicted primary founder will also be the most prevalent ST within the group. Together with strong bootstrap support, this can provide added confidence in the assignment of the primary founder. The average distance of each ST to all other STs in the eBURST group is also calculated by eBURST v3 and the primary founder is likely to be the ST with the minimum average distance to all other STs.

The eBURST diagrams display the patterns of descent within each group from the predicted founding ST (Figure 1B). The initial version of BURST placed the founder centrally and positioned SLVs and DLVs of the founder within concentric rings, whereas eBURST uses lines to show the radial links from the founder to each of its SLVs. A second difference is that only SLV links are shown in eBURST, and DLVs of the founder will therefore only be linked if the intermediate SLV on the path from the founder to the DLV is present in the input data. All isolates that are directly linked on a eBURST diagram will therefore differ at only a single locus and the cluster of linked STs should represent a clonal complex. A DLV of the predicted primary founder of a group will not be included if the linking SLV has not been sampled and using the stringent (conservative) default group definition some STs that might be part of the clonal complex will therefore not be shown. In the eBURST diagram, the circle representing the predicted primary founder is coloured blue and the areas of each of the circles indicate the prevalence of the ST in the input data (Figure 1B).

With the less stringent group definition of 5/7 shared alleles, more than one cluster of linked STs (clonal complex) may be displayed in the eBURST diagram, along with a number of unlinked individual STs. The lack of linking between two clusters within a single eBURST group implies that no ST within one cluster is a SLV of any ST in the other cluster(s). Similarly, unlinked individual STs will not be SLVs of any ST in the group. The SLVs and DLVs of any ST can be displayed on the eBURST diagram (see also here).

Subgroups and subgroup founders

In larger eBURST groups there may be several STs besides the predicted primary founder that have a number of SLVs of their own. For example, a SLV of the primary founder may have become successful and diversified to produce its own SLVs. A ST that appears to have diversified to produce multiple SLVs is called a subgroup founder. In the default setting for the eBURST diagram, a ST with at least two assigned descendent SLVs (i.e. three SLVs in total, as the link from the ST to its putative progenitor is also a SLV) is defined as a subgroup founder. In larger eBURST groups there may be many STs with at least two assigned descendent SLVs, and the number of descendent SLVs that define whether a ST is shown as a subgroup founder should be increased. The STs that are subgroup founders, according to the default definition of a subgroup, or a user-defined definition, are coloured yellow on the eBURST diagram (see examples below). The primary founder of the group is coloured blue.