Table 1.
List of epidemiologic clusters and their size.
Table 2.
Clustering performance for each stringency (s) value.
Table 3.
Impact of stringency setting on the partition (k) membership of genotypes linked to various epidemiologic clusters.
Fig 1.
Impact of minimum stringency (s) value on genetic linking of epidemiologically-linked isolates.
The hierarchical tree generated from our distance matrix was dissected into the minimum value of k (k = 21) predicted using the 94% s-value. The outer circle of colored bars indicates the boundary between each partition predicted and the inner circle of colored bars represents the epidemiologic linkage of various isolates, where each bar color represents a distinct epidemiologic cluster (grey represents isolates of unknown epidemiologic linkage). Epidemiological clusters of interest are labeled in the colored boxes. (A) At k = 21, we observe that the labeled epidemiologic clusters on the top right of the tree all belong to a single genetic partition, indicating that different epidemiologic clusters are being unnecessarily grouped into a single genetic partition.
Fig 2.
Impact of optimal stringency (s) value on genetic linking of epidemiologically-linked isolates.
The hierarchical tree generated from our distance matrix was dissected into the optimal value of k (k = 30) predicted using the 96.5% and 96.75% s-value. The outer circle of colored bars indicates the boundary between each partition predicted and the inner circle of colored bars represents the epidemiologic linkage of various isolates, where each bar color represents a distinct epidemiologic cluster (grey represents isolates of unknown epidemiologic linkage). Epidemiological clusters of interest are labeled in the colored boxes. (A) At k = 30, we observe that the labeled epidemiologic clusters on the top right of the tree are split between three different genetic partitions, while the two epidemiologic clusters on the bottom of the tree have 100% of isolates belonging to a single genetic partition.
Fig 3.
Impact of maximum stringency (s) value on genetic linking of epidemiologically-linked isolates.
The hierarchical tree generated from our distance matrix was dissected into the optimal value of k (k = 67) predicted using the 99.5% s-value. The outer circle of colored bars indicates the boundary between each partition predicted and the inner circle of colored bars represents the epidemiologic linkage of various isolates, where each bar color represents a distinct epidemiologic cluster (grey represents isolates of unknown epidemiologic linkage). Epidemiological clusters of interest are labeled in the colored boxes. (A) At k = 99.5%, we observe that the labeled epidemiologic clusters on the bottom of the tree have isolates split across multiple genetic partitions, suggesting that the maximum stringency value is unnecessarily splitting epidemiologic clusters between partitions.