Fig. 3

a Distribution of inter-cluster ensemble distances from cluster 6062, which contains the most frequent sequence. Clusters selected for comparison included clusters with >100 distinct sequences, >90% mean identity to the seed. To get a distance distribution when comparing clusters to cluster 6062, individual sequences of the same length from the given cluster and cluster 6062 were compared in an all-against-all fashion. As a reference, the median intra-cluster distance for cluster 6062 was 0.0898 (black line) and the first-quartile was 0.0536 (red line). b Representing all selected cluster pair-wise comparisons distance distributions in a heatmap shows that on average, clusters differ from other clusters by 0.2. In general, many of the structures are distinct from those of other cluster structures