Fig. 3
From: Getting insight into the pan-genome structure with PangTree
Distributions of the value of sequence compatibility with consensuses C1 (top), C2 (middle) and C3 (bottom), computed in consecutive iterations. Vertical green lines indicate resulting compatibility thresholds, i.e. right ends of the largest gaps between individual compatibility values (indicated by short vertical blue lines). C1 is calculated from all input sequences, while C2 and C3 – only from sequences having in the previous iteration compatibility above the threshold. The set of sequences exceeding the compatibility threshold has changed in the second iteration, since a more homogeneous subgroup was recognized with C2. The third iteration doesn’t change the respective set and the procedure converges. However, consensus C3 represents selected sequences better than C2, which is reflected by their larger average compatibility