Fig. 6

Left: a summary of 100 duplication episodes found in Génolevures dataset [49] by Algorithm 5 for the species trees from [54]. Right: 91 duplication episodes found in the species tree from [55]. D2 and D2^∗ denote one whole genome duplication (WGD) event suggested in [58, 59], while D1 and D1^∗ denote one WGD event from [57]. The number of episodes assigned to a single edge is presented on the side (blue italic), for example, our algorithm found 13 duplication episodes in the rooting edge in both trees. A gray histogram (the right side of a node) denotes the frequency of gene trees (Y axis) being involved into exactly x (X-axis starting from 1) episodes located on the corresponding node. The number above the highest bar denotes the maximal number of such gene trees. For example, the gray histogram in the left tree with the second bar of the size 960 denotes that there are 960 gene trees contributing to exactly 2 episodes at the current node. Bars of frequency lower that 10 are not shown. A red bar on the left of a node denotes the number of gene trees having at least one duplication event mapped to this node, i.e., the sum of bars of the corresponding gray histogram