Fig. 3: Structural properties of the generated phylogenetic trees.

Violin plots of a the number of branches, b total branch lengths, and c total stem lengths for neuroblastoma (NB), Wilms tumor (WT), and rhabdomyosarcoma (RMS) using either the maximum likelihood (ML) or parsimony (MP) method for phylogenetic reconstruction. Significance represented by P-values were calculated using the two-sided Mann–Whitney U-test. The box plots within the violin plots illustrates the interquartile range. The red dot is the median value. d In 14% of neuroblastomas, 15% of Wilms tumors, and 50% of the 8 rhabdomyosarcomas analyzed, the phylogenetic trees obtained using the ML and MP methods differed from one another. NB1, NB24, WT1, WT2, WT3, and WT5 are not included in this calculation since they did not display any private genetic alterations. Hence only event matrices could be generated but not phylogenetic trees. If including these cases, the proportions would be 12.8% for NB and 12.5% for WT. e Venn diagrams of how often discrepancies between ML/MP trees, back mutations, and parallel evolution occurred together in the same case/tree. Numbers indicate the number of tumors where a particular contradiction or combination of contradictions occurred. The information used to produce the plots can be found in Supplementary Data 1 and 2.