Fig. 10
From: A multi-labeled tree dissimilarity measure for comparing “clonal trees” of tumor progression
Trees inferred by SCITE and SiFit for acute lymphoblastic leukemia (ALL) patient dataset from [38] consisting of 115 single cells and 16 mutations. Unsurprisingly, due to large number of single-cells in this dataset, sequencing noise and similarities in the scoring schemes used in PhISCS and SCITE (see Appendix 1: Details of obtaining trees of tumor evolution for the real data sets) both methods report the same mutation tree so we only focus on SCITE in this discussion. The most notable difference among the two trees is in the placement and ordering of mutations in genes ZC3H3, XPO7 and BRD7P3 as well as in the ordering of mutations in genes FGD, RRP8, FAM105A, BDNF-AS and PCDH7. Furthermore, the relative order also differs for mutations in genes TRRAP and ATRNL1. However, in contrast to these important differences, the trees still share most of the major branching events in tumor evolution and have consistent ancestor–descendant order for most of the pairs of mutations. All these are reflected in MLTD-normalized score of 0.69 assigned to this pair of trees