Extended Data Fig. 3: Examples of early evolution.
From: Neuroblastoma arises in early fetal development and its evolutionary duration predicts outcome
a, Copy number profiles of an example tumor with near-triploid genome and without a timeable ECA. The color panels annotate genomic segments of equal copy number. Next to the copy number profile, the densities of non-amplified (green) and amplified (color-encoded) clonal mutations on segment of length ≥107 bp are shown, with the dashed line showing the kernel-density estimate of the distribution of non-amplified clonal mutations. Chromosomal segments of equal copy number were merged and Holm-corrected one-sided P values were computed based on a negative binomial distribution (**, padj < 0.01, exact P values are provided in Source Data). The bottom panel shows mutation densities at ECA and MRCA with 95% confidence bounds estimated by bootstrapping. The horizontal lines show mutation densities on gained segments with 95% confidence bounds computed from χ2 distributions. b, Mutation densities at ECA and MRCA of early-MRCA tumors with (left) and without (right) timeable ECA. Solid lines represent maximum likelihood estimates and shaded areas represent 95% confidence intervals, as obtained by bootstrapping. c and d, As in (a) but for a near-diploid tumor with timeable ECA (c) and for a near-tetraploid tumor without timeable ECA (d). Holm-corrected one-sided P values were computed based on a negative binomial distribution (**, padj < 0.01, exact P values are provided in Source Data). e, Mutation densities at MRCA in early-MRCA (n = 20) and late-MRCA primary tumors (n = 47, thereof n = 26 with ECA and n = 21 without ECA). f, Fraction of polysomic chromosomes in near-triploid (n = 33) and near-tetraploid (n = 12) tumors that were generated in a single oncogenic event. As a control, the fraction of disomic chromosome whose mutation density matched with the MRCA is shown for near-diploid tumors (n = 55). In (e) and (f), boxes show median, 25 and 75% percentiles, whiskers extend to the smallest and largest value within 1.5x interquartile range.
