Extended Data Fig. 3: Biological and clinical correlates of mutational patterns.
From: Clonal evolution during metastatic spread in high-risk neuroblastoma
(a) Scatter plots show the association between exposure to SBS18 (left) or SBS40 (right) and age at diagnosis amongst the diagnostic/t-resection tumors (n = 132) (Pearson correlation). (b) Box plot shows the mean expression of the genes in glutaminolysis signature associated with ROS accumulation22 across diagnostic tumors (n = 59) of different disease subtypes (left) and MYCN-A tumors (n = 56) from diagnosis, t-resection and relapse and further relapses (right). Median, upper and lower quartiles as well as comparisons with significant p-values from a two-sided Wilcoxon test are shown according to the significance levels in the legend. (c) For n = 151 tumors, scatterplots (left) shows correlation between number of SNVs and the number of predicted neoantigens, and (right) shows the relationship between the number of predicted neoantigens and immune infiltrates in the surrounding tumor microenvironment as assessed from RNAseq (Pearson correlation). (d) Barplots show the proportion of genome-wide SNVs for n = 45 tumors (left) and oncogenic driver SNVs for n = 54 tumors (right) attributed to different mutational signatures broken down by presence in post-therapy relapse tumors. For all scatterplots Pearson correlation and associated p-value with a blue linear regression line and 95% confidence interval in grey is shown. The data and script for the figure are available in Supplementary Table 1 and the GitHub repository.
