Fig. 2: Genes with altered expression associated with nearby SSV breakpoints.

a Schematic of the phenomenon of interest. SSV breakpoints outside of genes may result in altered cis-regulation. SSV breakpoints within genes may involve gene disruption or fusions. b For each of the indicated genomic region windows examined, numbers of significant genes (FDR < 10%) showing a correlation between expression and associated SSV event. Numbers above and below the zero point of the y-axis denote positively and negatively correlated genes, respectively. Linear regression models evaluated significant associations when correcting for tumor type (gray) and for both tumor type and gene-level CNA (black). For the 1 Mb region window, the model weights the relative gene distances of the breakpoints⁴. c Heat map of significance patterns for 627 genes significant for any region window (FDR < 10%, correcting for both tumor type and CNA). Red denotes significant positive correlation; blue, significant negative correlation. Genes listed are cancer-associated²⁰. d Venn diagrams representing the overlaps between the genes positively correlated (FDR < 10%) with SSV breakpoint in CTBBC pediatric brain cohort and the genes positively correlated (FDR < 10%) with SSV breakpoint in either TCGA-ICGC cohort (left, n = 2334 cases) or TCGA glioma cohort (right, n = 107 GBM/LGG cases). P values by one-sided Fisher’s exact test. e Significance of genes in TCGA glioma cohort (107 patients, x-axis), as compared to their significance in the CBTTC cohort (854 tumors, y-axis). Genes in the upper left quadrant reached significance only in the present study. f X-axis indicates the FDR in the most significant of 13 pediatric brain tumor types analyzed separately, and y-axis indicates the FDR when the 854 CBTTC tumors are analyzed as a combined cohort. Genes in the upper left quadrant reached significance only in the pan-cancer analysis. Genes in the lower right quadrant reached significance only in one or more single-type analyses. Color of data points represents the most significant tumor type. For d–f, significant genes are defined by 1 Mb region window, correcting for tumor type and CNA. For d, e, “cancer-related” is by COSMIC²⁰.