Fig. 5: PCAWG Central Nervous System meta-cohort donor-driver events in PCAWG-Scout. | Nature Communications

Fig. 5: PCAWG Central Nervous System meta-cohort donor-driver events in PCAWG-Scout.

From: A user guide for the online exploration and visualization of PCAWG data

Fig. 5

This Cytoscape-based visualization, available from the Study report, shows donors as hexagons and genes as circles. The PCAWG (Central Nervous System) CNS meta-cohort consists of the samples from cohorts identified by the codes CNS-GBM, CNS-Medullo, CNS-Oligo, CNS-PiloAstro. Edges represent driver events that either were validated or were only predicted by the PCAWG Drivers Working Group; validated driver events are colored green and predicted driver events are colored orange. Most of the driver events for IDH1 are shown as only predicted, whereas all of those for PTCH1 have been validated. The hexagon border thickness for each donor corresponds to the reported survival time. The size of each gene circle is proportional to the extent to which mutations found in the cohort for that gene have damage scores (MetaLR_score from DbNSFP v3.2a) higher than the scores for all possible SNVs that can happen over that gene (one-sided t test; the sample size varies from gene to gene). Genes IDH1, TP53, and DDX3X stand out as being more damaged than expected by chance. The circle color for each gene corresponds to its differential expression as calculated by PCAWG-Scout (FDR-corrected two-sided t statistic) when comparing IDH1-mutant samples with IDH1-wild type tumor samples; purple, light gray, and gold denote underexpression, no significant differential, and overexpression, respectively. The upper and lower bounds of the coloring gradient are defined on the basis of the entire genome, not just for the genes represented in the graph. IDH1 is among the most under-expressed, along with FN1 and CHEK2. FN1 and CHEK2 have driver events that co-occur in some samples with those for IDH1. PTCH1 is overexpressed in IDH1 mutants. Graphical esthetics of node border width, node size, node color, and edge color are configurable interactively. Details of the analysis and instructions for reproducing this visualization are available in Supplementary Note 3.

Back to article page