Fig. 6
Interplay between copy number level ITH, 5mC, and clinical parameters. A Scatterplots showing the correlation between CNV counts and global 5mC, and CNV counts and entropy, for ccRCC regions. Each dot represents a tumor region and is colored according to the SETD2 status of the tumor of origin. The barplot at the top of the panel shows that regions of SETD2 mt tumors have a significantly smaller number of CNVs. The side panels show a significant global hypermethylation in SETD2 mt tumors and a significantly higher entropy in SETD2 wt tumors. B Oncoprint showing presence and absence of CNVs associated with ccRCC as per Gerlinger et al. [1] and Gulati et al. 44] in our cohort of 12 tumors. Gross tumor characteristics (SETD2 mutational status, metastasis, and T stage) of all 12 tumors, and a region-specific result (H3K36me3 IHC status) are also shown. Tumors are grouped according to their SETD2 mutational status (wt to the left and mt to the right). P values are the outcome of χ2 comparison of each CNV between SETD2 wt and mt tumors. Some tumor regions could not be scored (‘NR’) and are left white. C Phylo(epi)genetic trees corresponding to the 5mC and CNV datasets. The 5mC tree is drawn using the standardized Euclidian distance measured for the methylation of 843,393 CpGs, and the CNV tree is drawn using the standardized Euclidian distance measured for the log2 signal intensities for 25,752 loci