Fig. 1: Combined whole-genome sequencing and DNA methylation pediatric brain tumor patient datasets to explore methylation differences involving germline SVs.

a Schematic of the study approach. We referred to the CBTN multi-omics datasets^20,33 to explore germline structural variation in pediatric brain tumor patients in relation to tumor DNA methylation patterns. Germline SV calls were generated from the blood normal sample (using whole genome sequencing, or WGS, data). Based on the corresponding tumor sample, the CBTN generated both DNA methylation profiles and gene expression profiles. We identified both CpG Islands (CGIs, left) and enhancers (right) with DNA methylation differences associated with nearby germline SV breakpoints, paying particular attention to genes with differential mRNA expression patterns inversely correlated with the methylation patterns. b In the CBTN cohort, combined DNA methylation array and germline SV data involved 1292 patients, broken down here by tumor histologic type. c For the 1292 patient blood samples, SV class distributions, as observed for all 2,554,847 germline SVs in the dataset (left), for the subset of germline SVs with breakpoint within 1 Mb of a CGI methylation probe (middle), and for the subset of germline SVs with breakpoints spanning a CGI midpoint (right). DEL, deletion; DUP, duplication; INV, inversion (SV types as called by Manta algorithm⁶⁷).