Fig. 2: Prevalence of cancer predisposition gene germline P/LP variants in pediatric CNS tumor patients.

A Number of patients with identified germline P/LP SNVs and InDels (N = 197 variants) by cancer predisposition gene (CPG) and tumor histology. B Number of CPG germline P/LP SNVs and InDels (N = 197) by call and call source. C Number of patients with identified P/LP structural variants (N = 18 variants) by gene, tumor histology, and type. DEL deletion, DUP duplication. D Number of patients harboring P/LP variants in syndrome-associated genes (N = 104), and clinical definition. RTPS Rhabdoid Tumor Predisposition Syndrome, FAP Familial Adenomatous Polyposis, MAS melanoma astrocytoma syndrome, BAP1 TPS BAP1 Tumor Predisposition Syndrome. E Percent of total patients (N = 830) harboring a CPG P/LP variant by tumor histology. F Odds ratios and associated p-values of CPG P/LP variant burden among pediatric CNS tumor cohort relative to PMBB and gnomAD cancer-free control cohorts. Odds ratios and associated p-values of P/LP variant burden in DNA repair genes from Knijnenburg et al.³³ among G entire cohort and H pediatric high-grade glioma cohort (HGG, excluding DIPG and DMG) relative to control cohorts. Odds ratios and other statistics were derived from two-sided Fisher’s exact tests with Bonferroni-adjusted multiple test correction. In odds ratio plots, points represent odds ratios or log10-transformed odds ratios, and error bars represent 95% confidence intervals. Dashed lines in p-value plots indicate Bonferroni-adjusted p < 0.05. N = 830 PBTA patients, N = 6295 PMBB participants, N = 74,023 gnomAD cohort participants, and N = 76 patients with HGG.