Global impact of germline structural variation on the cancer proteome

Proteome and transcriptome data combined can help assess the relevance of non-coding germline variants. Here, we combine germline Structural Variants (SVs) with mass spectrometry-based proteomics on tumors from 1637 cancer patients spanning various tumor tissues of origin to determine the extent SV breakpoint patterns involve differential protein expression of nearby genes. Rare and singleton SVs disrupting protein expression of known cancer susceptibility genes collectively involve 6% of patients. About 24% of the hundreds of genes with SV-associated non-coding cis-regulatory alterations at the mRNA level are similarly associated at the protein level. Both rare and common SVs may associate with differential protein expression within a specific tumor type or across multiple tissue types, including SVs differentially represented by patient ancestry. SVs involving altered methylation of CpG Islands or enhancers are also implicated in differential protein expression. Our results emphasize the contribution of germline SVs to cancer heterogeneity at the proteome level.