Assessing the Gene Regulatory Landscape in 1,188 Human Tumors

Abstract

Cancer is characterised by somatic genetic variation, but the effect of the majority of non-coding somatic variants and the interface with the germline genome are still unknown. We analysed the whole genome and RNA-Seq data from 1,188 human cancer patients as provided by the Pan-cancer Analysis of Whole Genomes (PCAWG) project to map cis expression quantitative trait loci of somatic and germline variation and to uncover the causes of allele-specific expression patterns in human cancers. The availability of the first large-scale dataset with both whole genome and gene expression data enabled us to uncover the effects of the non-coding variation on cancer. In addition to confirming known regulatory effects, we identified novel associations between somatic variation and expression dysregulation, in particular in distal regulatory elements. Finally, we uncovered links between somatic mutational signatures and gene expression changes, including TERT and LMO2 , and we explained the inherited risk factors in APOBEC-related mutational processes. This work represents the first large-scale assessment of the effects of both germline and somatic genetic variation on gene expression in cancer and creates a valuable resource cataloguing these effects.

A Framework for Transcriptome-Wide Association Studies in Breast Cancer in Diverse Study Populations

“…We lastly sought to evaluate the source of the significant eQTLs we detect in CBCS. Similar to previous pan-cancer germline eQTL analyses [25], we cross-referenced eGenes found in CBCS with eGenes detected in relevant healthy tissues from Genotype-Tissue Expression (GTEx) Project. We attributed all but 7 of the cis-eGenes from CBCS across both AA and WW women found in GTEx to one of these three tissue types (Figure 1B), with the effect sizes of the top eQTLs for these eGenes correlating very well between CBCS and GTEx (see Supplemental Figure 5 ).…”

Allelic imbalance reveals widespread germline-somatic regulatory differences and prioritizes risk loci in Renal Cell Carcinoma

“…We re-ran stratAS on each H3K27ac sample independently, restricted to peaks with high power to detect imbalance (having a total >50 reads across all heterozygous sites in the peak). We then built a logistic regression across all individual-peak pairs, with the presence/absence of a significant T-/d-asQTL as the outcome and various local features as predictors (mirroring similar analyses of allele-specific expression in previous work 22 ). We found that the presence of an SCNA loss was the feature most predictive of a T-asQTL (OR of 3.5, explaining 9.6% of variance) followed by the presence of an N-asQTL in the matched normal (OR of 5.24, but explaining 2.9% of variance) ( Supplementary Table S2 ).…”

“…The traditional approach of inferring hidden confounders using principal component or factor analysis 19 typically separates all tumor/normal samples and would thus result in an undefined or unstable interaction test. The allelic imbalance model has been applied extensively in studies of normal expression, but no formal interaction test has been proposed, and the previous applications to cancer have not modelled tumor-specific structural variation that can introduce biases 20–22 .…”

DeCompress: tissue compartment deconvolution of targeted mRNA expression panels using compressed sensing

“…Similarly, EA tumor-specific eGenes showed enrichments for cell death and proliferation ontologies, and immune-specific eGenes showed cytokine and lymph vessel-associated processes. We then cross-referenced bulk and tumor-specific cis -eGenes found in the CBCS EA sample with cis -eGenes detected in healthy tissues from GTEx: mammary tissue, fibroblasts, lymphocytes, and adipose (see Materials and Methods), similar to previous pan-cancer germline eQTL analyses ( 10 , 110 ). We attributed several of the bulk cis -eGenes to healthy GTEx tissue (all but 2), but tumor-specific cis -eGenes were less enriched in healthy tissues ( Supplemental Figure S21 ).…”