We develop a method to precisely identify regulatory QTLs that operate in a context-specific manner. We apply our method to chromatin and expression data from kidney cancer tumor/normal samples and identify wide-spread context-specific QTLs that appear to be driven by germline-somatic interactions. We connect these QTLs to GWAS loci and kidney cancer risk heritability, and we validate one locus using CRISPRi. We believe this is a powerful tool to identify important regulatory variants in many contexts.
stratAS and the analysis pipeline are available in the stratAS repository.