Fig. 2: Study design.

a Study design flow chart; b overview of the integrative-omics analysis. (1) Genetic prediction model building for blood DNA methylation levels; (2) associations of genetically predicted DNA methylation in blood and prostate cancer risk; (3) expression quantitative trait methylation; (4) genetic prediction models for blood and prostate tissue gene expression levels; (5) associations of genetically predicted gene expression in blood and prostate tissue with prostate cancer risk. Results in 1 were based on data of the Framingham Heart Study (FHS) (N = 1595). Results in 2 and 5 were based on the summary statistics of the PRACTICAL, CRUK, CAPS, BPC3, and PEGASUS consortia (N = 79,194 cases and 61,112 controls). Results in 3 were based on data of the FHS (N = 1367) and The Cancer Genome Atlas (N = 34). Results in 4 were based on data of the Genotype-Tissue Expression project (version 8).