Fig. 6: Systematic analysis of complex variants revealed two potential loci where novel correlated complex variant could explain PCa risk.

a Potential functional correlated complex variant identification pipeline based on coverage-based analysis of TCGA samples. b GWAS Manhattan plot of the rs2273669 related PCa risk region on chromosome 6. c GWAS Manhattan plot of the rs9287719 related PCa risk region on chromosome 2. Genome-wide significant associations indicated by individual dots above P = 5 × 10−8. d GWAS data MNLP conditioning neutralizes the rs2273669 effects, demonstrating the MNLP14 genotypes power. e GWAS data MNLP conditioning neutralizes the rs9287719 effects, demonstrating the MNLP16 genotypes power. GWAS identified correlated SNPs explain PCa risk at genome-wide significance showing 2 different risk regions (b, c). Conditioning GWAS associations in the locus on the corresponding identified correlated complex variant in the risk region explained all significant GWAS associations (d, e), supporting the potential causal variant role of the correlated complex variant. P values indicated for different approaches (Supplementary Fig. S9j).