Genome-wide association studies (GWAS) have been a useful tool in identifying numerous genetic loci that associate with increased risk for numerous cancers. However, as most of the identified risk variants are found in non-coding regions of the genome, the field has been slow in moving beyond identifying the risk variants to functionally determining the mechanism for cancer predisposition. The ultimate goal of genome-wide association studies is to identify and understand the role of these loci disease etiology to ultimately enable more effective screening and therapeutic treatments. Framed by better understanding GWAS results, my dissertation has 3 main aspects, where I: 1) developed a computational approach to characterize GWAS results using publically-available epigenomic databases, 2) identified novel germline susceptibility loci for myeloproliferative neoplasms, and 3) examined molecular mechanisms by which a prostate cancer single-nucleotide polymorphism may increase risk.