Before genome sequencing became routine, it was widely assumed that trait-relevant variation would reside mostly or exclusively within genes. Results from GWAS have uprooted this assumption. Many of the genomic regions found to be associated with disease are in noncoding areas—parts of the genome that do not encode protein, as reviewed by J. D. French and S. L. Edwards in “The Role of Noncoding Variants in Heritable Disease.” This phenomenon of noncoding variation has led to new challenges. If GWAS discovers a genetic change within a part of the genome that does encode a gene, then it is reasonable to posit that the mechanism of the phenotype being studied must be related to the function of that protein. An example was identified years before the first true GWAS project by Paloma Valverde and colleagues in their Nature Genetics publication “Variants of the Melanocyte-Stimulating Hormone Receptor Gene Are Associated with Red Hair and Fair Skin in Humans.” Red hair and variation in response to sun exposure in humans are both associated with changes in the MC1R gene, which encodes a protein that acts as a receptor for a hormone that regulates melanin pigmentation. For trait-associated genetic changes that do not reside within genes, though, the mechanism of variant impact—and, in particular, which genes may have their level of expression regulated by the associated non-coding variant—is much more difficult to discern. François Spitz reviews studies that seek to understand how well such distant relationships can predict the impact of a non-coding variant on the regulation of a given gene in “Gene Regulation at a Distance,” while Natali Papanicolaou and Alessandro Bonetti survey approaches for identifying the mechanisms whereby genetic changes in coding regions modulate disease phenotypes in “The New Frontier of Functional Genomics: From Chromatin Architecture and Noncoding RNAs to Therapeutic Targets.”