Finally, as GWAS typically focus on a single class of genetic variation—the common SNP contributions to disease risk—PRS typically will measure only the contribution of common single-nucleotide variants in an individual and not other classes of variation which may also impact genetic risk and be particularly relevant in certain individuals. For instance, an individual may carry one or more copy number variants which have a large impact on disease risk which will not be considered in a typical PRS. Also, rare pathogenic alleles are typically excluded from PRS derived from GWAS summary statistics, because GWAS typically include only “common” variants with a population frequency of 1% or more. Furthermore, PRS assumes an additive effect of individual risk alleles and does not model complex higher-order interactive (or epistatic) relationships between risk variants, simplifying the genetic model under which PRS are calculated. As our knowledge of the contributions of rare variants increases with large-scale sequencing studies 24, we may soon see the development of genomic risk predictions which encompass both common and rare single-nucleotide variants, structural variants, and even epigenetic factors and
