In a typical GWAS analysis, one might test ~2.5 million common SNPs in the genome, to see if any of these genetic variants are associated with a trait, such as a brain-derived measure, or a specific disease such as AD. Although not the only important type of genetic variation, SNPs can be measured using readily available genotyping arrays, and individually provide adequate statistical power as the variants are common enough to test their effects statistically. Because only selected SNPs capturing the variability of the genome are genotyped, many authors have argued that this genotyping technology is much less expensive than whole-genome sequencing. However, new technologies using low coverage sequencing with imputation may in some cases yield several times the effective sample size of GWAS based on SNP array data, and a commensurate increase in statistical power as described in Pasaniuc et al. (2012).
