Recently, attention has focused on the possibility that substantial portions of the unmeasured heritability of common conditions lie in uncommon variations in the human genome.32 The 1000 Genome Project, initiated by the NIH in 2008, is a systematic effort to catalog the spectrum of genetic variation that occurs with at least a 1% frequency in humans. 33,34,35 The 1000 Genome Project data suggest that substantial disease related variation is present at frequencies below 1%. For example, a study that sequenced a large portion of the protein coding regions of the genome with a high degree of accuracy found more than 500,000 SNPs, of which about 82% were novel and 86% were present in less than 0.5% of the study population.36 Further, of the 2% of variants predicted to change protein function, over 95% were rare. These observations are sobering, since they imply that, if SNPs’ effect sizes are small, reliably detecting associations between such SNPs and health conditions might require enormous populations.
