SNPs uncover a "holey" human genome

When Dylan Thomas wrote, "The best craftsmanship always leaves holes and gaps," the human genome certainly was not his muse. But it is for the many scientists who now hold tangible proof of his declaration.

Scientists at the Broad Institute of MIT and Harvard reviewed a collection of human DNA samples in which they had failed to detect single nucleotide polymorphisms (SNPs) and, upon closer inspection, realized that these samples contained large holes in regions of the genetic blueprint. In an online report published December 4 in Nature Genetics, these researchers describe a method that relies on SNPs to delineate gaps, or "deletions," in human DNA. By applying this method to data from the International HapMap Consortium they found that deletions are much more frequent than once imagined, both in their distribution across the genome and in their prevalence in the human population.

The completion of the human genome sequence put genetic variation between individuals in the limelight, and SNPs (single letter changes in the genetic code) initially took center stage. Other forms of genetic diversity, however, may be equally important. For example, addresses in the genome are typically present in two iterations, but some individuals lack one copy, or even both. These deletions often remove critical genetic instructions and can serve as a stepping stone toward disease. Broad scientists, led by David Altshuler who directs the Program in Medical and Population Genetics, set out to systematically determine the extent of deletion variability among humans.

Previous studies identified deletions by comparing DNA sequences from a handful of individuals to the finished human genome sequence. Broad researchers wanted to ascertain the frequency of these known deletions in a broader population and, in addition, to determine if other unidentified deletions are both detectable and common. Such a task requires a detailed view of an individual genome as well as a wide-angle view that encompasses many different genomes. SNPs — which have been cataloged, or genotyped, in a few hundred people in Phase I of the HapMap Project — provided the analytical tool to perform this work.

The scientists combed through a subset of the HapMap's discarded assays and found that deletions leave distinct imprints in SNP genotypes. These patterns recur in neighboring SNPs that are physically clustered on a chromosome and help outline the boundaries of a deletion. When applied to the entire HapMap sample collection this method revealed 541 putative deletion variants in the genome, most of which were not observed previously. These deletions ranged from 1 to 745 kilobases in length and about half were present in several unrelated individuals. The researchers then verified 90 of these candidates as true deletions. Among these, they found portions of genes frequently missing throughout the population, including ten genes that function in sex hormone metabolism, olfaction and drug detoxification.

To determine how the removal of genes influences normal physiology and disease, scientists must first measure deletion frequencies in different patient groups. SNP-based approaches are feasible, but only if the deletions are indivisibly tied to nearby SNPs in the genome. By examining the ten genes noted above, which are often disrupted by deletions, Broad researchers determined that a single SNP, known as a "tag" SNP, often served as a proxy for the deletion's location. This suggests that deletions emerged from unique events in human history and that they have been propagated in the population through shared ancestry. As deletions were passed to subsequent generations, they carried adjacent SNPs along with them and now, these SNPs provide a practical tool for identifying common deletions. In the future, studies that use SNPs to poll genetic variability may help illuminate the biological importance of deletions and, perhaps, their roles in human disease.

Paper(s) cited

McCarroll SA, Hadnott TN, Perry GH, Sabeti PC, Zody MC, Barrett JC,
Dallaire S, Gabriel SB, Lee C, Daly MJ, Altshuler DM & The
International HapMap Consortium. Common deletion polymorphisms in the human genome. Nature Genetics Advance Online Publication 4 December 2005 DOI:10.1038/ng1696