Skip to main content
Premium Trial:

Request an Annual Quote

Harvard Team Develops Method for Honing in on Natural Selection in the Genome

NEW YORK (GenomeWeb News) – Researchers from Harvard University have come up with a method for refining their view of recent signals of natural selection in the human genome.

Pardis Sabeti, an evolutionary biologist at Harvard University's FAS Center for Systems Biology, presented various aspects of the work during two talks at the American Society of Human Genetics meeting in Honolulu last week — one during a session on the medical uses of genomic sequencing and another during the meeting's closing symposium on Darwin and evolutionary biology.

A handful of human genes under natural selection have been identified in the past, Sabeti noted. One of the best known: mutations in the hemoglobin beta gene HBB that cause sickle cell anemia but also increase resistance to malaria.

Beyond such classic examples, though, researchers are developing methods for finding selection patterns all across the human genome. Because traits that increase an individual's reproductive success spread through populations more quickly than other mutations, Sabeti explained, it's possible to hunt down signals of natural selection in the genome.

Even so, she added, the specific genes behind these signals are sometimes hard to pin down. To remedy this, Sabeti and her team came up with a "composite of multiple signals" (CMS) approach that goes beyond the "long haplotype" approach she and her colleagues have used in the past.

The CMS method integrates data from three different tests for selection, combining long haplotype data with information from derived allele and differentiation tests, Sabeti explained. While each test alone provides some information about selection in the genome, putting all three together is even more informative, offering a further refined view of selection.

For instance, Sabeti said, the method helped the researchers narrow in from a 600,000 base region on chromosome 15 under selection in European individuals to a much smaller, 34,000 base signal involving a gene called SLC24A5. The gene is thought to explain roughly 30 percent of pigmentation differences between Europeans and Africans.

Among the other signals of selection the researchers identified were a signal in a chromosome 2 gene called EDAR that appears to be linked to hair-related traits and is under selection in Asian populations, and a signal in the chromosome 12 gene PAWR that is influenced by natural selection in African individuals and seems to be involved in immune system function and pathogen response.

In general, Sabeti said, genes involved in sensory processes are often under selection in Asian populations, while those contributing to immune system function tend to be under selective pressure in Africans. Meanwhile, she added, metabolism-related genes appear to be selected in all of the populations studied so far. "The picture of selection in the genome is getting more and more filled out," she said.

Although genome-wide analyses of selection are relatively new, researchers are already taking the first steps toward translating insights from such studies into public health efforts. For example, Sabeti described how selection signals in the arenavirus-response related gene LARGE in Nigerian individuals, reported in a 2007 Nature paper, led to new diagnostic testing facilities for an arenavirus called Lassa virus in Irrua, Nigeria.

Down the road, Sabeti said, the CMS approach should prove useful for picking additional signals of selection out of data from HapMap efforts, the 1000 Genomes Project, and other large population studies.