NEW YORK (GenomeWeb News) – In a paper appearing in the advance, online edition of Science today, Broad Institute and Harvard University researcher Pardis Sabeti and colleagues report on their composite of multiple signals, or CMS, approach for finding parts of the human genome under positive selection.
"It's clear that positive natural selection has been a critical force in shaping the human genome, but there are remarkably few examples that have been clearly identified," senior author Sabeti said in a statement. "The method we've developed makes it possible to zero in on individual genes as well as the specific changes within them that are driving important evolutionary changes."
Sabeti described the team's CMS approach and its applications at the American Society for Human Genetics meeting this past fall. The current paper offers additional details on the method, which combines data on long haplotypes, high-frequency derived alleles, and highly differentiated allele signatures.
Whereas applying other tests for selection individually tends to uncover large genomic regions of a few hundred thousand or million bases that are under selection, the CMS method offers a much more refined view.
When the researchers used CMS to assess 185 regions implicated in Northern European, West African, and East Asian populations by International HapMap project data, they found population-specific signals of positive selection that were a median of 55,000 bases in size. Within these regions, they were able to identify new and previously reported causal variants.
Among the signals identified in East Asian populations, for example, were several affecting sensory perception-related genes, including variants in the chromosome 10 gene PCDH15, which codes for the protocadherin 15 and has been shown to contribute to inner ear hair cell development.
Other population-specific signals, meanwhile, appear to involve variants affecting metabolism, skin pigmentation, and immune function-related genes.
Still others localized to regions between protein-coding genes, leading the researchers to speculate that these may affect regulatory elements in the genome.
Although there's still a ways to go before researchers understand how these selective processes specifically affect gene functions in various human populations, the researchers noted, using the approach to examine data from large-scale population sequencing efforts such as the 1000 Genomes Project should provide insights into evolutionary forces acting on the human genome.
"As more data on human genetic variation becomes available in the coming years, an even more detailed evolutionary picture should emerge," Sabeti said, calling the current CMS-enhanced view of human evolution "one piece of a much larger puzzle."
In addition, she and her co-authors noted that CMS also holds potential for looking at other types of selection, including older selection signatures and selection patterns in non-humans.