CMS is a tool for identifying regions that are under positive selection in the genome and for localizing the causal variant within these regions. It generates a composite score based on three distinct signatures of selection: long-range haplotypes, differentiated alleles, and high frequency derived alleles. By combining several, nearly-independent signals, CMS detects signals of positive selection at a much higher resolution than conventional tests, often at the level of single genes. It also has better power than haplotype-based methods for finding regions under selection, especially when the selection signal is fairly old (~20-35 kya).
CMS can be used on large-scale catalogues of human genetic diversity such as those from the International HapMap project to scan the whole genome for regions under selection. Now, with whole-genome sequencing data from the 1000 Genomes project, we can use CMS to test every nucleotide in the genome for evidence of selection. This gives us an unprecedented opportunity to discover novel signatures of positive selection in the human genome, and better understand recent human adaptation.
This website presents the results of a recent genome-wide survey of positive selection in data from the 1000 Genomes project. The Results section has the data from this analysis available for download, along with extensive annotations to identify promising candidate variants. The CMS Viewer tool allows users to explore this data interactively and download data for regions of interest.