The analysis, using the most comprehensive genetic data set from Native Americans to date, was conducted using three different statistical models, two of them created by UC Berkeley researchers. The first, developed by the lab of Yun Song, a UC Berkeley associate professor of statistics and of electrical engineering and computer sciences, takes into account the full DNA information available from the genomes in the study. A second method, developed by Rasmus Nielsen, a UC Berkeley professor of integrative biology, and graduate student Kelley Harris, requires much less computation, but relies on a summary of the genome data. These and a third method developed by researchers at the Wellcome Trust Sanger Institute, England, all yielded consistent results.