This project involved sequencing and analysing more than 100 canid genomes, including both modern samples, historical samples spanning the last 200 years and ancient samples, ranging from 15000 to 50000 years before present. An amalgamation of these data allowed us to answer elementary questions, such as how the various wolves in the world are related to each other, and how these wolves are related to other species in the genus.
To answer these questions, we constructed a phylogeny using randomly chosen regions of the genome, including man wolf populations, dog breeds, African, North American and Eurasian canids, such as dholes, golden jackals, and Ethiopian wolves. Since the phylogeny (relationships) between these canids are affected by gene flow/admixture between these canids, the amount of uncertainly in the phylogeny was summarized using methods used to examine the frequencies of various gene trees. Our investigations show that the gray wolves have been mixing with other canids, such as dogs, coyotes and golden jackals depending on the local canids that inhabit the same environment as the gray wolves. Specifically, the gray wolves from the Middle East have a substantial amount of mixing with golden jackals, and North American wolves share significant amounts of ancestry with coyotes. In terms of genetic contributions of the gray wolves to other canids, we hypothesize, based on the measures of genetic similarity, that the African golden wolf is a hybrid species, with genetic contributions from Eurasian gray wolves and Ethiopian wolves. Additionally, in terms admixture among other canid groups, we found evidence of gene flow from a as yet unknown canid into the ancestors of wolves and coyotes.
In North American wolves, we performed a study using 15 newly sequenced North American wolves and coyotes. We investigated the population structure among the different populations of North American wolves, and further estimated the amount of coyote admixture to the different wolf populations. Our results showed that the Polar wolf (which inhabits Greenland) is closely related to the Arctic wolves (inhabiting Ellesmere and Baffin Islands), and the Arctic/Polar wolf is a distinct population of North American wolves. We also find that extensive gene flow between coyotes and gray wolves wherever their ranges overlap. The taxonomically controversial red wolves and eastern timber wolves appear to be of hybrid origin, with varying combinations of gray wolves and coyotes.
To supplement the information obtained from the modern genomes, we sequenced 5 individuals of Pleistocene wolves, a wolf type that went extinct more than 10000 years ago. We also sequenced a single samples of a dog dating back to 10000 years ago - this dog, at the time of this report, is the oldest known dog to be whole genome sequenced. Based on the placement and shared genetic affinity between the ancient dog and the sled dogs, we surmised that the sled dogs form a basal lineage of dogs, which diverged from other dog lineages at least 10000 years ago. Further, the sled dogs have some genetic material from the Pleistocene wolves, and show signatures of selection related to adaptation to the arctic environment.
