Inferring human population history and gene flow from multiple genome sequences. S. Schiffels, R. Durbin Wellcome Trust Sanger Institute, Cambridge, Cambridgeshire, United Kingdom.

   The availability of complete human genome sequences from populations across the world has given rise to new population genetic inference methods such as the Pairwise Sequentially Markovian Coalescent (PSMC). While PSMC infers demographic history for times between 50kya and 2mya with good resolution, neither the more recent evolutionary history nor migration patterns across populations can be addressed. Here we present a new method that overcomes both of these shortcomings. The Multiple Sequentially Markovian Coalescent (MSMC) infers more recent evolutionary history within and across populations up to a few thousand years ago. MSMC models the pattern of mutations in multiple genome sequences under the sequentially Markovian coalescent with recombination. It fits local genealogical trees to the observed pattern, focussing on the first coalescence among any two individuals. We apply our method to genome sequences from several family trios from the 1000 Genomes project with African, European, Asian and Native American ancestry, inferring population sizes and migration rates as a function of time with high resolution. In particular, our method gives information about the separation of non-African populations after the out-of-Africa event, and observes the bottleneck involved in the peopling of the Americas.

You may contact the first author (during and after the meeting) at