Statistical phylogeography: approximate Bayesian computation

Last time we explored Migrate and IMa. We saw that they can (potentially) provide great insight into long-term patterns of gene exchange and, in the case of IMa, the pattern of population relationships. A great advantage of both packages, in principle, is that you don't have to tell them what you're looking for. They let the data speak for themselves. Of course, they let the data speak for themselves only to the extent that the demographic model underlying them is appropriate for the samples you've collected.

Sometimes you'll be investigating a problem for which you have some pre-defined scenarios that you'd like to compare. For example, you might wonder whether a set of populations are derived independently from a single, widespread ancestral population or from several different populations from different glacial refugia. For these purposes you need a different approach, and that's where Approximate Bayesian Computation (ABC) comes in. We'll talk about the principles of ABC tody and discuss one example of how it's been applied.

Online notes

Statistical phylogeography: approximate Bayesian computation

Return to lecture schedule