Inference of demographic parameters using importance sampling on coalescence history

International audience ; Inference of demographic parameters using importance sampling on coalescence history R. Leblois venue to SMBE 2008 was partially supported by a grant from the Biodiversity : Endemism in New Caledonia : A phylogenetic and population-level study of its origins (MNHN, OSEB) Few other demographic models such as single population with size fluctuations in time, n-population with constant migration and the possibility to consider multiple samples in time are almost ready to be tested. We are looking for students or post-docs to develop those models. Do not hesitate to contact us if you are potentialy interested. Migraine will thus be regularly updated to consider more demographic and mutational models so Check the Migraine web page regularly to find new model implementations and Download Migraine at : State of our art We developed and evaluated the performance of maximum likelihood (ML) analysis of allele frequency data in a linear array of populations under isolation by distance (IBD) and in a model of isolation with migration (IM) between two populations. Inference in a linear habitat under IBD The application on a linear IBD is the only algorithm that has been published so far (in Rousset & Leblois 2007 MBE). It has been validated using simulated and real data set analyses. Under this IBD model, the parameters are a mutation rate and either a dispersal rate in a stepping stone model or a dispersal rate and a scale parameter in a geometric dispersal model. An approximate procedure known as maximum product of approximate conditional (PAC) likelihood is found (i) to perform as well as ML (Fig. 2) and (ii) to be 10-1000x faster than ML, allowing the consideration of much larger habitat and more complex demographic models. Linear IBD simulation results Mis-specification biases may occur because the importance sampling algorithm is formally defined in term of mutation and migration rates scaled by the total size of the population, and this size may differ widely in the statistical model and ...