Title: Inferring the recombination landscape of the Neanderthal and other ancient hominins using an extended sequentially Markov coalescent 

Abstract:

By efficiently modelling the coalescent process with recombination along genomes, sequentially Markov models (SMC) allowed to infer the demographic history of a population from a few individual genomes. Such models, however, make strong assumptions ove the underlying evolutionary process. In particular, it is assumed to be homogeneous along the genome sequence, while it is well established that in most species, recombination, mutation and effective population sizes substantially vary along the genome. We here introduce a general framework, which we name "integrative sequentially Markov coalescent -- iSMC", allowing to model such variation together with change of effective population size in time. We first apply it to the modelling of spatially variable recombination rates, and demonstrate the ability of the new model to jointly infer demography and recombination landscape at various scales, using both simulations and data analyses from humans, drosophila and a fungal pathogen. Most surprisingly, we found out that the model is able to recover the recombination landscape with high accuracy when only one unphased diploid genome is analysed. This allowed us to infer the recombination landscape of several ancient hominins, and to study its evolution.