Inferring the Population Size History of 4 Cattle Breeds Using NGS Data and an Approximate Bayesian Computation Approach

Inferring the ancestral dynamics of effective population size is a long standing question in population genetics, which can now be resolved much more accurately thanks to the availability of massive genomic data in many species, including cattle. I present here a new Approximate Bayesian Computation approach, combining information from linkage disequilibrium and allele frequencies, to tackle this question. I evaluate the performance of this method using cross validation and show that it provides very accurate results up to at least 20,000 generations before present. I then apply it to large genomic samples in 4 cattle breeds (Angus, Fleckvieh, Holstein and Jersey), which have been produced by the 1000 bull genome project. This reveals interesting aspects of cattle history, from domestication to modern breed creation.