Genealogical trees are a central organising principle in genetics, capturing the relationships between a set of sampled genomes and their genetic ancestors. In recombining organisms, however, there is usually not a single tree describing these relationships along the genome, but rather a sequence of trees. This ensemble of correlated genealogical trees is usually referred to as an Ancestral Recombination Graph (ARG), and inferring ARGs has been the focus of concerted effort for over two decades. Until recently methods were extremely computationally expensive and were consequently rarely applied in practice. I will discuss recent computational breakthroughs that have brought ARGs to the forefront of population genetics research, and of the exciting new developments in simulation, inference and computation.
Dr Jerome Kelleher
Robertson Fellow in Biomedical Data Science, Big Data Institute, University of Oxford
My research revolves around developing efficient algorithms to solve fundamental problems in genomics, and implementing these algorithms in production quality, open source software. This programme takes advantage of the unique structure of genetic data, combining theoretical population genetics with classical computer science. I lead development of tskit, a growing library of fundamental operations for population and statistical genomics with a welcoming open source community.