Abstract
In principle, recombinant genealogies (Ancestral Recombination Graphs or "ARGs") contain all knowable information about the history of a set of genome sequences; they generalise the idea of a coalescent tree to a more complex genealogical network. They therefore present a compelling framework for unifying genetic ancestry with evolutionary history. Although the concept of an ARG was proposed several decades ago, it is only with the advent of more powerful simulation and inference algorithms that ARGs have become usable in population genetics. In particular, the tskit library, originally developed as an efficient way to store the result of whole genome simulations, can be used to store, process, and analyse various sorts of ARGs. These genetic genealogies can be either simulated via a combination of backward- and forward-time simulation (using tools such as msprime and SLiM), or inferred from phased sequence data. In this talk I will present a brief history of the concept of ARGs, highlight the availability of standardized demographic simulations of different species, discuss the current state-of-the-art when inferring ARG-like structures from sequence data, and speculate on future developments in this fast-growing area of population genetics.
For more information on Yan Wong, have a look at Wong's profile page.
Or you can see him presenting "Bang goes the theory" at BBC.