Speaker Steven Kelk (CWI) talks about the progress and challenges of efficiently constructing phylogenetic networks from rooted triplets.

Room M279

Traditionally the evolutionary relationships between species (or, more generally, any set of taxa) have been modelled with a phylogenetic tree. However, the existence of reticulation events (such as hybridisation, recombination and lateral gene transfer) has motivated increased interest in the more general concept of phylogenetic networks.

Phylogenetic networks can be modelled with directed acyclic graphs, which generalise trees. Steven Kelk: "We have been studying the following problem: if a biologist gives us, for certain size-3 subsets of the taxa, a hypothesis about how those three taxa evolved relative to each other (a rooted triplet), can we efficiently assemble a single phylogenetic network that includes all these taxa but contradicts as few of the original triplets as possible? In this talk I will review some of our recent progress in this area and chart the areas in which both algorithmic and biological challenges remain. The emphasis of this talk is primarily algorithmic/combinatorial but with biological motivation provided throughout and will be understandable for a general mathematical audience."