Inferring Phylogenetic Networks in the Genomic Era

While phylogenetic trees (i.e., branching diagrams that depict the evolutionary history of different organisms) have been essential for understanding species evolution, they do not fully capture certain evolutionary processes, such as hybridization. In these cases, a phylogenetic network, which extends a phylogenetic tree by allowing two branches to merge into one and create reticulations, is needed. However, existing methods for estimating networks from genomic data become computationally prohibitive as dataset size and topological complexity increase. In this talk, I present the performance of popular computational methods that detect hybridization from genomic data as an alternative to the network inference, discussing their significance and limitations. I then explain how phylogenetic networks generalize trees to represent complex evolutionary histories and explore the biological interpretations that can be drawn from various branching patterns. Finally, I introduce PhyNEST (Phylogenetic Network Estimation using SiTe patterns), a novel method that efficiently and accurately infers phylogenetic networks directly from sequence data using composite likelihood. PhyNEST is implemented as an open-source Julia package and is available at https://github.com/sungsik-kong/PhyNEST.jl.