Online Program

Comparing hybrid identification methods in phylogenetics (309919)

Keywords: phylogenetics, genomic data, pseudolikelihood, trait evolution

Phylogenetics studies the evolutionary history between organisms, and with the advances in sequencing technology, more genetic information has become available to infer gene flow events. These gene flow events can lead to hybrid speciation and transform the well known “Tree of Life” into a network. Analogous to phylogenetic tree inference, there are also methods to compute these network structures. However, these are often computationally expensive. As a more efficient alternative, hybrid methods compare triplets or quartets of species to identify any hybrids between them. Still, these hybrid methods have their own strengths and pitfalls, and their accuracy is impacted by factors such as depth of gene flow, branch lengths, and amount of genetic mixing. In this poster I will compare the performance of summary methods designed to detect hybridization events: HyDe, TICR, and MSCquartets. The ability of these methods to accurately detect various hybridization events will be compared using data from simulated networks with variability in introgression depths, numbers of available genes, and network sizes.