Testing Tree-Likeness of Phylogenetic Network Data with Cross-Validation

• Activity Number: 151 - Novel Methods and Tools in the Era of Big Omics Data
• Type: Contributed
• Date/Time: Monday, August 8, 2022 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Md Rashidul Hasan* and James Degnan
• Companies: University of New Mexico and University of New Mexico
• Keywords: phylogenetic networks; rooted triple; inferring level-1 networks; fold cross validation; information criterion; multispecies coalescent model

Abstract

Topological phylogenetic networks at the species level can be inferred from gene trees, evolutionary trees estimated at different loci. So far, several statistical inference methods have been used under the multispecies coalescent model of a species network. For selecting a model in phylogenetics, both the Akaike information criterion and Bayesian information criterion are frequently used, but these criteria may be poor at model singularities and near boundaries. We propose a rooted triple approach in the context of inference of evolutionary trees while using inter-taxon distances and k-fold cross validation by checking whether each triplet is tree-like or not. This is a modification of the NANUQ algorithm and can lead to a new statistical procedure for inferring level-1 networks.