Structurally informed evolutionary models improve phylogenetic reconstruction for emerging, seasonal, and pandemic influenza viruses

Xueting Qiu; Justin Bahl

Introduction 39 The importance of statistical phylogenetic methods to study the epidemiology, evolution 40 and ecology of rapidly evolving viral pathogens has been driven by the growing availability of 41 whole genome sequences (1). Precisely estimating the pattern of genetic variations is critical to 42 reconstruct the accurate pathogen phylogenies (2,3). Nucleotide substitution models have been 43 developed to describe the process of change from one nucleic state to another among viral 44 isolates, where they often allow for rate variations between transitions and transversions (4), or 45 incorporate nucleotide base frequencies with substitution rate parameters (5,6). Segmented The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. reporting domain-specific evolution rates and approximate selection pressure. The tree branches 94 with domain-specific rate ratio dC1+2/dC3 can inform the approximate selection pressure on each 95 strain, suggesting some biological explanations related to antigenic drift and emerging strains. 96 The structurally informed phylogenetic model may reveal novel biological insights of viral 97 evolution and have the potential to reveal more biological realism without over-parameterization.