Are theoretical advances in distance-based tree learning practical for phylogenetic inference?
The phylogenetic tree of relatedness for species, viral samples, or other biological taxa is essential information for all fields of evolutionary biology. Many different statistical and computational tools for phylogeny inference are used by practicing biologists, but none take advantage of a body of theoretical work on fast-converging algorithms that guarantee correctness with high probability under particular conditions. Here, we assess the utility of one of the most advanced of these algorithms when applied in reasonable biological situations. Our simulation study shows that realistic datasets will often not meet the assumptions of the algorithm, but also that the results are relatively robust to this problem. We additionally provide guidance on how the algorithm can be deployed when the true tree is not known, which is essential for any real-world application. Overall, our intention is to bring a class of methods to the attention of the phylogenetics community, and to make the algorithmic community aware of the needs of practicing biologists.