Novel metric for hyperbolic phylogenetic tree embeddings

Abstract Advances in experimental technologies, such as DNA sequencing, have opened up new avenues for the applications of phylogenetic methods to various fields beyond their traditional application evolutionary investigations, extending development, differentiation, cancer genomics, and immunogenomics. Thus, importance is increasingly being recognized, development a novel approach can contribute several areas research. Recently, use hyperbolic geometry has attracted attention artificial intelligence Hyperbolic space better represent hierarchical structure compared Euclidean space, therefore be useful describing analyzing tree. In this study, we developed metric that considers characteristics tree representation space. We performance proposed embeddings, general confirmed our method could used more precisely reconstruct distance. also demonstrate predicting nearest-neighbor node partial with missing nodes. Furthermore, based on integrate multiple trees nodes or imputing distances. This study highlights utility adopting geometric further advancing methods.