Bases-dependent Rapid Phylogenetic Clustering (Bd-RPC) enables precise and efficient phylogenetic estimation in viruses
Bin Ma,Huimin Gong,Qianshuai Xu,Yuan Gao,Aohan Guan,Haoyu Wang,Kexin Hua,Rui Luo,Hui Jin
DOI: https://doi.org/10.1093/ve/veae005
IF: 5.614
2024-01-01
Virus Evolution
Abstract:Abstract Understanding phylogenetic relationships among species is essential for many biological studies, which call for an accurate phylogenetic tree to understand major evolutionary transitions. The phylogenetic analyses present a major challenge in estimation accuracy and computational efficiency, especially recently facing a wave of severe emerging infectious disease outbreaks. Here, we introduced a novel, efficient framework called Bases-dependent Rapid Phylogenetic Clustering (Bd-RPC) for new sample placement for viruses. In this study, a brand-new recoding method called Frequency Vector Recoding was implemented to approximate the phylogenetic distance, and the Phylogenetic Simulated Annealing Search algorithm was developed to match the recoded distance matrix with the phylogenetic tree. Meanwhile, the indel (insertion/deletion) was heuristically introduced to foreign sequence recognition for the first time. Here, we compared the Bd-RPC with the recent placement software (PAGAN2, EPA-ng, TreeBeST) and evaluated it in Alphacoronavirus, Alphaherpesvirinae, and Betacoronavirus by using Split and Robinson-Foulds distances. The comparisons showed that Bd-RPC maintained the highest precision with great efficiency, demonstrating good performance in new sample placement on all three virus genera. Finally, a user-friendly website (http://www.bd-rpc.xyz) is available for users to classify new samples instantly and facilitate exploration of the phylogenetic research in viruses, and the Bd-RPC is available on GitHub (http://github.com/Bin-Ma/bd-rpc).
virology