Open AccessScalable Reconstruction of SARS-CoV-2 Phylogeny with Recurrent Mutations
Author(s) -
Daniel Novikov,
S. P. Knyazev,
Mark Grinshpon,
Pelin Icer,
Pavel Skums,
Alexander Zelikovsky
Publication year - 2021
Publication title -
journal of computational biology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.585
H-Index - 95
eISSN - 1557-8666
pISSN - 1066-5277
DOI - 10.1089/cmb.2021.0306
Subject(s) - phylogenetics , phylogenetic tree , scalability , mutation , biology , visualization , evolutionary biology , computer science , computational biology , artificial intelligence , genetics , gene , database
This article presents a novel scalable character-based phylogeny algorithm for dense viral sequencing data called SPHERE (Scalable PHylogEny with REcurrent mutations). The algorithm is based on an evolutionary model where recurrent mutations are allowed, but backward mutations are prohibited. The algorithm creates rooted character-based phylogeny trees, wherein all leaves and internal nodes are labeled by observed taxa. We show that SPHERE phylogeny is more stable than Nextstrain's, and that it accurately infers known transmission links from the early pandemic. SPHERE is a fast algorithm that can process >200,000 sequences in <2 hours, which offers a compact phylogenetic visualization of Global Initiative on Sharing All Influenza Data (GISAID).