CZ GEN EPI allows users to perform phylogenetic analysis through UShER and Nextstrain. The tool of choice will depend on the questions you are trying to address. UShER is an ultrafast phylogenetic placement tool used to place user-selected samples onto an existing and routinely updated phylogenetic tree. On the other hand, Nextstrain allows users to build phylogenetic trees tailored to their needs (see phylogenetic tree types). Below is a summary of distinguishing features between UShER and Nextstrain and a quick overview of each tool.
See tables below for a summary of distinguishing features between UShER and Nextstrain and questions that can be addressed with each phylogenetic analysis tool.
The main advantage of phylogenetic placements through UShER is speed (see guide on how to use UShER). This tool allows you to place your samples on subtrees with closely-related sequences found in public databases within minutes. This allows you to quickly infer relationships between closely related samples with comparable accuracy to that of phylogenetic tree building from scratch (e.g., Nextstrain trees). If you need to quickly identify public samples that are most closely related to your samples, UShER is the tool to use. However, your samples may land on different subtrees and you will not be able to see how your samples relate to each other (unless they are closely-related and cluster on the same subtree). Additionally, UShER does not incorporate any metadata such as collection date or location information into the tree. Use Nextstrain to see all the samples of interest in the same tree alongside contextual samples and integration of metadata.
Nextstrain is an open source toolkit for phylogenetic tree building and visualization. Nextstrain trees are built through CZ GEN EPI automatically or on-demand (see guides regarding tree types, how to build them, and tree viewing). Tree building allows you to evaluate how your samples relate to each other and other publicly available samples. Nextstrain trees also incorporate metadata such as sample collection time and geographic locations. This allows users to address questions regarding outbreak dynamics, such as timing between infections and spread. However, Nextstrain tree building can take up to 12 hours. Use UShER if you are interested in quickly seeing what are the most closely related samples to yours.
Please sign in to leave a comment.