tinselR-an R Shiny Application for Annotating Phylogenetic Trees.

Public health laboratories obtain whole-genome sequences of pathogens to confirm outbreaks and identify transmission routes. Here, we present tinselR, an open-source and user-friendly application for visualization and annotation of relatedness among pathogens with phylogenetic trees. By combining annotation and phylogenetics, we simplify one critical step in the pathogen analysis workflow.

ANNOUNCEMENT

The R programming language offers many powerful packages for phylogenetic analyses and visualization (e.g., ape [1] and ggtree [2]). Although R workflows are powerful, its command line interface can be difficult and time-consuming to master for creating data visualizations. Many researchers might opt instead to create a figure using software with a graphical user interface (GUI). By leveraging Shiny (3), R’s native library for creating interactive applications, we can harness the power of R and its phylogenetic packages while abstracting away some of the programming complexity and making it easier to visualize and annotate phylogenetic trees for the nonexpert.

To this end, we developed tinselR (pronounced tinsel-er), an R Shiny application that can be run locally or deployed to the cloud. tinselR’s minimum input requirement is a Newick-formatted phylogenetic tree, but it can also take a genetic distance matrix of single nucleotide polymorphisms (SNPs) along with user-defined metadata. Once the data are loaded, one can quickly modify the appearance of the plotted tree to include annotations, relabel tips, or add a heat map. The modified tree images are downloadable in PDF, PNG, and TIFF formats for presentations, publications, or other communication with collaborators and stakeholders.

Installation, example data, and additional resources. To install tinselR from GitHub, users will need to install the R package devtools (4) using R version ≥3.6. The R packages ggtree (2) and treeio (5) are also required and can be installed from Bioconductor using BiocManager. With the installation of these dependencies, tinselR is installable via the install_github command from devtools. Explicit installation commands are provided in Fig. 1a, and the final command (run_app) will launch the application locally. Note that install_github will also install other missing R dependencies. tinselR will accept Newick tree files from any program, e.g., RAxML (6), as input. Although it is possible to host R Shiny applications on a server, to date tinselR has been tested only by single users running the application locally. We recommend testing to ensure that tinselR performs as expected under multiuser conditions before providing access from a server for production purposes.

FIG 1

(a) Code which will install and launch the tinselR application. Please visit the GitHub page to determine the release version for installation and specify that in the devtools::install_github command or install the latest source using the command devtools::install_github(“jennahamlin/tinselR@*release”). (b) Example data set 1 displayed with SNP annotations and a heat map indicating collection source.

After launching tinselR, new users can explore the application using one of the preloaded data sets located in the “example data” tab. We provide three complete data sets once the application is launched (i.e., Newick-formatted tree, genetic distance matrix, and metadata file). After clicking on the example data tab, a user can select one of the data sets (e.g., example data set 1, 2, or 3) from the drop-down menu to test the application. We highlight the capabilities of tinselR using example data set 1 (Fig. 1b). The example data are from either Escherichia coli (NCBI BioProject PRJNA218110) or Salmonella enterica (NCBI BioProject PRJNA230403) with the number of isolates ranging from 14 to 19. The genomic data used in the example data sets were generated and used under the CDC institutional review board (IRB) protocol 7172.

Data availability.

Additional documentation for tinselR is located online at https://jennahamlin.github.io/tinselR/, which includes a vignette that describes the type of input data, example data, and a more detailed description for how to use the application. The source code and ability to file an issue are located online at https://github.com/jennahamlin/tinselR.