Team Leaders: Thomas Keller, Swamy Rakesh Adapa
Team Members: Atmaja Koorapati, Celine Atkinson, William Cromwell, Vijay Chauhan, Kyle Koller, Soa Fy Andriamandimby, Raúl A. González-Pech, Morgan Young
GVN/USF mentors:
The primary objective is to build an interactive application/ web app that can identify putative host-shifts where recent Covid sequences have evolved and jumped into a non-human host. We provide a variety of non-human sequences, and allow the user to upload a sequence of their choice. Then the new sequence is placed into the alignment and a phylogeny is built.
Host shifts will be identified as the novel sequence being placed in clades of other sequences of host-shift sequences, such as that of the human-mink shift in fall 2020.
We implemented the dashboard using an R package called flexdashboard, which allowed us to focus more of us efforts on the other individual components. Namely, the input, alignment, phylogenetic tree and visualization. The alignment was a custom Matlab script provided by one of the team members, Morgan Young. The tree estimation was done using the R package phangorn. Visualization of the tree was done using phylocanvas.
We provided a set of whole-genome Covid sequences that covered a variety of known human to non-human host shifts. These include the following species (ferret, mink, tiger,lion, more).
What did we find?
The main program is an Rshiny app. Therefore, you will need both a base R installation, as well as Rstudio. Once you have those downloaded, you will need to install the following packages:
phangorn
ape
DECIPHER
shiny
flexdashboard
phylocanvas
If you want to use some of the auxillary programs, the following packages are also required
fs
readr
purrr
dplyr
from bioconductor
Biostring