Rapid diagnostic tests (RDTs) are one component of Plasmodium falciplarim malaria case management in many parts of the world; however, deletion of histidine-rich protein genes pfhrp2/3 can result in false-negative RDT results in tests based upon detection of the histidine-rich protein 2 (HRP2) and its paralog histidine-rich protein 3 (HRP3). Projecting the spread of pfhrp2/3 deletions may allow local policymakers and international support agencies to plan for the possibility of increased false-negative RDT results.
This repository contains some of the computational work that was done in support of the pre-print "Global risk of selection and spread of Plasmodium falciparum histidine-rich protein 2 and 3 gene deletions" by Watson et al. as well as the "HRP2/3 Deletion Risk Explorer" app that allows users to explore how changes in model assumptions impact pfhrp2/3 deletion risk at a national level.
/
is limited to repository documentation and configuration files.
data/
contains data files that were used in preparation of the model, or the Deletion Risk Explorer.
R/
contains the source for the Deletion Risk Explorer.
task_1/
contains data and the Python Notebooks used to evaluate RDT data for the model.
utility/
contains various scripts used to processing the model data from hrpup for the Deletion Risk Explorer.
The HRP2/3 Deletion Risk Explorer is an R Shiny web application that can be accessed at URL, run locally in RStudio Desktop, or deployed to a web host such as shinyapps.io.
This repository uses Git Large File Storage, which will need to be installed:
sudo apt install git-lfs
Once LFS is installed, you can then pull the large files via:
git lfs pull
After this repository has been downloaded to your local machine and opened in RStudio, the dependencies can be installed via the following commands:
install.packages("devtools")
devtools::install_deps('.')
After which the app can be launched by either opening R/launch.R
and clicking "Run App" or by the following command:
runApp('R/launch.R')
The first time that the app runs warnings may be produced by CPL_gdal_init()
, but these can generally be ignored if the app loads successfully.
Additional language translations are welcomed and should be submitted as a pull request on GitHub after forking the rjzupkoii/WHO-Malaria-pfhrp23 repository on GitHub.
The directory R/lang
contains all of the necessary files for the user interface, structured as follows:
- Directories that are the two character ISO language code (ex.,
lang/es
for Spanish) containing Markdown files (ex.,explainer.md
) that are rendered as such, with some adjustments to the layout supplied bystyle.css
. All Markdown files inlang/en
need to be supplied as part of a new translation. - CSV files named as
translation_[ISO CODE].csv
(ex.,translation_fr.csv
for French) that contains two columns. The first of which isen
and contains the English labels in the user interface, and the second two character ISO code for the target language and contains the translated text. - The
translation.yml
file which supplies the translation for rendered maps. The top level key for each language is the two character ISO code, followed by fields forchoices
,regions
,labels
,risk_map
, andcomposite_map
.
When the loads for a user it defaults to English, after which the language can be changed by the user via the Change Language dropdown list. The list of language is determined by the contents of the choices
list supplied to the language
selectInput
control. New languages will need to be added to this choices
list and should be added in alphabetical order, with the control text being the name of the language as it would appear to a native speaker (i.e., French is listed as Français in the list). The value returned by the control is the two character ISO language code, which is the used to select the correct language information from the server.
Underlying model data from hrpup, which is released under CC-0.
World map data in R/data/R6_WHO_Compliant_map.rds
courtesty WHO Division of Data, Analytics and Delivery for Impact (DDI) and is released under the WHO publishing policies as CC BY-NC-SA 3.0 IGO.