|tPRiors| Bayesian true PRevalence inference via elicited priors - Shiny Contest Submission (Accepted for publication)

shiny Shiny Contest
1

|tPRiors| Bayesian true PRevalence inference via elicited priors

Authors: Konstantinos Pateras, Polychronis Kostoulas

Abstract: A web-based application which integrates statistical methods for Bayesian inference of infection prevalence based on prior elicitation named |tPRiors|.

Full Description: A web-based application which integrates statistical methods for Bayesian inference of infection prevalence based on prior elicitation named |tPRiors|. tPRiors is an application that incorporates elicited prior knowledge, transforms it to prior distributions and calculates posterior inferences for a variety of prevalence models. tPRiors makes the development and implementation of prevalence Bayesian inferences less cumbersome to researchers with or without strong statistical knowledge. More information would become available in the upcoming article <K Pateras and P Kostoulas, tPRiors: An R Shiny tool for generating prior and producing posterior distributions for disease prevalence>

Keywords: Disease prevalence, biostatistics, Bayesian statistics, epidemiology, healthcare, infectious diseases
Shiny app: https://kpateras.shinyapps.io/tPRiors
Repo: GitHub - kpatera/tPRiors: tPRiors: Bayesian prevalence estimation
RStudio Cloud: Posit Cloud

Thumbnail:
image

Full image:

image
image865×412 287 KB

3 Likes
2

I get an error 404 when following the link to the GitHub repository. Since the account exists, I assume that the repository is private?
Would you mind making the repository public? (note that the RStudio Cloud instance already makes your code public).

Thank you!

3

Thank you for noticing Kevin, I have made the GitHub repository public. Cheers!

1 Like
4

Brilliant! Comprehensive design. You have thought of allrequired user needs.
There are possible bugs (minor), but I believe, these are inevitable parts of the the development process.
Have you considered adding progress bar or some feedback in the UI when a long process runs in the background? It may help some impatient users like me :slight_smile:

5

Thank you for your kind words! Indeed, in the most recent version which is currently also being submitted for publication, I have tried to introduce some progress bars. My end result is not always intuitive, though. Have you applied similar features before? Do you have a specific preference?

Kostas

6

Hi Kostas,
I don't have any preference. However, if you ask me, I will suggest either of withprogress() or using shinycssloader. The later is simpler to implement.
Regards,
Asitav

1 Like
7

This Shiny web-application has now been accepted for publication in "BMC *Medical Research Methodology" journal.

Preprint can be found here: tPRiors: A tool for prior elicitation and obtaining posterior distributions of true disease prevalence | Research Square

Soon the published article will be available here: https://doi.org/10.1186/s12874-022-01557-1

There is also a reference for the participation in the shiny-contect of 2021 in the acknowledgements.

Best,
Konstantinos