The code and other support files associated with the Rossenkhan et al. paper re: infection timing estimation for AMP.
This software was originally run on 19 June 2020 as an operational step of the AMP trial sieve analysis. This software is presented here as a historical archive of what was used in that moment, for the purposes of reproducibility. Any use of this method in new research would be better accomplished with the "tsic3" package, available here:
[URL of tsic3]
This software was run on Ubuntu 18.04.5 LTS using R version 3.6.3 (2020-02-29)
The specific R scripts and their required libraries are listed below.
0_support
Early files from the tsic2 package (in the directory tsic2) and the library file tsic2_supplement.Rlib:
profvis
1_data_preparation
amp_dx_history_preprocessing.R:
tidyverse
2_run_analysis
run_hbayes.R:
ggplot2
tsic
run_weib3_idt.R:
ggplot2
Install required version of R. Install the tsic package: https://github.com/philliplab/tsic Install other required R packages. Create local directories for code and data.
Running the scripts requires configuring the "CONFIG" sections to map to the location of the files on your local filesystem.
This software also requires the installation of the "tsic" package written by Philip Labuschagne, which is available here:
https://github.com/philliplab/tsic
All script files need to be properly pointed to the source of the input files and folders. These path variables are indicated with the "CONFIG" tags in their comments.
The names of the directories in this repository indicate the order in which the scripts should be run:
0_support:
This directory contains code libraries required by these scripts, and includes an early (unreleased) version of the "tsic2" package. No code is run directly out of here.
1_data_preparation:
amp_dx_history_preprocessing.R: the script that processes the study file of lab test data. This script takes only a few seconds to run, and generates all files required by the next two scripts.
2_run_analysis:
run_hbayes.R: the script that estimates the infection dates using the Bayesian approach. This script takes 30-60 minutes to run, depending your computer.
run_weib3_idt.R: the script that estimates the infection dates using the Weibull method. This script takes 30-60 minutes to run, depending your computer.
From the command line starting in the code directory run the following commands:
R CMD BATCH amp_dx_history_preprocessing.R &
R CMD BATCH run_hbayes.R &
R CMD BATCH run_weib3_idt.R &
Same as demo instructions above.