Analysis of multiplex antibody endpoints in the MORDOR Niger trial
This repository includes R code to run all of the analysis for the paper:
Effect of biannual azithromycin distribution on antibody responses to malaria, bacterial, and protozoan pathogens in Niger
Arzika, A. M., Maliki, R., Goodhew, E. B., Rogier, E., Priest, J. W., Lebas, E., O’Brien, K. S., Le, V., Oldenburg, C. E., Doan, T., Porco, T. C., Keenan, J. D., Lietman, T. M., Martin, D. L., Arnold, B. F. & MORDOR-Niger Study Group. Nat. Commun. 13, 976 (2022). https://pubmed.ncbi.nlm.nih.gov/35190534/
https://www.nature.com/articles/s41467-022-28565-5
Should you have any questions about the files in this repository, please contact Ben Arnold at UCSF (ben.arnold@ucsf.edu).
This GitHub repository is mirrored on the Open Science Framework (OSF). The OSF project page includes additional study-related resources, including the pre-analysis plan, the datasets, and the compiled HTML computational notebooks created from the .Rmd
files:
The data have also been archived on Dryad: https://datadryad.org/stash/dataset/doi:10.7272/Q6VX0DSD
Arnold et al. (2021), Multiplex IgG antibody response and malaria parasitemia among children ages 1-59 months in the MORDOR Niger trial, 2015-2018, Dryad, Dataset, https://doi.org/10.7272/Q6VX0DSD
Following: https://www.nature.com/documents/nr-software-policy.pdf
All analyses were run using R software version 4.1.1 on Mac OSX Big Sur using the RStudio IDE (https://www.rstudio.com).
> sessionInfo()
R version 4.1.1 (2021-08-10)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Big Sur 11.6.1
You can download and install R from CRAN: https://cran.r-project.org
You can download and install RStudio from their website: https://www.rstudio.com
All R packages required to run the analyses are sourced in the file mordor-ab-Config.R
.
The installation time should be < 10 minutes total on a typical desktop computer.
To reproduce all analyses in the paper, we recommend that you:
-
clone the GitHub repository
-
Create a
data
subdirectory and copy the two datasets from OSF or Dryad -
Create a
figures
subdirectory to store output. -
All of the analysis scripts should run smoothly (scripts
03-xx
to16-xx
).
The first two data processing scripts will not run because they read in our internal datasets with personally identifiable information, and create the final analysis datasets (shared publicly). We have included the data processing scripts for transparency and completeness.
You can run the .Rmd
notebook scripts one-by-one or you can compile mordor-ab-run-all.R
, which is the file we used to run the final analyses (e.g., from the command line R CMD BATCH mordor-ab-run-all.R &
).
Running the all analyses on the above Mac desktop configuration required 31 minutes.
Note that the only script that takes very long is 08-mordor-ab-community-means.Rmd
because estimating the ICCs and 95% CIs by bootstrapping binomial mixed models is computationally slow.
Also note: we attempted to create a Binder virtual machine option for this project, but the underlying .rds
dataset is so large (30 MB) that it took too long to spawn a remote docker container on the Binder server, and we didn't have time to troubleshoot/optimize.
This project is covered by the CC0 1.0 Universal license.