https://github.com/mrc-ide/reestimate_covidifr_analysis

Repo for work on ICL Report 34 and subsequent Comm Med Methods Paper
https://github.com/mrc-ide/reestimate_covidifr_analysis

Last synced: about 2 months ago
JSON representation

Repo for work on ICL Report 34 and subsequent Comm Med Methods Paper

• Host: GitHub
• URL: https://github.com/mrc-ide/reestimate_covidifr_analysis
• Owner: mrc-ide
• Created: 2020-05-26T18:16:20.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2022-01-11T23:15:27.000Z (almost 3 years ago)
• Last Synced: 2023-08-01T05:14:21.196Z (over 1 year ago)
• Language: R
• Homepage:
• Size: 302 MB
• Stars: 9
• Watchers: 8
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Reestimating IFRs with Serology Data Research Compendium 

[![DOI](https://zenodo.org/badge/267118325.svg)](https://zenodo.org/badge/latestdoi/267118325)

The structure of this research compendium loosely follows the framework of an R-package for reproducing our analysis (for more information on research compendiums, see [ROpenSci Team](https://github.com/ropensci/rrrpkg)).  

An overview of the compendium is below: 

```

├── R                                 # "Helper" R functions specific to this research project 

├── README.md

├── analyses/                         # Scripts for data wrangling, fitting, figures, and tables

│   ├── ModFits/

│   │   ├── CareHomes_drakeworker.R

│   │   ├── ModFits_drakeworker_v2.R

│   │   ├── seroRev_ModFits_drakeworker_v2.R

│   ├── Rgn_Mod_Stan/

│   │   ├── run_regional_model.R

│   ├── SimWork/

│   │   ├── SeroRev_SimCurves_drakeworker_v2.R

│   │   ├── SimCurves_drakeworker_v2.R

│   │   ├── fit_conceptual_figure_delayeffects.R

│   │   ├── fit_seroday_concept.R

|

│   ├── 00-convalescent_seroAbs_v2.R

│   ├── 01-run_process_country_data_v2.R

│   ├── 02-descriptive_data_plots_v3.R

│   ├── 03-plot_methods_conceptual.R

│   ├── 04-analyze_overall_IFRs.R

│   ├── 05-analyze_age_IFRs_v2.R

│   ├── 06-analyze_carehome_IFRs.R

│   ├── 07-mod_posteriors.R

│   ├── 08-model_ppcs.R

│   ├── 09-collate_sims.R

│   ├── 10-plot_delay_concept.R

│   ├── 11-plot_seroday_concept.R

|

├── data/                                 # DO NOT EDIT ANY FILES IN THIS DIRECTORY BY HAND

│   ├── raw/

│   ├── derived/

│   ├── simdat/

│   ├── plot_aesthetics/

|

├── drake_clst/                          # Instructions for a slurm scheduler for resource allocation

|

├── reports/                             # Reports for simulations and study-specific runs                    

├── reestimate_covidIFR_analysis.Rproj   # R-proj

├── run_nightly_workers.sh               # Workers for generating MCMCMC fits

|                              

```

### Running the Code

Users will first need to perform model fitting for each of the included studies and/or the simulation runs with code provided in the `analyses/ModFits/` and `analyse/SimWork/` directories, respectively. These age-based models depend on the [`COVIDCurve` R-Package](https://github.com/mrc-ide/COVIDCurve). Individual "worker" scripts for models without and with seroconversion are available (you can send these out on a slurm cluster with the `run_nightly_workers.sh` script). From experience on a Linux-based cluster, fits take approximately 8-32 hours to complete. 

  

  

After you have produced the model fits, you can run the rest of the analyses in sequential order 01-11 (script `00-convalescent_seroAbs_v2.R` relies on data that is available upon request. The needed parameter results are provided in the `results/prior_inputs` directory).