OVERVIEW

This analysis estimates the mixture of transmissibility and immune escape for SARS-CoV-2 Omicron relative to Delta.

Methods Summary

1. Estimate the relative frequency of Omicron and Delta from S-Gene Target Failure (SGTF) and create a sample ensemble
2. Using sample ensemble and incidence data, including reinfections, create incidence ensemble
3. Estimate Rt for omicron and delta from ensembles, including Rt for omicron assuming shorter generation interval
4. Calculate ensemble of ratios
5. Compute NGM principle eigenvalue ratio for Omicron / Delta, assuming varying levels of immune escape for Omicron, including ratio for a shorter Omicron GI
6. Calculate transmissibility multiplier ensemble (= estimated ratio / NGM ratio)

Repository structure

Key folders

Folder	Purpose
R	R scripts each containing a step in the analysis workflow.
makefiles	Components of the workflow built using the scripts from the R folder. See the README for this folder for further details.
refdata	Key reference data used in this analysis. See the README in this folder for further details.
.devcontainer	Resources for reproducibility using vscode and docker.

Key files

File	Purpose
Makefile	Contains the workflow of this analysis. See the reproducibility section for details.
DESCRIPTION	Contains a summary of the dependencies used in this analysis readable from R.
sessioninfo.txt	Contains a full summary of the environment last used to generate results.

Reproducibillity

Reference & Analysis Directories

External source data for the analysis should be placed in refdata. By default, refdata will also be used as the root directory for intermediate and final analysis products. We recommend creating a makefiles/local.makefile which redefines REFDIR.

Dependencies

All R package dependencies can be installed using (in the working directory of the repository):

install.packages("devtools")
devtools::install_dev_deps(dependencies = TRUE)

Alternatively the dependencies can be installed manually (see the DESCRIPTION file for details of the dependencies and their sources). The provided docker image may be used to fully reproduce our analysis environment and this can itself be built using the provided Dockerfile. See the Docker documentation for more detail on using docker. For vscode users a .devcontainer has been provided which when used with the Remote development extension will automate setting up the supplied Dockerfile. The environment last used to generate results is also summarised in sessioninfo.txt.

Data requirements - NB: NOT CURRENT

This analysis requires the following data files which are not included in this repository:

File	Source
frequencies.rds
prov_ts_90_pub.rds

Make

This repository uses a Makefile to define the analysis workflow. This Makefile is split into component .makefile's stored in the makefiles folder. To reproduce the analysis (once the dependencies have been installed, and key data sources have been added) we can simply run the following in the command line:

make

To see a breakdown of workflow steps run the following instead:

make list

Individual steps can then be updated by running:

make <target>

Individual analysis steps can also be run interactively (see the R folder). See the makefiles folder for further details of the component workflow steps.