Code
cr0eso
As part of a recent project analysing outbreaks of COVID-19 in British Columbia long-term healthcare facilties,
we developed the R package 'cr0eso' for estimation of the basic reproduction number R0 in multiple outbreaks using an
SEIR model. This package, developed alongside Mike Irvine (BCCDC) and Sean Anderson (Fisheries and Oceans Canada), fits a Bayesian hierarchical
model to a collection of outbreaks in RStan. It is primarily designed for outbreaks
occurring in the same population but different facilities, e.g. healthcare, schools, workplaces or prisons. As well as
R0, it can also be used to estimate the effect of outbreak intervention strategies.
Serial intervals from genomic data
A team of us from the MAGPIE group have been developing a new method for estimating serial intervals using genomic data. As part of this, I've been
putting together shareable code to employ our approach, which will hopefully soon become an R package. This method estimates the distribution
of the serial interval from a collection of host symptom onset dates and their virus sequences, without needing any detailed information on who contacted who.
We provide a framework for performing this analysis in a set of case clusters, for example to explore if the serial interval is changing over time or in
different settings. The repository contains code to support our preprint,
and to apply our method to a simulated outbreak provided.