ekamau/bayesiantitercalc

Bayesian calculation of serum neutralization antibody titer

bayesiantiters

bayesiantiters is a Bayesian inference method that calculates a serum sample’s antibody concentration, $\phi$, from standard virus neutralization assay data using a mechanistic model. The method uses a dose-response relationship in a logistic function to simulate mortality of cell culture monolayers as a function of antibody concentration.

Given the limited information per sample that’s inherent in the experimental design of neutralization assays (number of replicates per dilution and number of dilutions), the advantage of Bayesian inference here is the use of probability distributions to incorporate uncertainty in the outcome.

Installation

You can install the development version of bayesianTiterCalc from GitHub with:

# install.packages("devtools")
devtools::install_github("ekamau/bayesianTiterCalc")

bayesiantitercalc's Issues

Add references for Reed and Muench paper to `reedmuenchTitrations` data

rstan options / functionality

How would we incorporate these rstan options in the package:

options(mc.cores = parallel::detectCores()).
rstan_options(auto_write = TRUE)

Do we need them? for large data?

Readme plot

Change so that we have more realistic data patterns

Change inputs to `sample_dose_response`

Currently we have m and n but would like lower and upper.

Combine 3 Stan models into 1

Remove global `Phi_Vals` from `sample_dose_response`

Remove `simData.plot` functionality

Add continuous integration testing

Create a `pkgdown` website for the package

See here for an example (and the way to create such a website).

Add vignette

Create a vignette using usethis::usethis::use_vignette().

An example vignette from the dplyr package is here.

Error in summary(fit)$summary : $ operator is invalid for atomic vectors

I am getting this error from the sampling function ... I can't figure out why, and the output of the model fitting function is class 'stanfit'..
Based on this website - https://cran.r-project.org/web/packages/rstan/vignettes/stanfit-objects.html - summary(fit)$summary should work.

I never get this when running the models 'normally'.

Add calculation of original Reed and Muench estimate

Remove output and other misc files that get created when Stan files are compiled