R/model-tQ-prior.R

In maomlab/BayesPharma: Tools for Bayesian Analysis of Non-Linear Pharmacology Models

#' Create a Prior for the tQ Enzyme Kinetics Model
#'
#' @description For both the `kcat` and `kM` parameters, the default priors are
#' `gamma(4, 1)`. We use the gamma distribution because it is naturally lower
#' bounded by `0`. The first parameter is the shape, `alpha=4`, and the second
#' is the rate, `beta=1`. The mean of gamma distributions is `alpha/beta` and
#' the variance is `alpha/beta^2`.
#'
#' @param kcat [brms::brmsprior] or `numeric`. Prior for `kcat` parameter. Given
#'   a numeric value, it will be used as a constant and not estimated.
#' @param kM [brms::brmsprior] or `numeric`. Prior for `kM` parameter. Given a
#'   numeric value, it will be used as a constant and not estimated.
#' @param ... additional [brms::brmsprior()] objects.
#'
#' @seealso [tQ_model], [tQ_formula], [tQ_init], and [tQ_stanvar]
#'
#' @export
tQ_prior <- function(
  kcat = brms::prior_string(prior = "gamma(4, 1)", lb = 0, nlpar = "kcat"),
  kM = brms::prior_string(prior = "gamma(4, 1)", lb = 0, nlpar = "kM"),
  ...) {
  
  c(
    prepare_prior(kcat, nlpar = "kcat"),
    prepare_prior(kM, nlpar = "kM"),
    ...)
}