R/binding.R

In BradyAJohnston/biochemr: Tools For User-Friendly & Reproducible Analysis of Biochemistry Experiments

#' Calculate Binding Affinities
#'
#' Calculated binding affinities (Kd) from experimental data that contains dose
#' and response information. Could be titrated concentration and resulting
#' fluorescence, or percentage shift in a EMSA or something similar.
#'
#' @param data data.frame or tibble containing dose response data.
#' @param conc Column with the concentration information.
#' @param resp Column containing the response information.
#' @param slope For all 1:1 binding interactions, the slope for the log -
#'   logistic curve should be 1. If the interaction has some cooperativity or it
#'   is not a 1:1 binding interaction, then this value will change and should be
#' @param group
#'   allowed to be free (set `slope = NA`).
#' @return a [tibble][tibble::tibble-package]
#' @export
#'
#' @examples
#' Puromycin %>%
#'   bio_binding(conc, rate, group = state)
bio_binding <-
  function(data,
           conc,
           resp,
           group = NULL,
           slope = 1) {
    if (!is.numeric(slope)) {
      if (!is.na(slope)) {
        stop(
          "Slope must be numeric (likely 1) or set to NA for the model to guess."
        )
      }
    }

    model <- drc::LL.4(
      names = c("slope", "min", "max", "kd"),
      fixed = c(slope, NA, NA, NA)
    )

    bio_dose_resp(
      data = data,
      .dose = {{ conc }},
      .resp = {{ resp }},
      .group = {{ group }},
      .model = model
      )
  }