R/krsa.R

In CogDisResLab/KRSA: KRSA: Kinome Random Sampling Analyzer

#' Main KRSA function. It performs the random sampling analysis and generates a Z score table and a count matrix
#'
#' Takes in a vector of the peptides that are considered "hits" and kinase-substrate and coverage dataframes and returns the Z score table
#'
#' @param peptides a vector of the peptides that are considered "hits"
#' @param itr number of iterations for the random sampling (default  = 2000)
#' @param seed seed number (default  = 123)
#' @param return_count boolean to return the kinase count matrix
#' @param map_file kinase-substrate dataframe
#' @param cov_file kinase coverage dataframe
#'
#' @return Z score tibble or list if return_count = TRUE
#'
#' @family core functions
#'
#' @export
#'
#' @examples
#' TRUE
krsa <- function(peptides, itr = 2000, seed = 123, return_count = F, map_file = KRSA_file, cov_file = chipCov) {
    message("Running KRSA ...")
    set.seed(seed)

    purrr::map_df(1:itr, krsa_sampling, cov_file, map_file, length(peptides)) -> temp

    temp %>%
        dplyr::group_by(Kin) %>%
        dplyr::summarise(SamplingAvg = mean(counts), SD = stats::sd(counts)) -> temp2

    cov_file %>%
        dplyr::group_by(Kin) %>%
        dplyr::summarise(
            Observed = sum(Substrates %in% peptides)
        ) -> temp3

    dplyr::left_join(temp2, temp3) %>%
        dplyr::mutate(Z = (Observed - SamplingAvg) / SD) %>%
        dplyr::arrange(dplyr::desc(abs(Z))) %>%
        dplyr::filter(!Kin %in% c("BARK1", "VRK2")) %>%
        dplyr::select(Kin, Observed, SamplingAvg, SD, Z) %>%
        dplyr::rename(Kinase = Kin) -> fin

    if (return_count == T) {
        return(list(count_mtx = temp, KRSA_Table = fin))
    } else {
        fin
    }
}