R/CellType_Composition_Heatmap.R

In qiuyugong-aifi/QCBATCH: This packages is used to perform QC of bacth effect

#' Title
#'
#' @param combined_dataset
#' @param cell_type
#' @param group
#'
#' @return
#' @export
#'
#' @examples
CellType_Composition_Heatmap <- function(combined_dataset, cell_type = "seurat_pbmc_type", group = "pbmc_sample_id", Sort = "False") {
  if (typeof(combined_dataset) == "S4") {
    combined_dataframe <- combined_dataset[[]]
  } else {
    combined_dataframe <- combined_dataset
  }

  cell_composition <- combined_dataframe %>%
    dplyr::select((!!as.name(cell_type)), (!!as.name(group))) %>%
    dplyr::group_by((!!as.name(group)), (!!as.name(cell_type))) %>%
    dplyr::tally() %>%
    dplyr::mutate(percent = n / sum(n)) %>%
    dplyr::select((!!as.name(group)), (!!as.name(cell_type)), percent)

  cell_composition <- data.frame(spread(cell_composition, key = (!!as.name(cell_type)), value = percent))

  if (Sort == "True") {
    cell_composition$visit <- substr(cell_composition$Donor_Visit_Detail, 8, 40)
    cell_composition <- cell_composition[order(cell_composition$visit), ]
    cell_composition <- cell_composition[1:(length(cell_composition) - 1)]
  }
  rownames(cell_composition) <- cell_composition %>%
    dplyr::select((!!as.name(group))) %>%
    pull()
  cell_composition <- cell_composition[, 2:ncol(cell_composition)]
  cell_composition[is.na(cell_composition)] <- 0

  fig <- plot_ly(
    x = colnames(cell_composition), y = rownames(cell_composition),
    z = as.matrix(cell_composition), type = "heatmap", width = 1000, height = length(rownames(cell_composition)) * 20
  )

  return(fig)
}