R/mergeSCEColData.R

In compbiomed/singleCellTK: Comprehensive and Interactive Analysis of Single Cell RNA-Seq Data

#' @title Merging colData from two singleCellExperiment objects
#' @description Merges colData of the singleCellExperiment objects
#'  obtained from the same dataset which contain differing colData.
#'  (i.e. raw data and filtered data)
#' @param inSCE1 Input SingleCellExperiment object. The function will output this
#'  singleCellExperiment object with a combined colData from inSCE1 and inSCE2.
#' @param inSCE2 Input SingleCellExperiment object. colData from this object
#'  will be merged with colData from inSCE1 and loaded into inSCE1.
#' @param id1 Character vector. Column in colData of inSCE1 that will be
#'  used to combine inSCE1 and inSCE2. Default "column_name"
#' @param id2 Character vector. Column in colData of inSCE2 that will be
#'  used to combine inSCE1 and inSCE2. Default "column_name"
#' @return SingleCellExperiment object containing combined colData from
#'  both singleCellExperiment for samples in inSCE1.
#' @examples
#' sce1 <- importCellRanger(
#'     cellRangerDirs = system.file("extdata/", package = "singleCellTK"),
#'     sampleDirs = "hgmm_1k_v3_20x20",
#'     sampleNames = "hgmm1kv3",
#'     dataType = "filtered")
#' data(scExample)
#' sce2 <- sce
#' sce <- mergeSCEColData(inSCE1 = sce1, inSCE2 = sce2, id1 = "column_name", id2 = "column_name")
#' @export
mergeSCEColData <- function(inSCE1, inSCE2, id1 = "column_name", id2 = "column_name") {
  # For default case & column_name not in colData
  if (id1 == "column_name" && "column_name" %in%
    names(SummarizedExperiment::colData(inSCE1))) {
    columnNameColExist1 <- TRUE
  } else {
    columnNameColExist1 <- FALSE
    SummarizedExperiment::colData(inSCE1)$column_name <- colnames(inSCE1)
  }

  if (id2 == "column_name" && "column_name" %in%
    names(SummarizedExperiment::colData(inSCE2))) {
  } else {
    SummarizedExperiment::colData(inSCE2)$column_name <- colnames(inSCE2)
  }
  not.in.sce1 <- c(setdiff(
    names(SummarizedExperiment::colData(inSCE2)),
    names(SummarizedExperiment::colData(inSCE1))
  ), id2)
  not.in.sce1 <- not.in.sce1[!is.null(not.in.sce1)]

  coldata.not.in.sce1 <- SummarizedExperiment::colData(inSCE2)[, c(not.in.sce1),
    drop = FALSE
  ]

  coldata.sce1 <- SummarizedExperiment::colData(inSCE1)

  if (is.null(id1) | is.null(id2)) {
    if (is.null(rownames(coldata.not.in.sce1))) {
      stop("Unable to match between singleCellExperiment objects.
              Please define id1/id2 within the function,
              or assign a column name for the singleCellExperiment object.")
    } else {
      coldata.not.in.sce1$cell <- rownames(coldata.not.in.sce1)
      coldata.sce1$cell <- rownames(SummarizedExperiment::colData(inSCE1))
      id1 <- "cell"
      id2 <- "cell"
      placeholder <- TRUE
    }
  } else {
    placeholder <- FALSE
  }

  coldata.merge <- base::merge(coldata.sce1,
    coldata.not.in.sce1,
    all.x = TRUE,
    sort = FALSE,
    by.x = id1,
    by.y = id2
  )

  coldata.merge <- coldata.merge[match(
    colnames(SingleCellExperiment::counts(inSCE1)),
    coldata.merge[, id1]
  ), ]

  rownames(coldata.merge) <- coldata.merge[, id1]

  if (placeholder == TRUE) {
    coldata.merge[, id1] <- NULL
  }

  if(!columnNameColExist1){
    coldata.merge$column_name <- NULL
  }
  SummarizedExperiment::colData(inSCE1) <- S4Vectors::DataFrame(coldata.merge)
  return(inSCE1)
}