#' Import the quantification results of many RNAseq quantifiers, including 'alevin' for single-cell data.
#'
#' @param path
#' @param ...
#' Most likely the first step in your DTUrtle analysis.
#' Can perform multiple scaling schemes, defaults to scaling schemes appropriate for DTU analysis.
#'
#' @return
#' @param files Vector of files to be imported. Optionally can be named to keep the samples names.
#' @param type Type of the quantification data. All tools supported by tximport can be selected. If you have single-cell data, the use of `alevin` is proposed.
#' - `'salmon'`
#' - `'alevin'`
#' - `'kallisto'`
#' - `'rsem'`
#' - `'stringtie'`
#' - `'sailfish'`
#' - `'none'`
#' @param ... Further parameters to the specific tximport call. See `tximport()` for available parameters.
#'
#' @return - For bulk data: A combined count matrix for all specified samples.
#' - For single-cell data (`type='alevin'`): A list of count matrices per sample. Can be combined and optionally added to a Seurat object with TODO.
warning("\nImport of alevin files currently does not support using scaling methods.\nPlease note, that in tagged-end single-cell protocols (like 10X chromium) it is assumed\nthat there is no length effect in the fragment generation process - thus making a scaling unnecessary.")
warning("It is recommended to use the 'countsFromAbundance' scaling schemes 'dtuScaledTPM' or 'scaledTPM',\nto correct for increased counts of longer transcripts.\nIf you are using a tagged-end protocol, the use 'no' is suggested.")
}
}else{
if(hasArg("tx2gene")){
message("Using 'dtuScaledTPM' for 'countsFromAbundance'.")
args$countsFromAbundance<-"dtuScaledTPM"
}else{
message("Using 'scaledTPM' for 'countsFromAbundance'. If you specify a 'tx2gene' file, 'dtuScaledTPM' can be used.")
args$countsFromAbundance<-"scaledTPM"
}
}
if(hasArg(txOut)){
if(args$txOut==F){
warning("Unless you exactly know what you are doing it is not recommended to set txOut to FALSE.\nDownstream analysis may fail!")
}
}else{
args$txOut<-T
}
args$files<-files
args$type<-type
return(do.call(tximport,args)$counts)
}
}
#TODO: Check if seurat_obj is seurat object!
#' Title
#TODO: combine to matrix without Seurat! (same extension scheme)
#' Combine list of sparse transcription count matrices.
#'
#' @param tx_mat
#' @param seurat_obj
#' Only needed when dealing with single-cell data. Adds a cellname extension if necessary.
#' Can optionally add the combined matrix to a existing Seuratobject (keeping the cellname extension of the object and matching the cells).
#'
#' @return
#' @param tx_list List of sparse transcription count matrices, as returned by `import_counts()` for single-cell data.
#' @param cell_extensions Optional list of cellname extensions that are added to the cellnames of one sample. The cellnames and the extension are separated by an underscore '_'.
#' @param seurat_obj Optional seurat object, where the combined matrix is added as an assay. This has the advantage, that the cells are matched and subsetted if necessary. Currently only Seurat 3 objects are supported.
#' @param assay_name If the combined matrix should be added to an existing Seurat object, the name of the assay can be specified here.
#'
#' @return Either a combined sparse transcription count matrix or a seurat object which the combined sparse transcription count matrix as an assay.
stopstr<-ifelse(length(dup_cells)>10,paste0("Found ",length(dup_cells)," duplicated cellnames even after cellname extension!"),paste0("Found duplicated cellnames even after cellname extension:\n\t",paste0(dup_cells,collapse="\n\t")))
message("Of ",ncol(tx_mat)," cells, ",length(common_cells)," (",round(length(common_cells)/ncol(tx_mat)*100),"%) could be found in the provided seurat object.")
message(length(uniq_cells)," (",round(length(uniq_cells)/ncol(tx_mat)*100),"%) are unique to the tx files.")
message("Of ",ncol(tx_list)," cells, ",length(common_cells)," (",round(length(common_cells)/ncol(tx_list)*100),"%) could be found in the provided seurat object.")
message(length(uniq_cells)," (",round(length(uniq_cells)/ncol(tx_list)*100),"%) are unique to the tx files.")
#' @param n_core Set the number of CPU cores that should be used in the statistical computations.
#' @param ... If `filtering_strategy=='own'` specify the wished parameters of `dmFilter()` here.
#'
#' @return `DRIM` object with the key results, that can be used in the DTUrtle steps hereafter. The object is just a easily accesible list with the following items:
#' @return `drim` object with the key results, that can be used in the DTUrtle steps hereafter. The object is just a easily accesible list with the following items:
#' - `drim`: The results of the DRIMSeq statistical computations (`dmTest()`).
#' - `design_full`: The design matrix generated from the specified `pd` columns
#' - `design_full`: The design matrix generated from the specified `pd` columns.
#' - `cond_levels`: A copy of the specified comparison groups.
#' - `group`: ?
#' - `group`: Vector which sample/cell belongs to which comparison group.
#' - `pct_exp_tx`: Data frame of the expressed-in percentage of all transcripts. [TODO: split by group?]
#' - `pct_exp_gene`: Data frame of the expressed-in percentage of all genes. [TODO: split by group?]
