Merge pull request #72 from yerkes-gencore/writeDGEResults

Write dge results

DESCRIPTION

@@ -22,6 +22,7 @@ Imports:
     magrittr,
     Matrix,
     methods,
+    openxlsx,
     plotly,
     plyr,
     reshape2,

NAMESPACE

@@ -42,13 +42,15 @@ export(runSoupX)
 export(run_scDblFinder)
 export(scoreCC)
 export(update_sc)
+export(writeDGEResults)
 import(DoubletFinder)
 import(Seurat)
 import(SingleR)
 import(SoupX)
 import(dplyr)
 import(ggplot2)
 import(grDevices)
+import(openxlsx)
 import(scDblFinder)
 importFrom(Matrix,rowSums)
 importFrom(S4Vectors,mcols)
@@ -78,6 +80,7 @@ importFrom(ggplot2,stat_smooth)
 importFrom(ggplot2,theme)
 importFrom(ggplot2,theme_classic)
 importFrom(gtools,mixedsort)
+importFrom(here,here)
 importFrom(magrittr,"%>%")
 importFrom(methods,show)
 importFrom(plotly,ggplotly)

R/writeDGEResults.R

@@ -0,0 +1,77 @@
+#' Write differential expression test results to excel sheet
+#'
+#' Take a list of differential expression results tables returned from `Seurat::FindMarkers()`
+#'  or a similar function and write them to an excel worksheet, with one sheet for each result.
+#'
+#' @param results List of outputs from `Seurat::FindMarkers()` or a similar function
+#' @param sheet_names List of names for sheets, defaults to names of `results`
+#' @param output_name Name of file
+#' @param outdir Location of file
+#' @param p_val_colname Name of column with adjusted P value data
+#' @param drop_NA Exclude values with NA for adjusted P value?
+#' @param write_rownames Write rownames of `results` to sheet?
+#'
+#' @return `NULL`
+#' @export
+#'
+#' @import openxlsx
+#' @importFrom here here
+#'
+#' @examples
+#' \dontrun{
+#' dge_results$NK_adaptive.post.1D3_vs_cnt <- FindMarkers(object = cd8_obj,
+#'                                                        group.by = 'condition.timepoint',
+#'                                                        ident.1 = 'post-ATI.1D3', ident.2 = 'post-ATI.control',
+#'                                                        subset.ident = 'NK (adaptive)',
+#'                                                        features = dge_genes,
+#'                                                        logfc.threshold = 0, min.pct = 0,
+#'                                                        test.use = 'MAST', assay = 'RNA')
+#' dge_results$NK_adaptive.pre.1D3_vs_cnt <- FindMarkers(object = cd8_obj,
+#'                                                       group.by = 'condition.timepoint',
+#'                                                       ident.1 = 'pre-ATI.1D3', ident.2 = 'pre-ATI.control',
+#'                                                       subset.ident = 'NK (adaptive)',
+#'                                                       features = dge_genes,
+#'                                                       logfc.threshold = 0, min.pct = 0,
+#'                                                       test.use = 'MAST', assay = 'RNA')
+#' writeDGEResults(dge_results)
+#' }
+writeDGEResults <- function(results,
+                            sheet_names = names(results),
+                            output_name = paste0("differentially_expressed_genes.xlsx"),
+                            outdir = here::here('outputs'),
+                            p_val_colname = 'p_val_adj',
+                            drop_NA = FALSE,
+                            write_rownames = TRUE){
+  outfile <- file.path(outdir, output_name)
+  message(paste0('Writing results to ', outfile))
+  wb <- openxlsx::createWorkbook('ENPRC Gencore')
+  mapply(FUN = .addWorksheet_DGEres,
+         result = results,
+         sheet_name = sheet_names,
+         MoreArgs = list(wb = wb,
+                         drop_NA = drop_NA,
+                         p_val_colname = p_val_colname,
+                         write_rownames = write_rownames)
+  )
+  openxlsx::saveWorkbook(wb, outfile, overwrite = TRUE)
+}
+
+.addWorksheet_DGEres <- function(wb,
+                                 result,
+                                 sheet_name,
+                                 drop_NA,
+                                 p_val_colname,
+                                 write_rownames){
+  if (nchar(sheet_name) > 31) {
+    sheet_name <- substr(sheet_name, 1, 31)
+  }
+  if (drop_NA) {
+    result <- result[!is.na(result[[pval_col_name]]),]
+  }
+  result <- result[order(result[[pval_col_name]]),]
+  openxlsx::addWorksheet(wb, sheet_name)
+  openxlsx::writeData(wb,
+                      sheet = sheet_name,
+                      x = as.data.frame(result),
+                      rowNames = write_rownames)
+}