Restore mirToRanges functionality

DESCRIPTION

@@ -52,6 +52,7 @@ Suggests:
     impute,
     knitr,
     magrittr,
+    miRNAmeConverter,
     org.Hs.eg.db,
     RColorBrewer,
     readr,

NAMESPACE

@@ -34,6 +34,7 @@ export(trimColData)
 export(uniformBuilds)
 import(methods)
 importFrom(BiocBaseUtils,checkInstalled)
+importFrom(BiocBaseUtils,isCharacter)
 importFrom(BiocBaseUtils,isScalarCharacter)
 importFrom(BiocBaseUtils,isScalarNumber)
 importFrom(BiocBaseUtils,lifeCycle)

R/simplifyTCGA.R

@@ -1,7 +1,7 @@
 #' @importFrom GenomicFeatures genes
 #' @importFrom GenomeInfoDb keepStandardChromosomes seqlevelsStyle
 #' seqlevelsStyle<-
-#' @importFrom BiocBaseUtils isScalarCharacter
+#' @importFrom BiocBaseUtils isScalarCharacter isCharacter
 NULL
 
 .checkHas <-
@@ -30,17 +30,23 @@ NULL
 
 .convertTo <- function(x, which, FUN, keep, unmap) {
     for (i in which(which)) {
-        lookup <- FUN(rownames(x[[i]]))
+        assay <- x[[i]]
+        lookup <- FUN(rownames(assay))
         ranges <- lookup[["mapped"]]
-        rse <- x[[i]][names(ranges), ]
+        rowidx <- mcols(ranges)[["rowIdx"]]
+        rowidx <- Filter(Negate(is.na), rowidx)
+        if (!is.null(rowidx) && length(rowidx))
+            rse <- `rownames<-`(assay[rowidx, ], names(ranges))
+        else
+            rse <- assay[names(ranges), ]
         # rowData not merged with mcols of RHS in `rowRanges<-` method
         mcols(ranges) <-
             S4Vectors::DataFrame(rowData(rse), S4Vectors::mcols(ranges))
         SummarizedExperiment::rowRanges(rse) <- ranges
         x <- c(x, setNames(S4Vectors::List(rse),
             paste0(names(x)[i], "_ranged")))
         if (length(lookup[["unmapped"]]) && unmap) {
-            se <- x[[i]][lookup[["unmapped"]], ]
+            se <- assay[lookup[["unmapped"]], ]
             x <- c(x, setNames(S4Vectors::List(se),
                 paste0(names(x)[i], "_unranged")))
         }
@@ -67,6 +73,25 @@ NULL
     res
 }
 
+#' @name hidden-helpers
+#' @keywords internal
+.makeMiRNAListRanges <- function(x, gn) {
+    checkInstalled("miRNAmeConverter")
+    nc <- miRNAmeConverter::MiRNANameConverter()
+    mirna_version <-
+        miRNAmeConverter::assessVersion(nc, names(gn))[1L, "version"]
+    trout <- miRNAmeConverter::translateMiRNAName(
+        nc, x, versions = mirna_version
+    )
+    new_x <- trout[[paste0("v", mirna_version, ".0")]]
+    res <- list(unmapped = setdiff(x, trout[["input"]]))
+    rowIdx <- match(tolower(trout[["input"]]), tolower(x))
+    gn <- gn[match(new_x, names(gn))]
+    mcols(gn)[["rowIdx"]] <- rowIdx
+    res[["mapped"]] <- gn
+    res
+}
+
 .getGN <- function(gen) {
     stopifnot(isScalarCharacter(gen))
 
@@ -92,9 +117,9 @@ NULL
 #' @rdname hidden-helpers
 #' @keywords internal
 .getRangesOfMir <- function(x) {
-    stopifnot(isScalarCharacter(x))
+    stopifnot(isCharacter(x))
 
-    mirnas_gr <- .get_hsa_gff3(x)
+    mirnas_gr <- .get_hsa_gff3("hg19")
 
     miR <- mirnas_gr[
         mcols(mirnas_gr)[["type"]] %in% c("miRNA", "microRNA", "tRNA")
@@ -103,7 +128,7 @@ NULL
     seqlevelsStyle(miR) <- "NCBI"
 
     names(miR) <- mcols(miR)[["Name"]]
-    .makeListRanges(x, miR)
+    .makeMiRNAListRanges(x, miR)
 }
 
 #' @rdname hidden-helpers
@@ -209,9 +234,7 @@ NULL
 #'   [`RangedSummarizedExperiment`][SummarizedExperiment::RangedSummarizedExperiment-class]
 #'   objects
 #'
-#' @author L. Waldron
-#'
-#' @md
+#' @author L. Waldron, M. Ramos
 #'
 #' @examples
 #'
@@ -275,10 +298,16 @@ symbolsToRanges <- function(obj, keep.assay = FALSE, unmapped = TRUE) {
 #'
 #' @export
 mirToRanges <- function(obj, keep.assay = FALSE, unmapped = TRUE) {
-    lifeCycle(
-        package = "TCGAutils",
-        cycle = "defunct",
-        title = "simplifyTCGA"
+    checkInstalled("Bioc.gff")
+
+    can.fix <- .isFixable(mae = obj, pattern = "^hsa")
+
+    .convertTo(
+        x = obj,
+        which = can.fix,
+        FUN = .getRangesOfMir,
+        keep = keep.assay,
+        unmap = unmapped
     )
 }
 

R/utils.R

@@ -45,7 +45,7 @@
     gff_lines <- readLines(file, n = 50)
     genome_line <- grepv("genome-build-id", gff_lines)
     gnm <- strsplit(genome_line, ":\\s+")[[1L]] |>
-        tail(n = 1L) |>
+        utils::tail(n = 1L) |>
         trimws()
     if (!length(gnm))
         NA_character_