Expand granges method

R/intra-range-methods.R

@@ -79,7 +79,7 @@ setMethod("flank", "GenomicRanges",
 
 
 ### - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-### promoters() and terminators()
+### promoters(), terminators() and expand()
 ###
 
 ### Returns an IRanges **instance**.
@@ -155,6 +155,41 @@ setMethod("terminators", "GenomicRanges",
                                  site="TES")
 )
 
+### Expand ranges in a GRanges object upstream and downstream by specified numbers of bases, taking account of strand.
+### Unstranded ranges are treated like they on the "+" strand
+.GenomicRanges_expand = function(x, upstream = 0, downstream = 0, trim_out_of_bounds = TRUE) {
+
+  # Check that upstream and downstream are vectors of either length 1 or with the same length as x
+  if(!length(upstream) %in% c(1, length(x))){
+    stop("upstream should be a vector of length 1 or the length of x")}
+  if(!length(downstream) %in% c(1, length(x))){
+    stop("downstream should be a vector of length 1 or the length of x")}
+
+  # Check if any regions would have negative widths after adjustment
+  if(any(width(x) + upstream + downstream < 0)){
+    stop("Some regions would have a negative width after adjustment. This is not permitted.")
+  }
+
+  # Shift genomic regions upstream either by upstream (strand is + or *) or downstream (strand is -) 
+  x <- GenomicRanges::shift(x, 
+    -ifelse(GenomicRanges::strand(x) == "-", downstream, upstream))
+
+  # Expand GRanges so that their width equals their original size plus upstream and downstream
+  x <- GenomicRanges::resize(x, 
+    width = (GenomicRanges::width(x) + upstream + downstream), fix = "start", ignore.strand = T)
+
+  # Remove any out-of-bounds regions and return x
+  if(trim_out_of_bounds){
+    x <- GenomicRanges::trim(x)
+  }
+  return(x)
+} 
+
+setMethod("expand", "GenomicRanges",
+    function(x, upstream=0, downstream=0, trim_out_of_bounds=TRUE)
+        .GenomicRanges_expand(x, upstream, downstream, trim_out_of_bounds)
+)
+
 
 ### - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
 ### reflect()

inst/unitTests/test_intra-range-methods.R

@@ -230,6 +230,34 @@ test_promoters_GenomicRanges <- function()
     checkIdentical(seqinfo(gr), seqinfo(current))
 }
 
+test_expand_GenomicRanges <- function()
+{
+
+    # Test zero-length GRanges
+    gr0 <- GRanges(tx_id=integer(0))
+    checkIdentical(expand(gr0, 5, 2), gr0)
+
+    # Test that giving 0 for upstream and downstream returns original GRanges
+    gr <- setNames(GRanges(c("chr1:100-200:+", "chr1:100-200:*", "chr1:100-200:-"), tx_id=seq_len(3)), letters[1:3])
+    checkIdentical(expand(gr, 0, 0), gr)
+
+    # Test that expanding works in strand-wise manner
+    target <- setNames(GRanges(c("chr1:0-250:+", "chr1:0-250:*", "chr1:50-300:-"), tx_id=seq_len(3)), letters[1:3])
+    checkIdentical(expand(gr, 100, 50), target)
+
+    # Test that regions are contracted when negative values given to upstream or downstream
+    target <- setNames(GRanges(c("chr1:150-175:+", "chr1:150-175:*", "chr1:125-150:-"), tx_id=seq_len(3)), letters[1:3])
+    checkIdentical(expand(gr, -50, -25), target)
+
+    # Test providing vectors with length > 1 for upstream and downstream
+    target <- setNames(GRanges(c("chr1:75-250:+", "chr1:0-225:*", "chr1:0-250:-"), tx_id=seq_len(3)), letters[1:3])
+    checkIdentical(expand(gr, c(25, 100, 50), c(50, 25, 100)), target)
+
+    # Test that an error is thrown when upstream or downstream ahve length > 1, but not equal to the length of gr
+    testthat::expect_error(expand(gr, c(25, 100), c(50, 25, 100)))
+    testthat::expect_error(expand(gr, c(25, 100, 50), c(25, 100)))
+}
+
 test_restrict_GenomicRanges <- function()
 {
     gr <-  make_test_GRanges()