Data Engineering and BI courses are free this week!
## duplicated()
## ------------
"duplicated"(x, incomparables=FALSE, fromLast=FALSE, method=c("auto", "quick", "hash"))
## match()
## -------
"match"(x, table, nomatch=NA_integer_, incomparables=NULL, method=c("auto", "quick", "hash"), ignore.strand=FALSE)
## order() and related methods
## ----------------------------
"order"(..., na.last=TRUE, decreasing=FALSE)
"sort"(x, decreasing=FALSE, ignore.strand=FALSE, by)
"rank"(x, na.last=TRUE, ties.method=c("average", "first", "random", "max", "min"))
## Generalized element-wise (aka "parallel") comparison of 2 GenomicRanges
## objects
## ------------------------------------------------------------------------
"compare"(x, y)?`Ranges-comparison` in the IRanges
package for a description of these arguments.
TRUE or FALSE.
"first"
is supported for now.
as.env(x);
the resulting variables are passed to order to generate the
ordering permutation.
duplicated() and unique()
on a GenomicRanges object are conforming to this. The "natural order" for the elements of a GenomicRanges object is to
order them (a) first by sequence level, (b) then by strand, (c) then by
start, (d) and finally by width.
This way, the space of genomic ranges is totally ordered.
Note that the reduce method for GenomicRanges uses this
"natural order" implicitly. Also, note that, because we already do (c)
and (d) for regular ranges (see ?`Ranges-comparison`),
genomic ranges that belong to the same underlying sequence and strand are
ordered like regular ranges.
order(), sort(), and rank() on a GenomicRanges
object are using this "natural order".
Also ==, !=, <=< code="">, >=, < and >
on GenomicRanges objects are using this "natural order".
gr0 <- GRanges(
seqnames=Rle(c("chr1", "chr2", "chr1", "chr3"), c(1, 3, 2, 4)),
ranges=IRanges(c(1:9,7L), end=10),
strand=Rle(strand(c("-", "+", "*", "+", "-")), c(1, 2, 2, 3, 2)),
seqlengths=c(chr1=11, chr2=12, chr3=13))
gr <- c(gr0, gr0[7:3])
names(gr) <- LETTERS[seq_along(gr)]
## ---------------------------------------------------------------------
## A. ELEMENT-WISE (AKA "PARALLEL") COMPARISON OF 2 GenomicRanges OBJECTS
## ---------------------------------------------------------------------
gr[2] == gr[2] # TRUE
gr[2] == gr[5] # FALSE
gr == gr[4]
gr >= gr[3]
## ---------------------------------------------------------------------
## B. duplicated(), unique()
## ---------------------------------------------------------------------
duplicated(gr)
unique(gr)
## ---------------------------------------------------------------------
## C. match(), %in%
## ---------------------------------------------------------------------
table <- gr[1:7]
match(gr, table)
match(gr, table, ignore.strand=TRUE)
gr %in% table
## ---------------------------------------------------------------------
## D. findMatches(), countMatches()
## ---------------------------------------------------------------------
findMatches(gr, table)
countMatches(gr, table)
findMatches(gr, table, ignore.strand=TRUE)
countMatches(gr, table, ignore.strand=TRUE)
gr_levels <- unique(gr)
countMatches(gr_levels, gr)
## ---------------------------------------------------------------------
## E. order() AND RELATED METHODS
## ---------------------------------------------------------------------
order(gr)
sort(gr)
sort(gr, ignore.strand=TRUE)
## TODO: Broken. Please fix!
#sort(gr, by = ~ seqnames + start + end) # equivalent to (but slower than) above
score(gr) <- rev(seq_len(length(gr)))
## TODO: Broken. Please fix!
#sort(gr, by = ~ score)
rank(gr)
## ---------------------------------------------------------------------
## F. GENERALIZED ELEMENT-WISE COMPARISON OF 2 GenomicRanges OBJECTS
## ---------------------------------------------------------------------
gr2 <- GRanges(c(rep("chr1", 12), "chr2"), IRanges(c(1:11, 6:7), width=3))
strand(gr2)[12] <- "+"
gr3 <- GRanges("chr1", IRanges(5, 9))
compare(gr2, gr3)
rangeComparisonCodeToLetter(compare(gr2, gr3))
Run the code above in your browser using DataLab