sgedgesByGene-methods: Extract the edges and their ranges from a SplicingGraphs...

In Bioconductor/SplicingGraphs: Create, manipulate, visualize splicing graphs, and assign RNA-seq reads to them

Extract the edges and their ranges from a SplicingGraphs object

Description

sgedgesByGene and sgedgesByTranscript both extract the edges and their ranges of all the genes from a SplicingGraphs object. They return them in a GRangesList object named with the gene ids, and where the items are grouped by gene (for sgedgesByGene) or by transcript (for sgedgesByTranscript).

Alternatively, intronsByTranscript extracts the intronic edges and their ranges of all the genes from a SplicingGraphs object. It returns them in a GRangesList object named with the gene ids, and where the items are grouped by transcript.

Usage

sgedgesByGene(x, with.exon.mcols=FALSE, with.hits.mcols=FALSE,
                 keep.dup.edges=FALSE)

sgedgesByTranscript(x, with.exon.mcols=FALSE, with.hits.mcols=FALSE)

## S4 method for signature 'SplicingGraphs'
intronsByTranscript(x)

Arguments

x	A SplicingGraphs object.
with.exon.mcols	Whether or not to include the exon metadata columns in the returned object. Those columns are named: exon_id, exon_name, exon_rank, start_SSid, and end_SSid. They are set to NA for edges of type intron.
with.hits.mcols	Whether or not to include the hits metadata columns in the returned object. See ?countReads for more information.
keep.dup.edges	If FALSE (the default), then within each group of the returned object, edges with the same global edge id are merged into a single element. Use keep.dup.edges=TRUE if this merging should not be performed.

Value

A GRangesList object named with the gene ids and where the items are grouped by gene (for sgedgesByGene), or by transcript (for sgedgesByTranscript and intronsByTranscript). In the latter case (i.e. grouping by transcript), the names are not unique.

The items that are being grouped are the splicing graph edges of type exon and intron (no artificial edges) for sgedgesByGene and sgedgesByTranscript, and the introns for intronsByTranscript.

When the grouping is by transcript (i.e. for sgedgesByTranscript and intronsByTranscript, items are ordered by their position from 5' to 3'.

About duplicated edges: A given edge can typically be shared by more than 1 transcript within the same gene, therefore sgedgesByTranscript typically returns an object where the same global edge id shows up in more than 1 group. However, the same global edge id is never shared across genes. By default sgedgesByGene removes duplicated edges, unless keep.dup.edges=TRUE is used.

Author(s)

H. Pagès

See Also

This man page is part of the SplicingGraphs package. Please see ?`SplicingGraphs-package` for an overview of the package and for an index of its man pages.

Examples

## ---------------------------------------------------------------------
## 1. Make SplicingGraphs object 'sg' from toy gene model (see
##    '?SplicingGraphs')
## ---------------------------------------------------------------------
example(SplicingGraphs)
sg

## 'sg' has 1 element per gene and 'names(sg)' gives the gene ids.
names(sg)

## ---------------------------------------------------------------------
## 2. sgedgesByGene()
## ---------------------------------------------------------------------
edges_by_gene <- sgedgesByGene(sg)
edges_by_gene
## 'edges_by_gene' has the length and names of 'sg', that is, the names
## on it are the gene ids and are guaranteed to be unique.

## Extract the edges and their ranges for a given gene:
edges_by_gene[["geneB"]]
## Note that edge with global edge id "geneB:3,4" is an intron that
## belongs to transcripts B1 and B2.

edges_by_gene0 <- sgedgesByGene(sg, keep.dup.edges=TRUE)
edges_by_gene0[["geneB"]]
## Note that edge "geneB:3,4" now shows up twice, once for transcript
## B1, and once for transcript B2.

## Keep the "exon metadata columns":
sgedgesByGene(sg, with.exon.mcols=TRUE)
## Note that those cols are set to NA for intronic edges.

## ---------------------------------------------------------------------
## 3. sgedgesByTranscript()
## ---------------------------------------------------------------------
edges_by_tx <- sgedgesByTranscript(sg)
edges_by_tx

## 'edges_by_tx' is typically longer than 'sg'.
## IMPORTANT NOTE: One caveat here is that the names on 'edges_by_tx'
## are the gene ids, not the transcript ids, and thus are typically NOT
## unique!

## Select elements of a given gene:
edges_by_tx["geneB"]  # not a good idea
edges_by_tx[names(edges_by_tx) %in% "geneB"]  # much better :-)
## Note that edge with global edge id "geneB:3,4" is an intron that
## belongs to transcripts B1 and B2.

## Keep the "exon metadata columns":
sgedgesByTranscript(sg, with.exon.mcols=TRUE)
## Note that those cols are set to NA for intronic edges.

## ---------------------------------------------------------------------
## 4. intronsByTranscript()
## ---------------------------------------------------------------------
in_by_tx <- intronsByTranscript(sg)
in_by_tx

## 'in_by_tx' has the length and names of 'edges_by_tx'. The same
## recommendation applies for selecting elements of a given set of
## genes:
in_by_tx[c("geneB", "geneD")]  # not a good idea
in_by_tx[names(in_by_tx) %in% c("geneB", "geneD")]  # much better :-)

## ---------------------------------------------------------------------
## 5. Comparing the outputs of unlist(), intronsByTranscript(), and
##    sgedgesByTranscript()
## ---------------------------------------------------------------------
ex_by_tx <- unlist(sg)
in_by_tx <- intronsByTranscript(sg)
edges_by_tx <- sgedgesByTranscript(sg)

## A sanity check:
stopifnot(identical(elementNROWS(in_by_tx) + 1L,
                    elementNROWS(ex_by_tx)))

## 'edges_by_tx' combines 'ex_by_tx' and 'in_by_tx' in a single
## GRangesList object. Sanity check:
stopifnot(identical(elementNROWS(edges_by_tx),
                    elementNROWS(ex_by_tx) + elementNROWS(in_by_tx)))

Bioconductor/SplicingGraphs documentation built on Nov. 3, 2024, 8:11 a.m.