Counts reads overlapping genomic features. Uses
GenomicAlignments::summarizeOverlaps() if available, otherwise shells out
to featureCounts from the Subread package.
Usage
bb_count_reads(
bam_paths,
annotation,
method = c("auto", "internal", "featureCounts"),
threads = 4L,
feature_type = "exon",
attr_type = "gene_id",
paired = FALSE
)Arguments
- bam_paths
Character vector. Paths to BAM files. Optional vector names supply unique sample IDs when BAM basenames are not unique.
- annotation
Character. Path to GTF/GFF annotation file.
- method
Character.
"auto"tries GenomicAlignments first, then featureCounts;"internal"forces GenomicAlignments;"featureCounts"forces the external tool.- threads
Integer. Number of threads for featureCounts. Default
4.- feature_type
Character. Feature type to count (GTF column 3). Default
"exon".- attr_type
Character. Attribute to group by. Default
"gene_id".- paired
Logical. Count paired-end fragments. Default
FALSE. The featureCounts backend requires Subread 2.0.2 or later in paired mode.
Value
A numeric matrix of counts (genes x samples). Rownames are gene IDs, colnames are derived from BAM file names.
Examples
# \donttest{
# Internal counting needs the optional Bioconductor packages Rsamtools,
# GenomicAlignments, GenomicFeatures and txdbmaker.
needed <- c("Rsamtools", "GenomicAlignments", "GenomicRanges",
"GenomicFeatures", "txdbmaker", "SummarizedExperiment")
if (all(vapply(needed, requireNamespace, logical(1), quietly = TRUE))) {
dir <- tempfile("bambamR-count-")
dir.create(dir)
sam <- file.path(dir, "s1.sam")
writeLines(c(
"@HD\tVN:1.6\tSO:coordinate", "@SQ\tSN:chr1\tLN:2000",
"r1\t0\tchr1\t100\t60\t8M\t*\t0\t0\tACGTACGT\tIIIIIIII",
"r2\t0\tchr1\t300\t60\t8M\t*\t0\t0\tTTGCAACG\tIIIIIIII"
), sam)
bam <- Rsamtools::asBam(sam, destination = file.path(dir, "s1"))
gtf <- file.path(dir, "genes.gtf")
writeLines(c(
paste("chr1", "example", "exon", "90", "150", ".", "+", ".",
"gene_id \"G1\"; transcript_id \"T1\";", sep = "\t"),
paste("chr1", "example", "exon", "280", "350", ".", "+", ".",
"gene_id \"G2\"; transcript_id \"T2\";", sep = "\t")
), gtf)
bb_count_reads(bam, gtf, method = "internal")
unlink(dir, recursive = TRUE)
}
#> Counting reads with GenomicAlignments...
#> Import genomic features from the file as a GRanges object ...
#> OK
#> Prepare the 'metadata' data frame ...
#> OK
#> Make the TxDb object ...
#> Warning: genome version information is not available for this TxDb object
#> OK
# }