Build a binned-scored GRanges object from one or many bigWig files. The aggregating function per bin can be min, max, sd, mean.
Usage
bw_bins(
bwfiles,
bg_bwfiles = NULL,
labels = NULL,
per_locus_stat = "mean",
bin_size = 10000,
genome = "mm9",
selection = NULL,
norm_mode = "fc",
remove_top = 0,
default_na = NA_real_,
scaling = "none",
canonical = FALSE
)Arguments
- bwfiles
Path or array of paths to the bigWig files to be summarized.
- bg_bwfiles
Path or array of paths to the bigWig files to be used as background.
- labels
List of names to give to the mcols of the returned GRanges object. If NULL, file names are used.
- per_locus_stat
Aggregating function (per locus). Mean by default. Choices: min, max, sd, mean.
- bin_size
Bin size.
- genome
Genome. Available choices are mm9, hg38.
- selection
A GRanges object to restrict binning to a certain set of intervals. This is useful for debugging and improving performance of locus specific analyses.
- norm_mode
Function to apply to normalize bin values. Default fc: divides bw / bg. Alternative: log2fc: returns log2(bw/bg).
- remove_top
Return range 0-(1-remove_top). By default returns the whole distribution (remove_top == 0).
- default_na
Default value for missing values
- scaling
If none, no operation is performed (default). If relative, values are divided by global mean (1x genome coverage).
- canonical
Use only canonical chromosomes (default: FALSE)
Value
A GRanges object with each bwfile as a metadata column named after labels, if provided, or after filenames otherwise.
Details
bwfiles and bg_bwfiles must have the same length. If you are using same background for several files, then file paths must be repeated accordingly.
norm_mode can be either "fc", where values will represent bigwig / bg_bigwig, or "log2fc", values will represent log2(bigwig / bg_bigwig) per bin.
Examples
# Get the raw files
bw <- system.file("extdata", "sample_H33_ChIP.bw", package="wigglescout")
bw2 <- system.file("extdata",
"sample_H3K9me3_ChIP.bw", package="wigglescout")
bw_inp <- system.file("extdata", "sample_Input.bw", package="wigglescout")
# Sample bigWig files only have valid values on this region
locus <- GenomicRanges::GRanges(
seqnames = "chr15",
IRanges::IRanges(102600000, 103100000)
)
# Run single bw. The larger the bin size, the faster this is.
bw_bins(bw, bin_size = 100000, labels = c("H33"), selection = locus)
#> Warning: Not all bigWig files or the genome tiling provided share the same sequence info.
#> Common to all (1): chr15
#> Missing in some of the files: chr1, chr2, chr3, chr4, chr5 ...
#> This can be due to different versions of the same reference genome, or to completely different organisms. Make sure these match!
#> GRanges object with 6 ranges and 1 metadata column:
#> seqnames ranges strand | H33
#> <Rle> <IRanges> <Rle> | <numeric>
#> [1] chr15 102500001-102600000 * | 1.343961
#> [2] chr15 102600001-102700000 * | 0.934239
#> [3] chr15 102700001-102800000 * | 1.956845
#> [4] chr15 102800001-102900000 * | 2.130207
#> [5] chr15 102900001-103000000 * | 2.010765
#> [6] chr15 103000001-103100000 * | 3.355660
#> -------
#> seqinfo: 35 sequences from an unspecified genome; no seqlengths
# Use of some parameters
bw_bins(bw, bg_bwfiles = bw_inp, bin_size = 100000,
labels = c("H33"), norm_mode = "log2fc", selection = locus)
#> Warning: Not all bigWig files or the genome tiling provided share the same sequence info.
#> Common to all (1): chr15
#> Missing in some of the files: chr1, chr2, chr3, chr4, chr5 ...
#> This can be due to different versions of the same reference genome, or to completely different organisms. Make sure these match!
#> GRanges object with 6 ranges and 1 metadata column:
#> seqnames ranges strand | H33
#> <Rle> <IRanges> <Rle> | <numeric>
#> [1] chr15 102500001-102600000 * | 0.384133
#> [2] chr15 102600001-102700000 * | -0.160072
#> [3] chr15 102700001-102800000 * | 0.912455
#> [4] chr15 102800001-102900000 * | 1.014908
#> [5] chr15 102900001-103000000 * | 0.781969
#> [6] chr15 103000001-103100000 * | 1.590060
#> -------
#> seqinfo: 35 sequences from an unspecified genome; no seqlengths
# Multiple bigWig
bw_bins(c(bw, bw2),
bin_size = 100000,
bg_bwfiles = c(bw_inp, bw_inp),
norm_mode = "log2fc",
selection = locus)
#> Warning: Not all bigWig files or the genome tiling provided share the same sequence info.
#> Common to all (1): chr15
#> Missing in some of the files: chr1, chr2, chr3, chr4, chr5 ...
#> This can be due to different versions of the same reference genome, or to completely different organisms. Make sure these match!
#> GRanges object with 6 ranges and 2 metadata columns:
#> seqnames ranges strand | sample_H33_ChIP sample_H3K9me3_ChIP
#> <Rle> <IRanges> <Rle> | <numeric> <numeric>
#> [1] chr15 102500001-102600000 * | 0.384133 -1.50269
#> [2] chr15 102600001-102700000 * | -0.160072 -1.32163
#> [3] chr15 102700001-102800000 * | 0.912455 -2.07487
#> [4] chr15 102800001-102900000 * | 1.014908 -2.66062
#> [5] chr15 102900001-103000000 * | 0.781969 -2.39259
#> [6] chr15 103000001-103100000 * | 1.590060 -3.99871
#> -------
#> seqinfo: 35 sequences from an unspecified genome; no seqlengths