Windowed sequencing depth across the genome

Bins the genome into non-overlapping windows and computes the average (or median) sequencing depth in each window for each sample. Returns a tidy data.frame suitable for plotting or downstream QC analysis.

Usage

coverageDepth(
  object,
  window,
  method = c("mean", "median"),
  log2_transform = FALSE
)

Arguments

object

A commaData object.

window

Positive integer. Window size in base pairs.

method

Character string. Aggregation method within each window. One of "mean" (default) or "median".

log2_transform

Logical. If TRUE, the depth values are log2-transformed (using \(log2(depth + 1)\) to handle zeros). Default: FALSE.

Value

A data.frame with one row per (chromosome window, sample), containing:

chrom

Chromosome name.

window_start

First base of the window (1-based).

window_end

Last base of the window (1-based).

sample_name

Sample identifier.

depth

Mean or median sequencing depth in the window.

log2_depth

Log2-transformed depth (only present if log2_transform = TRUE).

Details

Depth is computed only at positions with observed methylation sites. Windows with no sites have depth = NA.

If genome size information is stored in genome(object), windows are sized to fit the chromosomes exactly (the last window may be smaller than window). If genome information is absent, only the range spanned by observed sites is covered.

See also

varianceByDepth, methylomeSummary

Examples

data(comma_example_data)
cd <- coverageDepth(comma_example_data, window = 10000L)
head(cd)
#>     chrom window_start window_end sample_name    depth
#> 1 chr_sim            1      10000      ctrl_1 63.00000
#> 2 chr_sim        10001      20000      ctrl_1 72.47826
#> 3 chr_sim        20001      30000      ctrl_1 67.16129
#> 4 chr_sim        30001      40000      ctrl_1 90.48148
#> 5 chr_sim        40001      50000      ctrl_1 75.58621
#> 6 chr_sim        50001      60000      ctrl_1 92.30769