Split DRAGEN coverage-bin metric names into numeric bounds

Parses the coverage-bin rows of a DRAGEN *_coverage_metrics.csv (PCT of <region> with coverage [lo: hi)) into their numeric depth bounds. The open, cumulative bins ([Nx: inf)) return NA for the upper bound, which distinguishes them from the bucketed bins ([lo: hi)). DragenCov uses this NA to partition the bins into the metricsbins (bucketed) and metricscumu (cumulative) long tables.

Usage

dragen_cov_bin_split(v)

Arguments

v

(character())
Vector of raw coverage-bin metric names.

Value

(tibble) with integer columns cov_lo and cov_hi (cov_hi is NA for the open inf upper bound).

Examples

dragen_cov_bin_split(c(
  "PCT of genome with coverage [ 100x: inf)",
  "PCT of QC coverage region with coverage [  20x:  50x)"
))
#> # A tibble: 2 × 2
#>   cov_lo cov_hi
#>    <int>  <int>
#> 1    100     NA
#> 2     20     50