v0.2.1

v0.2.1

• allow unsorted bed as input to --by
• allow setting precision with env var, e.g. MOSDEPTH_PRECISION=5
• moderate performance increase.
• build using holy-build box to better support systems with older libc.

thresholds argument and CRAM

0.2.0

• ~2X speed improvement for CRAM by not decoding unused base-qualities.
• add new --thresholds argument. See README for usage. thresholds and quantization are highly recommended over
• when using quantize, labels now indicate the range of depths encompassed by that region.
• for quantize and per-base, mosdepth will output all chromosomes in the bam header, even if they have no alignments.
  this makes it so that mosdepth output for files aligned to the same reference have the same number and order of chromosomes
  and same total base coverage
  per-base output when possible as they are more compact and faster to output.
• fix bug with dist output for chroms larger than 2.1 billion bases that also affected total output in some cases.

mosdepth prefix

This is a breaking release. A given output will be unchanged, but the command-line usage to get it will be different. I apologize for the change in use, but not for the increase in utility. In order to expose the utility of mosdepth across more (and multiple simultaneous) uses, mosdepth now requires a prefix for outputs rather than sending it to stdout. Usage as of this release looks as below. Where previous usage was e.g.

mosdepth $bam > $output

current usage becomes:

mosdepth $output_prefix $bam

This is explained in more detail in the help:

mosdepth 0.1.9

  Usage: mosdepth [options] <prefix> <BAM-or-CRAM>

Arguments:

  <prefix>       outputs: `{prefix}.mosdepth.dist.txt`
                          `{prefix}.per-base.bed.gz` (unless -n/--no-per-base is specified)
                          `{prefix}.regions.bed.gz` (if --by is specified)
                          `{prefix}.quantized.bed.gz` (if --quantize is specified)

  <BAM-or-CRAM>  the alignment file for which to calculate depth.

Common Options:
  
  -t --threads <threads>     number of BAM decompression threads [default: 0]
  -c --chrom <chrom>         chromosome to restrict depth calculation.
  -b --by <bed|window>       optional BED file or (integer) window-sizes.
  -n --no-per-base           dont output per-base depth (skipping this output will speed execution).
  -f --fasta <fasta>         fasta file for use with CRAM files.

Other options:

  -F --flag <FLAG>            exclude reads with any of the bits in FLAG set [default: 1796]
  -q --quantize <segments>    write quantized output, see docs for description.
  -Q --mapq <mapq>            mapping quality threshold [default: 0]
  -h --help                   show help

better support for beds with non-region lines

• handle track lines
• report message on bad bed lines (and skip).

NOTE that the default binary below requires a fairly modern version of GLIBC. If that fails for you with an error like:
/lib64/libc.so.6: version GLIBC_2.14' not found`

then use the mosdepth_old_glibc binary.

remove extra tab

when using --by, there was an extra tab between the chrom and start columns. This fixes that.

bugfixes related to `distribution` and large window sizes

• fixed when a window-size larger than the chromosome length was specified, windows would be output until overflow.
• limit size of dist file by only reporting novel values
• cap memory use even for very deep coverage.

thanks to @tomsasani for reporting several issues.

MT

fix off-by-1 error that affected MT chromosome.

options

see: #3 (comment)

• new -F flag allows setting exclude flag
• if --by argument is a bed file with >= 4 columns, the 4th column is transferred to the output.

usability improvements

linux binary below.

mosdepth inital release.

the binary below is for linux.