MullinsLab / dwindling

A tool to summarize the NGS read count stats at each stage of a pipeline

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Dwindling Reads

Usage:

dwindling-reads [--csv] --stage <stage1> <files> [--stage <stage2> <files> [--stage <stage3> <files>]
dwindling-reads --help

This program collates summary information from a set of next-gen sequencing paired-end reads at each stage of your pipeline.

Each stage is specified by --stage, followed by two to four arguments. The first argument is an arbitrary name of your choosing to identify that stage. Both FastQ (uncompressed and gzipped) and BAM files are supported. FastQ stages must specify at least two files for forward and reverse reads. An optional third file may be specified for single-end orphans, such as those commonly produced by quality trimmers. BAM stages must specify only a single .bam file.

CSV output may be piped into dwindling-plot to produce an SVG plot of the read counts through the various stages.

Options:

--csv     output CSV instead of formatted text
--help    print usage message and exit

Examples:

dwindling-reads --stage raw     raw/ABC_R1.fq.gz raw/ABC_R2.fq.gz \
                --stage sickle  trimmed/ABC_R1.fq trimmed/ABC_R2.fq trimmed/ABC_orphans.fq \
                --stage mapped  ABC.bam

Example output:

196300 raw = 98150 forward + 98150 reverse

161640 sickle = 80820 forward + 80820 reverse (= 82.34% of raw)
  6278 single-end orphans (= 3.20% of raw)
 34660 discarded, including orphans (= 17.66% of raw)

161910 mapped = 80820 forward + 80820 reverse (= 100.17% of sickle)
  -270 discarded (= -0.17% of sickle)

(In this case, bwa mem found 270 reads with secondary mappings so the reads are growing not dwindling!)

dwindling-plot example

plot example

Installation

dwindling-reads

Requires:

  • Perl 5.14 or newer
  • zcat (distributed with gzip)
  • wc (distributed with coreutils)
  • samtools

The first three are generally pre-installed on most Linux/Unix operating systems.

In addition, several Perl libraries from CPAN are needed and described in the cpanfile. The quickest way to satisfy the Perl prerequisites is by running:

make bundle

from within a clone of the git repository. This command will download cpanm and use it to install the required dependencies into the inc/ directory. The dwindling-reads program will automatically use libraries in this directory if they exist.

dwindling-plot

Requires:

You can generally just install the two R packages like so:

Rscript -e 'install.packages(commandArgs(T))' ggplot2 Cairo

To install them system-wide, use sudo Rscript.

About

A tool to summarize the NGS read count stats at each stage of a pipeline

License:Other


Languages

Language:Perl 85.6%Language:R 10.8%Language:Makefile 3.6%