A collection of Python tools for filtering and visualizing Next Generation Sequencing reads.
Runs under Python 3.6 to 3.9 (though 7 of 1682 tests fail under 3.7 due to the unittest
module and mocking of open
).
Change log
On Linux (at least) you will need to first:
sudo apt install zlib1g-dev libbz2-dev liblzma-dev
Then you should be able to install via
$ pip install dark-matter
Some additional information is available for Linux, OS X, and Windows.
Not all of the following are mandatory. It depends which part of the dark-matter code you try to use.
run-bowtie2.py
can use Picard
to mark duplicates.
run-bowtie2.py
uses gatk if you call
it with --markDuplicatesGATK
or --callHaplotypesGATK
. If you want to do
this you'll need to download the gatk
zip file, unzip it, and either put
the directory where you unzip it into your shell's path or else move the
contents of the zip file into a directory already in your path.
You'll need
bcftools if you
want to make consensuses using the make-consensus.py
script. Either
follow
the installation instructions
or, if you use brew (or Linux brew), brew install bcftools
.
The run-bowtie2.py
and run-bwa.py
scripts both make use of
samtools.
The compare-sequences.py
script (if called with --align
) requires the
needle
and (possibly) stretcher
from
EMBOSS.
If you are using dark matter in an
iPython Notebook, you should install
the requirements in requirements-ipynb.txt
.
If you are using dark matter as a developer, you should install the
requirements in requirements-dev.txt
.