Pararead is a python package that simplifies parallel processing of DNA sequencing reads (BAM or SAM files), by parallelizing across chromosomes. Pararead is built for developers of python scripts that process data read-by-read. It enables you to quickly and easily parallelize your script.
Pararead is hosted on pypi. Install with:
pip install --user pararead
Or, within active environment:
pip install --upgrade pararead
In the examples folder you can find count_reads_basic.py, which will count the number of reads in a sam/bam file in parallel.
Run this on your bam file like this:
python examples/count_reads.py file.bam -O output.txt --cores 2
Look at the code to see how this is implemented.
The main model provided is an abstract class called ParaReadProcessor, for which concrete children are created by implementing a __call__ method. This creates a callable instance that is then mapped over chromosomes.
The concept is generally described in this early blog post, which initiated the project that eventually became pararead. More details will be forthcoming.