From the original repo:
A multi-threaded (~180% cpu for compressed files when pairs are input) tool to remove filtered fastq reads.
The tool looks for filtered reads in either read of a pair, and will then remove both reads of the pair.
Memory usage is tunable. Approximately 250M RAM per 100,000 read pairs.
cwl included.
To build:
chmod +x
./build.sh