Hostile removes host sequences from short and long read (meta)genomes, consuming single or paired FASTQ from files or stdin. Batteries are included – a human reference genome is downloaded when run for the first time. Hostile is precise by default, removing an [order of magnitude fewer microbial reads](https://log.bede.im/2023/08/29/precise-host-read-removal.html#evaluating-accuracy) than existing approaches while removing >99.5% of real human reads from 1000 Genomes Project samples. For best possible retention of microbial reads, optionally use an existing index masked against bacterial and/or viral genomes, or make your own using the built-in masking utility. Read headers can be replaced with integers (using `--rename`) for privacy and smaller FASTQs. Heavy lifting is done with fast existing tools operating on a stream. In benchmarks, bacterial Illumina reads were decontaminated at 32Mbp/s (210k reads/sec) and bacterial ONT reads at 22Mbp/s, using 8 alignment threads. In typical use, Hostile requires 4GB of RAM for decontaminating short reads (Bowtie2) and 13GB for long reads (Minimap2). Further information and benchmarks can be found in the [paper](https://doi.org/10.1093/bioinformatics/btad728) and [blog post](https://log.bede.im/2023/08/29/precise-host-read-removal.html). Please open an issue to report problems or otherwise [reach](https://bsky.app/profile/bedec.bsky.social) [out](mailto:b@bede.im) for help and advice, and please cite the paper if you use Hostile in your work.
0 commit comments